Studies on the Reduction Behavior of Iron Oxide Pellet Fines with Hydrogen Gas: Mechanism and Kinetic Analysis

The present study contributes to the current worldwide activities aiming to replace fossil carbon in steel making processes with hydrogen causing considerable reduction of greenhouse gas emissions. Compacts prepared from iron oxide pellets fines were isothermally reduced in pure hydrogen gas and a mixture of hydrogen and argon in the temperatures range from 700 to 1100 °C. The total weight loss produced during the reduction process was continuously recorded using thermogravimetric analysis (TG) technique. The findings demonstrated that the temperature has a considerable impact on the conversion and reduction rates. At a given temperature, the reduction rate was accelerated as the amount of H2 increased in the reducing gas. The results indicated that H2 content does not have an effect on reduction behavior, when it is higher than 80%. The reduction reaction of samples was shown to takes place in a step wise manner from hematite to metallic iron. The reduction kinetic and mechanism were deduced from the application of mathematical models and the morphological structure of the reduced samples and correlated with the apparent activation energy (Ea) values. The Ea values at the early, intermediate and final stages were 16.36, 29.24 and 49.35 kJ/mole, respectively. The early stage of the reduction process was controlled by chemical reaction, whereas the gaseous diffusion was controlled the latter stage. At the intermediate stage, the reduction process was controlled by mixed mechanism of gaseous diffusion and chemical reaction.


Introduction
Carbon dioxide emissions are currently one of the most serious environmental challenges, with one of the consequences being global warming. The iron and steel industries contribute for approximately 7% of total world CO 2 emissions [1]. On average, each ton of crude steel produced emits about 1.9 tons of CO 2 . As a result, the iron and steel sectors are exploring novel methods to minimize CO 2 emissions while also making the iron and steel manufacturing process more environmentally friendly and economically viable [2]. Steelmaking companies are setting for themselves a target to become a low-carbon or carbon neutral industry. The use of hydrogen as a reducing agent [3][4][5][6], carbon capture storage and utilization [7][8][9][10] biomass as a reducing agent [11,12] are accredited as possible ways to decrease fossil-based CO 2 emissions. Hydrogen is a highly effective reducing agent and a clean energy source, making it a promising energy medium for the twenty-first century. Its ability to diffuse and reduce makes it ideal for improving the thermodynamic and kinetic conditions of iron oxide reduction, resulting in faster reduction rates. In addition, the lower viscosity and volume of H 2 reduce the density and viscosity of the gas, leading to better thermal energy utilization. One of the methods explored to reduce CO 2 emissions is producing Direct Reduced Iron (DRI) using green hydrogen. Also, Injecting hydrogen directly into the blast furnace is a key step towards achieving low-carbon ironmaking and green sustainable development. However, several challenges are associated with replacing fossil fuels with hydrogen, including cost, transportation, and storage. Moreover, reducing iron ores with hydrogen is an endothermic process requiring higher energy consumption [13][14][15].
The International Energy Agency (IEA) provides some important data [16] about the Direct Reduced Iron (DRI) production sustainable development scenario up to 2070; some of these data are excerpted and given in Table1.
It can be seen that DRI production by 100% H 2 technology is expected to be a major key player for achieving the goal of low-carbon or carbon neutral steel industry.
Carbon monoxide and hydrogen both have the ability to reduce iron oxides where hydrogen reduction generates water vapor rather than carbon dioxide. Until the moment, the reduction of iron oxide in direct reduction (DR) plants is accomplished with a combination of H 2 and CO by reforming the natural gas. Hence, it looks that the only reducing agent used by DR plants going forward will be hydrogen gas by making some modifications and the technical and economic aspects is considered. In addition, the production of hydrogen from renewable resources like biomass makes the procedure more environmentally friendly favorable. In light of this, hydrogen-based iron ore reduction has gained a lot of interest recently [17,18].
