Skip to content
Publicly Available Published by De Gruyter December 25, 2020

The electrochemical reduction mechanism of Fe3O4 in NaCl-CaCl2 melts

  • Hui Li , Lei Jia , Weigang Cao , Jinglong Liang EMAIL logo , Le Wang and Hongyan Yan

Abstract

In order to study the process of Fe3O4 reduction by melt electro-deoxidation. Electrochemical method was used to analyze the reduction mechanism of Fe3O4 in NaCl-CaCl2 melts. The effects of cell voltage and time on the product were discussed through constant cell voltage electrolysis. The results showed: (1) The reduction of solid Fe3O4 to metallic Fe is a two-step process for obtaining electrons. (2) The transformation process (600 min, 0–1.0 V) of the electrolysis products with the increase of the cell voltage is as follows: Fe3O4 → FeO → FeO + Fe → Fe. (3) The intermediate product Ca2Fe2O5 was formed (2.0 V, 10–300 min), which inhibited the deoxygenation process in the early stage of the reaction. When the electrolysis time exceeds 60 min, the main reaction is the reduction of Ca2Fe2O5 to Fe.

1 Introduction

The higher energy consumption and CO2 emissions in the traditional steel smelting process, so the development of new methods for preparing Fe is urgent (Dolganov et al. 2020). The Ultra-Low CO2 Steelmaking (ULCOS) program (including molten salt electrolysis) has been established in Europe (Allanore et al. 2007; Allanore et al. 2010). The first industrial-scale deployment of Boston Metal Molten Oxide Electrolysis (MOE, 1873 K) to produce ferroalloy technology has completed a series a financing of US $20 million to reduce CO2 emissions during steel production. Electric energy is relatively clean and pollution-free, so molten salt electrolysis has great potential in metal preparation.

The use of FFC Cambridge (Chen, Fray, and Farthing 2000) process to directly electrolyze Fe2O3 to prepare Fe has been extensively studied (D.Y. Zhang et al. 2020). The solid Fe2O3 had been used as material by Li GM (Li, Wang, and Chen 2009) to prepare Fe at 800 °C in pure CaCl2 melt. Parameters: Electrolysis voltage 1.8–3.2 V, electrolysis time 2–20 h. The reduction step is Fe2O3 → FeO → Fe. Fe3O4 was detected in the experiment, which is considered to be a combination of FeO and Fe2O3. The filling of Mo pores with Fe/Fe2O3 (8:1) was studied by Gao HP (Gao et al. 2013) as a material in 1108 K pure CaCl2 melt. The diffusion coefficient was 8.2 × 10−6 cm2 S−1 and the activation energy was about 67.8  kJ mol−1. Fe3O4 was not detected under the condition of constant potential and the reduction step was Fe2O3 → FeO → Fe. The studied the preparation of dendritic metallic Fe by Zou X L (Zou et al. 2015) and Haarberg (Haarberg and Yuan 2014) direct electrochemical reduction Fe2O3 in alkaline solutions. The Fe3O4 intermediate product was detected. The reduction step was Fe2O3 → Fe3O4 → Fe. According to the Fe-O phase diagram, Fe3O4 is a necessary process in the process of Fe oxidation. When the temperature is higher than 570 °C, the decomposition of Fe2O3 is carried out in three steps: Fe2O3 → Fe3O4 → FeO → Fe. At present, there are few reports on electrochemical experiments with solid Fe3O4 as material to research. Recently, our team studied the electrochemical behavior of Fe3+ in the NaCl-CaCl2 melts in the form of Fe3O4 dissolution. Fe3+ was reduced to Fe in two steps (Fe3+ → Fe2+ → Fe) (Li et al. 2019). The solubility of metal oxides in the melts is not high and the electrolytic efficiency is low (S.Q. Jiao et al. 2020). The results showed that the solubility of Fe2O3 in the NaCl-CaCl2 melts is about 1.5 wt% at 1000 °C (Hu et al. 2018).

In this paper, the solid Fe3O4 has been used as the material, and the reduction process is tested by electrochemical method in the NaCl-CaCl2 melts at 800 °C. The relationship between the conditions and the reduction process will provide theoretical support for the use of solid Fe3O4 to directly electro-reduce the green preparation of Fe.