Currently, researchers are investigating the utilization of hydrogen gas rather than metallurgical coke in iron-making. Three Swedish companies are researching the utilization of H 2 gas in iron and steel manufacturing process. This process is known as the Hydrogen Breakthrough Ironmaking Technology (HYBRIT). The goal of this approach is to completely replace metallurgical coke with hydrogen gas [19,20]. Given the growing need for DRI in the steelmaking process, more focus is being placed on promoting the research of the DR process [21]. For example, investigating the reduction kinetics of iron oxide in gases like H 2 , CO, and CO/H 2 gas mixtures.
Pritesh Garg et al. [22] conducted a study on the reduction of iron oxide pellet powder using different H 2 /H 2 O gas mixtures at temperatures ranging from 1023 K to 1373 K. They used a quantitative steam generator setup with a thermo-gravimetric analysis (TGA) setup to control the use of H 2 O in the reducing gas mixture. The study analyzed the effect of different mole fractions of H 2 and H 2 O on the reduction kinetics for the global reduction of Fe 2 O 3 . The results showed that the reaction was controlled by a chemical reaction in the first and second stages when the H 2 O mole fraction was lower. However, when the H 2 O mole fraction was higher (20%), the reaction became diffusion-controlled towards the end of the second stage. The apparent activation energy for the second stage increased with increasing H 2 O content in the reducing gas. Stepwise reduction experiments were also conducted by controlling the H 2 mole fraction at 50% and 90% to decouple the Fe 2 O 3 -Fe reaction.
Daniel et al. [23] study the kinetics of hydrogen reduction of iron ore fines between 873 K and 1073 K. They determined that the reduction kinetics cannot be represented by a single simple gas-solid reaction model. The rate controlling mechanisms depend on the temperature and level of reduction. The reduction reaction takes place in three steps, Piotrowski et al. [24] studied the influence of gas composition on the kinetics of iron oxide reduced in hydrogen atmosphere. They proved that the reduction rate rises with increasing temperature and the amount of H 2 in the gas mixture. The results reveal that the mechanism controlling the reduction process switches from being surface-controlled to being gas diffusion-controlled when a thin layer of magnetite and wustite is produced on the surface.
The effect of H 2 -H 2 O on the isothermal reduction of iron ore pellets was studied at temperatures between 700 and 1100 °C [25]. The performance of a reducing gas mixture composed of CO-CO 2 -H 2 -H 2 O-N 2 was compared to one containing simply CO-CO 2 -N 2 . The findings showed that raising the reduction temperature resulted in a greater reduction level in both reducing atmospheres. As the temperature increased, the porosity and surface areas also increased, which improved the reduction process. Reducing gases, with H 2 and H 2 O, enhanced porosity and surface areas in the reduced pellets, resulting in better reduction efficiency. However, gas compositions with over 2% water vapor at 800°C and 900°C did not result in a significantly better reduction performance than a CO-CO 2 -N 2-reducing atmosphere, as the water vapor may have obstructed the active sites of wustite Metolina et al. [26] investigate the reduction of industrial hematite pellets in hydrogen atmosphere in order to evaluate the impact of process factors on overall reduction reaction. The findings revealed a significant interaction between the pellet size and temperature, which had a major effect on the reduction behavior. The findings revealed that the pellet size and temperature had a major effect on the reduction process, with a significant interaction between them. As the temperature rises and the size of the particles gets smaller, the rate of reduction dramatically increases. In contrast, within the area covered by this investigation, porosity had no significant influence on the process. The pellet volume stays nearly constant in the temperature range from 600 to 1000 °C. A significant rise in pellet volume was seen at 1000, which was connected with the development of cracks. Yong Lee et al. [27] study the non-isothermal reduction kinetics of Fe 2 O 3 nano powder agglomerates in H 2 gas at different heating rates. They concluded that, during the reduction reaction the phase change happened in the order of Fe 2 O 3 to Fe 3 O 4 and Fe 3 O 4 to Fe. Based on the non-isothermal (TG) measurements, the Kissinger-Akahira-Sunose (KAS) pattern was applied to determine the apparent activation energy E a . The sintering of the oxide particles during the reduction stage (from Fe 3 O 4 to Fe) slowed down the reaction rate, which abruptly increased the activation energy. The catalytic effect of the produced Fe particles in promoting reduction of the surrounding oxide particles is what caused the reaction rate to accelerate as the grain of metallic iron grew.
El-Geassy et al. investigated the reduction behavior of iron oxide in H 2 /CO gas mixtures at temperatures ranging from 800 to 1100 °C [28]. They demonstrated that, the reaction mechanism was shown to be highly influenced by the original structure (in particular, the porosity and grain size) of the briquette and the composition of the gas. At the initial stage, in case of reduction of dense briquettes with H 2 gas the reduction reaction is chemically controlled. When the reduction was carried out using either CO or CO/H 2 mixtures, the reduction process was regulated by a combination of chemical reaction and gaseous diffusion, with the gaseous diffusion contributing more as the amount of CO increased in the reducing gas.
Nowadays, one of the main problems confronting global steel production is the management of wastes including mill scales, dust, sludge, and fines produced at various steps of the iron and steel sectors. The generation of pellet fines during palletization, transportation and handling of the pellets still one of the major challenges in steelmaking process. Currently, a considerable amounts of iron bearing by-products are produced during DRI production, electric arc furnace, casting, and rolling processes at steel manufacturing companies. In 2020, the amounts of produced fines all over world was estimated by about 20Mt. Utilizing these wastes is an important method for resource conservation and pollution reduction [29][30][31][32].
Recently, hydrogen has emerged as a key issue in both the global steel industry and academic fields. It provides a viable option that might be highly efficient in reducing CO 2 emissions as well as coke consumption in the iron and steel-making by optimizing the use of hydrogen and hydrogen-rich sources like natural gas and biomass in the iron and steel sector. Such solutions need extensive research in order to understand and properly assess their impact on the iron and steelmaking and decreasing CO 2 emissions. Understanding kinetic analysis is beneficial in the DR process by hydrogen gas.
Despite there have been numerous studies on the reduction of iron oxides using hydrogen, there is still a need for a better understanding of the reaction mechanisms because there are many variables involved, such as the temperature at which reduction occurs, the composition of the reducing gas, the composition of the iron oxide or iron ore, porosity, particle size, etc. This study focuses on the reduction of fines generated as waste from industrial pellets used in DR plants using an H 2 -Ar gas mixture in laboratory experiments conducted at temperatures from 700 to 1000 °C. Subsequently, the characterization of reduced samples is performed to evaluate the influence of the different factors on the reduction process and predict the reduction kinetics and mechanisms.

Raw Materials
The iron oxide pellets fines utilized in this investigation were obtained from a DR plant (Ezz El-Dekheila Steel Company, Alexandria, Egypt) as a waste material generated during the transportation and processing of pellets.

Compacts Preparation
Iron oxide pellets fines were moistened with 6% distilled water, then similar weights compressed at 20 kg/cm 2 in a cylindrical mould with a hydraulic press to produce compacts of similar size and shape. The formed compacts were dried at 110 °C for 24 h and stored in a desiccator for subsequent characterization.
A preliminary reduction experiment at 1100 °C using H 2 gas until the weight of the sample reached a constant was used to determine the total reducible oxygen content (w o ) of iron oxide pellet fines. (Eq. 1) The extent of the reduction at time t (R t %) was then estimated using Eq. 2.
where w i is the initial sample weight, w f is the final weight of sample, w t is the sample weight at time t, and w o is the total oxygen content.