2 Experiment

2.1 Materials and reagents

Pure NaCl (99.95%), CaCl2 (99.95%), Fe3O4 (99.95%), AgCl (99.5%) were dried in a vacuum drying oven at 260 °C for 24 h. High-purity graphite rods (Ф15 × 100 mm) were polished with 400, 1000, and 2000 mesh sandpaper in sequence, and the surface of the graphite was repeatedly washed with deionized water and absolute ethanol. The graphite rods were placed in a vacuum drying oven at 120 °C for 48 h. An Ag/Ag+ reference electrode was used for electrochemical experiments. NaCl, CaCl2, and AgCl with a molar ratio of 48: 48: 4 are filled into a aluminum silicate tube protected by Ar (99.999%). Subsequently, Ag wire (Ф0.5 mm, 99% purity) was placed in the tube after salt melting, and the outer part of the tube was connected with nickel wire.

Two working electrodes are used for electrochemical testing. (1) Metal hole electrode (MCE-Fe3O4): The electrochemical test working electrode uses a double-hole molybdenum bar that can be filled with Fe3O4. (Purity > 99.9%, the thickness is 0.5 mm, the width is 2 mm, and the length is 50 mm) (Rong et al. 2014). In order to avoid Fe3O4 falling off after being immersed in melts, it was sintered in a resistance furnace at 800 °C for 300 min to prepare an MCE electrode. (2) Fe Coated electrode (FCE- Fe3O4): Repeatedly immersing the stainless steel wire (Ф0.5 mm) into the ethanol suspension containing Fe3O4 powder to prepare an FCE electrode (3 g Fe3O4 placed in 5 mL ethanol ultrasonic vibration 15 min) (Tang et al. 2013; Yin et al. 2011).

The steel mold with a diameter of 15 mm was adopted for constant cell voltage electrolysis stainless, and presses Fe3O4 powder into a porous cylindrical sheet (8Mpa, 14.8 mm in diameter, 1.4 mm in thickness, 0.8 g in mass, 38% in apparent pore). As shown in Figure 1. It was sintered in a resistance furnace at 800 °C for 300 min, and the test piece was wrapped with 400 mesh stainless steel mesh and connected with a stainless steel iron rod (Ф5 mm). An alumina crucible containing 200 g of equimolar NaCl-CaCl2 (Ф55 mm × 100 mm, purity 99.5%) was placed in a resistance furnace at a heating rate of 5 °C/min. After the temperature in the furnace was raised to 300 °C and kept for 12 h to remove residual moisture, the temperature was then raised to 800 °C and kept for 2 h to prepare for the experiment. The nickel plate cathode and the graphite rod anode constitute a pre-electrolysis loop, and 2.6 V electrolysis is applied for 12 h to remove some trace water, Zn, Mn and other elements. The above experiments were conducted under an atmosphere protected by Ar (> 99.999%).

Figure 1: (a) Metal molybdenum strip after warp cutting; (b) MCE-Fe3O4 working electrode and cathode before sintering.
Figure 1:

(a) Metal molybdenum strip after warp cutting; (b) MCE-Fe3O4 working electrode and cathode before sintering.

2.2 Experimental methods

The CHI660E electrochemical workstation was used for experimental testing. Experimental electrode: working electrodes FCE-Fe3O4 and MCE- Fe3O4 (WE immersion depth of 2–3 mm), counter electrode graphite (CE immersion depth of 30–35 mm), reference electrode Ag/Ag+ (RE immersion depth of 3–4 mm). Cyclic voltammetry (CV) and square wave voltammetry (SWV) experiments were carried out to study the reduction mechanism of Fe3O4. The GWINSTEK PSM-3004 DC power supply was used for constant compliance voltage electrolysis, the compliance voltage is controlled between 0.3–3.0 V, the whole experiment process is carried out in an atmosphere filled with argon after vacuuming. When the reduction reaction is completed, the electrode is pumped away from the melts and cooled with the furnace under the protection of argon. The electrolytic product was cleaned in distilled water for 60 min with ultrasonic waves, then dried in a drying cabinet at 110 °C for 120 min. After drying, it was characterized.

Due to the accuracy, purity, oxide film, adsorption of atoms and molecules of the electrode surface, it has a great influence on the electrode potential. The electrode potential can be changed up to 1 V. Therefore, in the electrochemical test, the electrode surface is activated by the method of large scan speed and multi-turn scanning. Figure 2 (a) is the open circuit potential diagram (OCP-t) when the MCE electrode is immersed in the NaCl-CaCl2 melts interface and flows into Ar gas. When the working electrode contacts the melts interface, the open circuit potential stabilizes. When argon was blown in, there was a slight fluctuation, and then the open circuit potential slowly decreased and finally stabilized at about −0.077 V. We can judge the depth of immersion into the liquid surface through this phenomenon. Figure 2 (b) potential time diagram is a polarization experiment to test the reversibility of the MCE electrode. After the external conductor is discharged by short circuit for 1 s, use an external power supply to charge for 1 s. The working electrode reacted quickly and returned to the open circuit potential value within 3 s. After the open circuit potential is stabilized, use the AC impedance method to measure the total resistance of the melts. Experimental parameters (range and frequency: 105–10 Hz, 121, amplitude 0.025 V, applied voltage is the open circuit potential value). Figure 2 (c) is the Nyquist diagram of the AC impedance under the NaCl-CaCl2 melts. The impedance at the high frequency end is 1.41 Ω, which can be used as positive feedback compensation in electrochemical experiments.

Figure 2: Electrochemical performance test of MCE electrode in NaCl-CaCl2 melts (a) Open circuit potential diagram (OCP-t) (b) MCE electrode reversibility test (c) AC impedance Nyquist diagram.
Figure 2:

Electrochemical performance test of MCE electrode in NaCl-CaCl2 melts (a) Open circuit potential diagram (OCP-t) (b) MCE electrode reversibility test (c) AC impedance Nyquist diagram.

Figure 3: SWV at different frequencies (1–50 Hz) pure salt and filling Fe3O4 in NaCl-CaCl2 melts at 800 °C, WE: Mo, RE: Ag/Ag+, CE: graphite rod.
Figure 3:

SWV at different frequencies (1–50 Hz) pure salt and filling Fe3O4 in NaCl-CaCl2 melts at 800 °C, WE: Mo, RE: Ag/Ag+, CE: graphite rod.

2.3 Characterization test

The product and morphology were analyzed by scanning electron microscope JEM-2800F, energy spectrum, X-ray diffraction spectrum (XRD, X-ray 6000 with Cu Kα1 radiation λ = 1.5405 Å, scanning speed 10°/min).

3 Results and discussion

3.1 Square wave voltammetry (SWV) of MCE-Fe3O4

Square wave voltammetry is highly sensitive and can effectively suppress the problem of capacitive background currents that occur in cyclic voltammetry at high scan rates. By judging the reversibility of the reaction, the number of transferred electrons of the reaction can be calculated (Weng et al. 2017). A square wave voltammetry test was carried out in a NaCl-CaCl2 melts at 800 °C with MCE electrodes filled with 1 mg Fe3O4 and without filled Fe3O4 as working electrodes. The test parameters are as follows: step potential 1 mV, amplitude 30 mV, scanning frequency 1–50 Hz. The results are shown in Figure 3. The reduction peaks R1 and R2 appear near −0.77 V and −1.11 V. As the scanning frequency increases, the peak potentials of the reduction peaks R1 and R2 remain basically unchanged.

Figure 4 (a) is the peak shape fitting of the SWV curve at f = 20 Hz, and W1/2 is obtained. The reversibility of the reaction is judged by the relationship i ∼ f1/2 and the change of the peak potential with the frequency in Figure 4 (b).The current density of the reduction peak R1 has a linear relationship with the square root of the frequency, and the number of electron transfers of the reduction peak R1 is calculated by Eq. (1) (Li et al. 2018; Y.K. Wu, Chen, and Wang 2020), n ≈ 1.85. It shows that the peak R1 corresponds to the reduction of 2 electrons obtained by Fe3O4 to FeO, and the peak R2 is the reaction of transferring 2 electrons from FeO to Fe. The reaction equation is shown in 2–3.

(1)W1/2=3.52×RTnF

R (8.314 J·mol−1 K−1) is molar gas constant, T (K) is temperature, n is electron transfer number, F (96485C·mol−1) is Faraday constant.

(2)Fe3O4+2e=3FeO+O2
(3)FeO+2e=Fe+O2
Figure 4: (a) Peak shape fitting graph of square wave voltammetry curve (f = 20 Hz) (b) The relationship between the peak current densities of R1 and R2 the square root of frequency.
Figure 4:

(a) Peak shape fitting graph of square wave voltammetry curve (f = 20 Hz) (b) The relationship between the peak current densities of R1 and R2 the square root of frequency.