Characterization Tools
The chemical composition was analyzed by X-ray fluorescence (XRF Advanced Axios, PAN analytical, Netherland). The different phases in iron oxide pellets fines and in the reduced samples were identified by X-ray diffraction technique (XRD, PAN Analytical Empyrean, Eindhoven, Netherlands) PW 1730 equipped with Cu radiation (λ = 1.5Ǻ, at 40 kV and 30 mA). The morphology and micro-structures of samples were examined by both reflected light microscope, RLM (Zeiss Axio-scope A1), and Scanning Electron Microscope, (SEM) (JEOL-JSM-5410, 10 kV, Japan). XPS analyses were conducted using K-Alpha X-ray photoelectron spectrometer, Thermo Fisher Scientific (USA), provided with a monochromatic Al Kα X-ray source. The instrument was operated at a pressure of 1 × 10 -7 Pa. The wide survey scans were obtained at pass energy of 200 eV to determine the elemental composition on the mineral surfaces. The narrow high-resolution scans were obtained at pass energy of 50 eV. The data were processed using the A vantage software. The chemical composition and the different phases of the sample were analyzed using X-ray fluorescence (XRF) and X-ray diffraction (XRD). The results are illustrated in Table 2, and Fig. 1, respectively. The main phase present in the sample was hematite (Fe 2 O 3 ), accounting for approximately 96 wt%. The particle size distribution of the iron oxide pellet fines is given in Table 3, revealing that more 60% of the sample was composed of particles in the < 100 μm size range.

Reduction System
Thermogravimetry (TG) monitors the reaction behavior of samples is represented schematically in Fig. 2. The major components of the reduction apparatus include a gas supply system, a vertical tube furnace, fitting with an alumina reaction tube, an automated sensitive balance and temperature controller. The output of the balance is coupled to the recording system for continual weight loss monitoring. The control parts consist of two different thermocouples (T1 and T2). The thermocouples T1 and T2 are used to measure the furnace and the sample temperatures, respectively. A gas supply system and gas purification unit is used to remove moisture and other contaminants from the gases.

Experimental Procedures
In each experiment, a compact (iron oxide pellets fines) is heated in argon atmosphere with a flow rate of 500 ml/ min from room temperature up to the target temperature and held at that temperature for around 5 min. Under these conditions, the Ar is then replaced by H 2 and keeps the same flow rate of Ar gas. Throughout the experimental tests the weight loss of samples was continually measured by thermogravimetric analysis (TG) technique until the weight reaches a constant value demonstrating the completion of the experiment. After the experiment is finished, H 2 gas is switched off, and Ar gas is introduced again while the furnace is cooled down to the room temperature. The reduction product is removed and stored in a desiccator for subsequent analysis. Figure 3 illustrates the effect of gas composition on the isothermal reduction behavior of samples at 900 °C. It can be shown that the reduction extent and the reaction rate significantly affected with the increasing of H 2 concentration in the gas mixture, which is consistent with the experimental and numerical results obtained by other researchers [33][34][35]. Figure 3 indicates that in the case of pure H 2 , the reduction reaction is completed at approximately 40 min and reach approximately 100% reduction extent. It can also be noticed that for all reduction curves obtained for using gas mixtures with different H 2 contents; the rate of reduction starts rapid and slightly decelerate, whereas at about 15 min of reduction a considerable slowdown in reduction rate was observed. Table 4, shows the reduction extents that were achieved after 15 min of reduction using a mixture of reducing gas with different H 2 contents. It is obvious that as the H 2 content in the gas mixture increases the achieved reduction extent increased. By increasing the H 2 content more than 80%, no considerable improvement in the reduction extent was noticed. This indicates that H 2 content > 80% has no influence on the reduction behavior.

Effect of H 2 Content on the Reduction Behavior
The relationship between the rate of reduction and reduction fraction for different samples reduced at 900 °C is shown in Fig. 4. The reduction rate is calculated by the first derivative of the reduction fraction as a function of time. The reduction reaction can be divided into three distinct stages. During stage I, the dotted line indicates the reduction   fraction of approximately 0.11, representing the theoretically expected reduction from hematite to magnetite. Stage II marks the reduction of magnetite to wustite, with a reduction fraction of about 0.33. Finally, stage III represents the transformation of wustite to metallic iron. In the first stage, the rate of reduction is highest and reaches about 30 mg/min. The reduction rate becomes smaller in the second stage and increases with increase the ratio of H 2 in the gas mixture. In the third stage, the reduction rate has the lowest value compared to the other two stages and it becomes slow with the increasing of reduction fraction. The reduction at other temperatures has a similar behavior to that obtained at 900 °C.