The anode reaction of graphite is as follows:

The O2− removed from the cathode is discharged at the anode to produce CO2 (Liu et al. 2018). Part of CO2 is released, and a part of CO2 is dissolved into NaCl-CaCl2 melt to form CO32−, which combines with Ca2+ in melt to produce CaCO3.

3.2 Cyclic voltammetry (CV) of MCE-Fe3O4 and FCE-Fe3O4

Figure 5 (a) is the CV of MCE and MCE-Fe3O4 electrode with increasing scan rate (0.1–1.0 V·s−1) in NaCl-CaCl2 melts at 800 °C. The scanning range is −2.4–0.6 V. The CV of the MCE electrode shows an oxidation and reduction peak A’/ A corresponding to Na+/Na near −1.9/−2.4 V. In the forward scanning process, the peak current during Na oxidation is lower than that during Na reduction, that is: Ipa < Ipc. This is due to the volatilization of Na in the melts at 800 °C. In the forward scan to around 0.6 V, the peak B’/ B is the redox peak of chlorine gas (Weng et al. 2016). In the range of −2.0–0 V, the background current is less than 0.1 A, there is no obvious redox peak, and it has good electrochemical stability. It can be used as the potential scanning window of MCE-Fe3O4. The cyclic voltammetry curve of the MCE- Fe3O4 electrode shows that the reduction peak R1, peak R2 and oxidation peak D is appear. It is suggested that the solid Fe3O4 is reduced to Fe in two steps at the cathode (Fe3O4 → FeO → Fe).

Figure 5: CVs of different scan rates of (a) MCE- Fe3O4 (b) coating FCE-Fe3O4 in 800 °C NaCl-CaCl2 melts.
Figure 5:

CVs of different scan rates of (a) MCE- Fe3O4 (b) coating FCE-Fe3O4 in 800 °C NaCl-CaCl2 melts.

It can be seen from Figure 5 (a) that the current also increases with the acceleration of the scan rate. When the scan rate increases, the time required to reach the peak potential is shortened, the thickness of the transient diffusion layer becomes thinner, the concentration gradient is larger, the diffusion rate is accelerated, resulting in an increase in peak current. In terms of Fe3O4 deoxygenation reduction, the faster scan rate prevents O2− from Fe3O4 transferred to the graphite anode in a short time, but remains at the cathode, forming a large concentration gradient and increasing the current. In the reverse scanning process, these O2− participate in the oxidation process of Fe, and a large oxidation peak current appears. When a smaller scan rate of 0.2 V·s−1 is used, the negative potential shift causes the over potential of the reaction to increase, which accelerates the electrochemical reaction rate, the ion concentration gradient at the solid-liquid interface decreases, and the current value decreases.

Figure 5 (b) is the CV of FCE and FCE-Fe3O4 electrode with increasing scan rate (0.01–0.05 V·s−1) in NaCl-CaCl2 melts at 800 °C. The CV curves of the FCE- Fe3O4 electrode showed peaks R1 and R2 not found in pure salts near −1.2 V and −1.6 V, respectively, corresponding to oxidation peaks D1 and D2. The reduction process of FCE-Fe3O4 at the cathode is similar to the reduction mechanism of the MCE working electrode. The scan rate has little effect on the potential of the peak R1, but the potential of the peak R2 is obviously shifted to the left. This is because the electrochemical reaction rate is slow at a small scan rate, and the ion diffusion rate is relatively fast. With the negative scan of the potential, a reaction with a wider peak shape and a smaller peak current appears. When the scan rate increases, the electrochemical reaction speed is faster. At this time, the ion diffusion rate is small, the reaction peak is sharp, and the peak current is large (Li et al. 2020). The negative shift of peak R2 is also affected by the electrochemical reaction speed of R1. When the reaction speed of R1 is fast, the O2− concentration at the three-phase interface is high, which will hinder the reaction in the second step.