Effect of Temperature on the Reduction Behavior
The reduction curves of samples reduced with 80% H 2 -20% Ar gas mixture at temperatures from 700 to 1000 °C are shown in Fig. 5. Considering each curve; the reduction rate is greatest in the beginning stages and progressively decreases as the reaction progresses. At the initial stages, the calculated reduction extent was more noticeable up to a particular level dependent on the employed temperature. During the last stages, the rate of oxygen elimination was extremely slow exhibiting a considerable decrease in the reduction rate until the completion of reduction reaction resulting in a plateau shape. The reaction takes longer time at low temperatures, whereas the reduction time decreases as the temperature rises. Furthermore, the level of reduction that exhibited at this plateau increased with increasing temperature. The high reduction rate during the beginning and middle stages is caused by Fe 2 O 3 being reduced to Fe 3 O 4 , which is then reduced to FeO. The reduction rate slows down in the last stages due to the conversion of FeO to Fe.

Reduction Kinetics and Mechanism
The previous studies [5] indicate that the reduction of iron oxide is supposed to begin at the surface and progresses to the core and with time the diameter of the unreacted core decreases until it reaches zero. The reduction kinetics can be expressed by a shrinking, unreacted core model to explain the reduction process. Initially, the assumption is that iron ore particles are entirely composed of hematite. As the reduction progresses, an iron layer is formed, and the unreacted core of the pellet particle shrinks in diameter. Iron ore particles have a layer of reactant gas at their surfaces. External diffusion, internal diffusion, and chemical reaction are the three steps of the reduction reaction.
1. Reactant gas components diffuse to the surface of pellet from the main gas phase through gas boundary, which is named external diffusion. 2. Gas component adsorbs on the surface of pellet by physical, chemical methods and diffuses to the reaction front interface, which is called internal diffusion. 3. Gas component has chemical reaction with pellet at the front interface and produces oxidation product of gas. 4. The gas oxidation product diffuses back to the surface of pellet (internal diffusion) and transfers to the main gas phase via the gas boundary (external diffusion).
In order to predict the limiting mechanism in the gas-solid chemical reactions of iron ore reduction, the mathematical models were tested against experimental results and correlated with the apparent activation energy values, and the morphological characteristics of the samples [36][37][38]. where K represents the reaction rate constant, A is the pre-exponential factor, E a denotes the apparent activation energy, R refers to the gas constant, and T is the absolute temperature.
The rate of reaction (K) is obtained from the reduction curves illustrated in Fig. 5. The correlations between lnk and (1/T) are displayed throughout the early, intermediate, and the last stages of reaction, as given in Fig. 6 The E a value during the early stages suggests that the reduction reaction seems to be chemical reaction control, while the E a value demonstrates that the gaseous diffusion is limiting mechanism in the later stages. At the intermediate stage, the reduction process is controlled by mixed mechanism of gaseous diffusion and chemical reaction. Other studies have found similar results in reduction experiments [39,40].
One of the mathematical equations that are frequently used for deducing kinetics and mechanisms in gas-solid chemical reactions is the grain model proposed by Szekely et al. [39]. This study determined the rate-controlling step in the whole reduction reaction by testing the mathematical formulations for gaseous diffusion, chemical reaction, and mixed control. The following are the three mathematical equations [39][40][41]: (a) for gaseous diffusion mechanism ∶ K = X + (1 − X) where K represents the reaction rate and X refers the reduction fractional.