3.3 The effect of voltage on Fe3O4 electrolysis products

The reduction mechanism of Fe3O4 in the NaCl-CaCl2 melts was studied, and the Fe3O4 sintered at 800 °C was subjected to constant cell voltage electrolysis (0.3–1.0 V). In order to avoid the decomposition of melts, according to the theoretical decomposition pressure, the calculation in Table 1 should be less than 3.2 V. Figure 6 is the XRD pattern of Fe3O4 after electrolysis at 600 min in the NaCl-CaCl2 melts at 800 °C. It can be seen from Figure 5 that after immersing in NaCl-CaCl2 melts for 600 min, Fe3O4 did not undergo a chemical reaction, but when the applied voltage was 0.3–0.4 V, Fe3O4 deoxidized and reduced to FeO; after applying 0.5 V for 600 min, it produced little metallic Fe. When the electrolysis voltage is further increased to 0.6 V, the diffraction peak of Fe is more intense, but there is still little FeO in the electrolysis product that has not been reduced. When the electrolysis voltage is 0.8 V, all FeO is reduced to metallic Fe.

Table 1:

Equilibrium electrode potential of Fe-containing compounds and chlorides at 800 °C.

□Reaction(T = 800 °C)ΔGΘ(kJ/mol)EΘ/V(vs. O2/O2−)
2Fe3O4 = 6FeO + O2(g)370.320−0.96
1/2Fe3O4 = 3/2Fe + O2(g)149.277−0.99
2FeO = 2Fe  +  O2(g)299.517−1.01
2CaO = 2Ca + O2(g)1045.428−2.71
2CaO·Fe2O3 = 4FeO + 2CaO + O2(g)378.910−0.98
2NaCl = 2Na + Cl2(g)575.887−3.24
CaCl2 = Ca + Cl2(g)635.340−3.29
Figure 6: XRD pattern of the electrolysis product of Fe3O4 electrolyzed at constant cell voltage at 800 °C of 600 min in NaCl-CaCl2 melts.
Figure 6:

XRD pattern of the electrolysis product of Fe3O4 electrolyzed at constant cell voltage at 800 °C of 600 min in NaCl-CaCl2 melts.

Cyclic voltammetry was used to analyze the behavior of the oxide but a graphite rod was used as the counter electrode during electrolysis. However, science the same graphite rod was used as the counter electrode during cyclic voltammetry, the behavior of the oxide should be consistent with the electrolysis process.

Figure 7 (a) is an SEM image of 0.3 V electrolysis for 600 min. From the figure, it can be seen that FeO has obvious particle boundaries, and the particle size is between 10 and 25 μm. Figure 7 (b–d) are SEM images of electrolysis for 600 min after applying 1.0–3.0 V cell voltage. When the electrolytic voltage of 1.0 V is applied, the size of the Fe particles is uniform, about 5 μm. With the increase of the electrolytic voltage to 1.5 V, the growth rate of reduced metal Fe is greater than the nucleation rate. It is Fe particle of about 10 μm. When an electrolytic voltage of 3.0 V is applied, it appears as more fine Fe particles with a diameter of about 1 μm, because the nucleation rate of reduced Fe is greater than the growth rate when a higher electrolytic voltage is applied. From the above analysis, it can be seen that the choice of electrolysis voltage can effectively control the generation and morphology of the product, and change the particle size and continuity of the Fe (Tang et al. 2016).

Figure 7: SEM images of Fe3O4 sample electrolyzed by constant cell voltage in NaCl-CaCl2 melts at 800 °C for 600 min(a) 0.3 V (b) 1.0 V (c) 1.5 V (d) 3.0 V (inset: EDS test data).
Figure 7:

SEM images of Fe3O4 sample electrolyzed by constant cell voltage in NaCl-CaCl2 melts at 800 °C for 600 min(a) 0.3 V (b) 1.0 V (c) 1.5 V (d) 3.0 V (inset: EDS test data).

The electrolysis of Fe3O4 samples with constant cell voltage under a two-electrode further illustrates the results obtained by electrochemical tests. With the gradual increase of voltage, the reduction potential of Fe3O4 → FeO is first reached. The oxygen at the vertex of the regular octahedron in Fe3O4 is ionized and diffused to the interface of Fe3O4| NaCl-CaCl2 (Chen and Fray 2004). When the cell voltage increases to the reduction potential 0.5 V of FeO → Fe, metallic Fe begins to be produced. It is concluded that as the voltage continues to increase, the reduction driving force increases, and the reduction speed will also accelerate, which is consistent with the analysis results in CVs.