In order to predict the limiting mechanism in the gas-solid chemical reactions of iron ore reduction, the abovementioned mathematical equations were applied against experimental results and correlated with the E a values to confirm the validity of reduction mechanism as presented in Figs. 7, 8 and 9. When the chemical reaction formula was tested in the early stages, linear correlations were observed as given in Fig. 7. During the intermediate stage, linear correlations were achieved using the mixed control mechanism as depicted in Fig. 8. At the final stages, linear correlations were acquired by the application of a gaseous diffusion mechanism as shown in Fig. 9. These match the mechanisms proposed by the E a values.
In order to investigate the reduction mechanism, the isothermal reduction temperature 900 °C was selected and the reduction reaction was interrupted at different reduction extents; 10, 25, 50, 75 and 100%. The reduction extent was calculated relative to the maximum mass loss obtained after 60 min. The samples subjected to interrupted reduction were cooled down to room temperature in Ar atmosphere to terminate the reaction progress and prevent the oxidation. The reduced samples were characterized with XRD and XPS to determine the phase transformation during the reduction process, while their microstructures evolution was observed by optical microscopy and SEM-EDS.
(5) (b) for chemical reaction mechanism: K = 1 − (1 − X) 0.5 (6) (c) for mixed control mechanism: 6 Arrhenius plots for samples reduced by 80%H 2 -20% Ar Fig. 7 Application of mechanism control mathematical formulations at the initial stages of reduction with 80%H 2 -20% Ar The XRD patterns of partial reduction mode of the various samples are presented in Fig. 10. In the beginning of the reduction reaction, after 10% and 25% reduction extents hematite and magnetite were clearly the predominant phases in the samples. As the reduction reaction progressed and the reduction extent becomes 50%, the main phases are magnetite and wustite. This indicates the reduction of hematite to magnetite according to the reaction (Eq. 7) and also the appearance of wustite phase behind magnetite phase is a result of magnetite reduction to wustite according to reaction (Eq. 8).
Increasing the reduction time enhances the reduction of magnetite to wustite and reduction of wustite to metallic iron according to reactions (Eqs. 8 and 9) as noticed in the case of 75 and 100% reduction extents.
When the reduction extent reaches about 75% only wustite and iron phases can be detected. The intensity of wustite phase decreases gradually by increasing the reduction time till the complete reduction 100% achieved where only metallic iron phase can be detected. The reduction reaction was noticed to be occurring in a step wise manner from hematite to metallic iron (Fe 2 O 3 -Fe 3 O 4 -FeO-Fe); this was consistent with the findings reached by other researchers [23,42].
The Fe 2 O 3 , Fe 3 O 4 , FeO and Fe contents of partially reduced samples was detected by quantitative XRD analysis using Rietveld refinement [43],  Also, XPS measurements were used to explore the surface oxidation state of reduced samples interrupted at different reduction extents and to confirm the reduction mechanism suggested by XRD. Figure 11 shows the XPS spectrums of Fe at different reduction extents. The position of XPS peaks of Fe 2p 3/2 for different oxidation state of iron; Fe +3 , Fe +2 , Fe 0 has been investigated by many researchers [44][45][46] and the position of these peaks is between 706.6 and 711.2 eV according to valence state of iron. From Fig. 11 it can be seen that at 10% reduction extent the XPS peak of Fe appeared at 710.7 eV which indicates that the major valence state of this sample is Fe +3 of hematite. Also, two satellite peaks are clearly detected at 719.8 and 732 eV which are characteristic to hematite [44]. While increasing the reduction extent up to 50% enhances the formation of FeO and this is confirmed by the shift of the peak from 710.7 to 710.5 eV and appearance of two satellite peaks at 718.4 and 731.8 eV [45]. With further 75% reduction extent the peaks shifted from 710.7 to 710.3 eV. The complete reduced sample (100%) possess a XPS peak at binding energies of 706.3 which is well matched with zero valent iron (i.e. metallic iron) [46].
Based on the phase composition resulted from XRD and XPS analysis for the interrupted reduced samples at different reduction extents, the reduction reaction was noticed to be occurring in a step wise manner from hematite to metallic iron. (Fe 2 O 3 -Fe 3 O 4 -FeO-Fe).
The phase transformation and the morphological features of the partially reduced samples at 900 °C were examined by the reflected light microscope (RLM), as shown in Fig. 12a-f. Figure 12a shows that the reduction occurred in a topochemical mode, forming three distinct layers (outer, middle, and core layers). At the beginning of the reduction process, the reducing gas was adsorbed on the surface of Fe 2 O 3 , reducing it to Fe 3 O 4 , which was then reduced to wustite, and finally to metallic iron, forming a relatively porous structure due to the removal of oxygen, promoting the gaseous diffusion process. Near the surface, the resistance to gas mixture diffusion was minimal, and therefore, the chemical reaction resistance had a significant effect as a controlling mechanism. Figure 12b depicts a sharp interface between the middle and the core layer, and the compact's outer layer, comprised of the iron-wustite phase, diffused in the wustite grains. Consequently, iron was nucleated and grew on the surface of wustite particles, allowing for the development of the iron-wustite intergrowth layer. As the diameter of metallic iron layers around the wustite grains increased, more resistance to the reducing gas diffusion was built up. As a result, the contribution of gaseous diffusion in the reduction mechanism increased. Thus, it can be concluded that the rate-controlling mechanism at the intermediate reduction stage is a combined effect of chemical reaction and gaseous diffusion mechanisms. Figure 12c illustrates that the outer layer consists of metallic iron and wustite with many macropores, allowing good access to reducing gas to the inner layers. The reducing gas readily diffused via the outer shell, so iron was nucleated and grew up. As the reduction proceeded, the metallic iron grains increased and connected to each other, and the structure included small grains of wustite together with metallic iron, as observed in Fig. 12d. As the reduction time increased, the internal structure became denser due to the coalescing and integration between metallic iron grains, as observed in Fig. 12e. This caused the reaction rate to slow in the final stage, indicating maximum resistance to gas diffusion. Therefore, gas diffusion controls the reduction process during the final stages. The SEM images of samples reduced at a temperature 900 °C for various times are shown in Fig. 13a-e. The morphological structure of reduced samples shows a gradual change of the grains with the progress of the reduction reaction. Initially, the internal structure indicates the formation of a relatively porous structure due to the removal of oxygen which promotes the process of gaseous diffusion. The main intermediate phase was magnetite according to XRD analysis. The reduction from hematite to magnetite and even to wustite resulted in the formation of many pores due to the loss of oxygen in lattice. The grains gradually become large and then the structure becomes relatively dense and less porous was observed as the reduction reaction proceeded. With further reduction, the amount of metallic iron (may contain some wustite) increased and coalesced to each other resulting in the formation of dense structure. The slowing down of the reduction rate in this stage resulted from the formation of a dense outer layer, which hindered the diffusion of reducing gas.

Conclusion
Nowadays, researchers are interested in utilizing hydrogen as a reducing agent to decrease the amount of CO 2 emissions in the iron and steel sector. Additionally, investigating the kinetics mechanism of the reduction processes is very important to determine the precise material and energy consumption. This study investigated the isothermal reductions of rejected fines generated from iron oxide pellets used in DR plants in H 2 -Ar gas mixture. Based on the experimental Fig. 13 SEM images of samples partially reduced at 900 °C at different reduction extents: a 10%, b 25%, c 50%, d 75%, and e 100% results, the reduction kinetic and mechanism were deduced. The main conclusions are: 1. The rate of reduction and reduction conversion is greatly influenced by temperature. 2. At any applied temperature, the reduction rate increased as H 2 content increased. 3. H 2 content has no effect on reduction behavior when it is higher than 80%. 4. The reduction reaction was shown to proceed step wisely from Fe 2 O 3 to Fe (Fe 2 O 3 → Fe 3 O 4 → FeO → Fe). 5. At the early stages, the reduction process was controlled by chemical reaction, while in the intermediate stage, the reduction reaction was controlled by mixed gas diffusion and chemical reaction mechanisms. Gas diffusion was the rate-controlling mechanism during the latter stages. 6