3.4 The effect of time on Fe3O4 electrolysis products

The process of Fe3O4 electro-deoxidation reduction was explained, electrolysis experiments were carried out at 2.0 V cell voltage for different times, and the results are shown in Figure 8. Ca2Fe2O5 appeared when the applied voltage was 2.0 V for different times of electrolysis for 10–600 min. Due to the influence of Ca2+ that Fe3O4 first formed Ca2Fe2O5. As the applied voltage reaches the reduction potential of Fe, a part of FeO and Fe will appear. In the short 10 min electrolysis process, due to the short electrolysis time, there are only diffraction peaks with little Fe. As the electrolysis time is extended to 30 and 60 min, the amount of metallic Fe generated increases, and the peaks intensity of Ca2Fe2O5 and FeO decreased slightly; after 120 min of electrolysis, the diffraction peak of FeO disappeared, only Fe and little Ca2Fe2O5 remained. Ca2Fe2O5 gradually reduced and the diffraction peak of metallic Fe increased as time went.

Figure 8: XRD pattern of Fe3O4 sample electrolyzed at 2.0 V with constant cell voltage in NaCl-CaCl2 melts at 800 °C for 10–300 min (a) Scan range:10–90° (b) Scan range:30–50°.
Figure 8:

XRD pattern of Fe3O4 sample electrolyzed at 2.0 V with constant cell voltage in NaCl-CaCl2 melts at 800 °C for 10–300 min (a) Scan range:10–90° (b) Scan range:30–50°.

It is worth noting that the diffraction peak of FeO disappeared from 120 to 300 min, and only a small amount of Ca2Fe2O5 and most of Fe existed. This point is satisfied by the thermodynamic data, such as Eqs. (4) and (5). The thermodynamic data the ΔGΘ of Ca2Fe2O5 decomposing into FeO is greater than that of FeO decomposing into Fe at 800 °C. According to the theoretical decomposition voltage corresponding to Table 1, the theoretical decomposition voltage of FeO reduction to Fe is −1.01 V, which is just the theoretical decomposition voltage of Ca2Fe2O5 reduction to FeO is −1.13 V, so there is no FeO after 120 min.

(4)2FeO=2Fe+O2(g)
ΔGθ=388kJ·mol1 T=800°C
(5)2(2CaO)·Fe2O3=4FeO+4CaO+O2(g)
ΔGθ=439kJ·mol1 T=800°C

The formation of Ca2Fe2O5 is an extremely easy process. Its generation is not due to the chemical reaction between CaO and Fe3O4 produced by the hydrolysis of CaCl2, but when the voltage is energized, Ca2+ moves to the cathode and adsorbs on the surface of Fe3O4. All the reactions involved are shown in Eqs. (6–8):

(6)2Fe3O4+2Ca2++10e=Ca2Fe2O5+4Fe+3O2
(7)Ca2Fe2O5+2e=2Ca2++2FeO+3O2
(8)FeO+2e=Fe+O2

Figure 9 is the SEM image and EDS data of the Fe3O4 substrate at a constant cell voltage of 2 V in melts of NaCl: CaCl2 = 0.48: 0.52 at 800 °C for 10–300 min. The particles gradually become finer with time. The O contenting the electrolysis products keeps decreasing and the Fe content keeps increasing.

Figure 9: SEM images of Fe3O4 sample electrolyzed by constant cell voltage of 2 V in NaCl-CaCl2 melts at 800 °C (a) 10 min (b) 30 min (c) 60 min (d) 120 min (e) 180 min (f) 300 min.
Figure 9:

SEM images of Fe3O4 sample electrolyzed by constant cell voltage of 2 V in NaCl-CaCl2 melts at 800 °C (a) 10 min (b) 30 min (c) 60 min (d) 120 min (e) 180 min (f) 300 min.

Within the first 60 min of electrolysis time, interconnected Ca2Fe2O5 is mainly formed, and its structure becomes denser and the porosity decreases. According to the de-oxidation of TiO2, CaTiO3 is easily formed during the electrolysis process, which has a certain effect on the effective migration of ions (Dolganov et al. 2020). In the early stage of Fe3O4 de-oxidation, Ca2Fe2O5 is easily formed and has an effect on the migration of ions, the Ca content in this period of time is higher and the O content decreases less (Li et al. 2010). During the electrolysis to 120–180 min, the interconnected Ca2Fe2O5 is gradually deoxidized and decomposed into granular metallic Fe with a certain porosity, which accelerates the mass transfer and electron conduction of O2− from the inside of the solid cathode to the molten salt. When the electrolysis time is 300 min, due to the removal of oxygen, the reduced Fe atoms gradually nucleate and grow, forming metallic Fe with a particle size of 5–10 μm.

2Figure 10 is a line graph of the changes of Fe, O and Ca content with the electrolysis time. From the figure, the O content gradually decreases with the increase of the electrolysis time. During the first 60 min of electrolysis, Ca2Fe2O5 is mainly formed with low Fe content. At 60–120 min of electrolysis, Fe content increases rapidly and in the stage of rapid deoxygenation and reduction. After 120 min of electrolysis, the Fe content increases slowly, because the O2− ions generated through the rapid de-oxidation stage are too late to migrate and gather near the cathode, which hinders the progress of electrolytic de-oxidation.

Figure 10: EDS of Fe3O4 sample electrolyzed by constant cell voltage in NaCl-CaCl2 melts at 800°C for 300 min.
Figure 10:

EDS of Fe3O4 sample electrolyzed by constant cell voltage in NaCl-CaCl2 melts at 800°C for 300 min.

3.5 Current efficiency

With the increase of voltage, the current first reaches the maximum due to the electron accumulation. Fe3O4 can be reduced at 1 ∼ 3 V. When Fe3O4 begins to reduce, the outermost electrons reach saturation, and active ions appear at the three-phase interface, and O2− migrates to anode discharge. After electrolysis at 1 V for 38 min, the applied energy is the same as the energy required for the activation ion migration, and the current approaches equilibrium. The reaction rate is the highest in the first 7 min of 2 V electrolysis. With the increase of time, the reaction proceeds to the inside, the reaction resistance increases gradually, and the current decreases slowly. At 3 V, the nucleation, growth and decomposition of Ca2Fe2O5 hinder the migration of O2− and slow down the reaction rate. Finally, the current tends to be stable. Metal iron can be obtained by electrolysis for 8 h at a constant cell voltage of 1.0 V. The electrolytic efficiency reaches 96.4% and the electrolytic energy consumption is 3.4 kW·h·kg−1.

Figure 11: I-t plots of electrolysis under 3 V voltages for 8 h in 800 °C NaCl-CaCl2 melts.
Figure 11:

I-t plots of electrolysis under 3 V voltages for 8 h in 800 °C NaCl-CaCl2 melts.

Table 2:

Current efficiency and energy consumption calculation for electro-reduction of Fe3O4 at different electrolytic conditions.

Electrolytic conditionsMass of Fe3O4(g)Consumed electric (mA h)Current efficiency(%)Energy consumption (kW h/kg-Fe)Actual product

-Fe(g)
Theoretical

-Fe(g)
1 V–8 h0.81822.296.43.40.530.56
2 V–8 h0.83955.882.817.00.460.56
3 V–8 h0.84069.868.931.60.380.56

4 Conclusion

In the NaCl-CaCl2 melts with an equimolar ratio of 800 °C, the solid Fe3O4 was taken as the material, and the reduction mechanism of the solid Fe3O4 was studied by electrochemical methods using FCE-Fe3O4 and MCE-Fe3O4 as working electrodes, respectively. The products of voltage electrolysis are studied, and the estimates are as follows:

  1. The process of reducing solid Fe3O4 to Fe is a two-step electron-gaining process, which transfers 2 mol of electrons per step. The reduction process is: Fe3O4 → FeO → Fe.

  2. The effect of different voltages on the Fe3O4 product is as follows: when the cell voltage is less than 0.4 V is FeO, the electrolytic product of 0.5–0.6 V contains FeO and Fe, and when it is greater than 0.8 V, it is all reduced to Fe. When the cell voltage reaches the reduction potential of Fe, the product particles become finer as the voltage increases.

  3. Electrolysis at 2.0 V for different times, the initial formation was mainly dense Ca2Fe2O5. When the time is longer than 60 min, the process of reducing Ca2Fe2O5 to Fe, and the electrolysis time is 60–120 min, the de-oxidation speed is the fastest.


Corresponding author: Jinglong Liang, North China University of Science and Technology, 063009, Tangshan, Hebei, China, E-mail:

Funding source: National Natural Science Foundation of China

Award Identifier / Grant number: 51674120

Funding source: Hebei Province Higher Education Science and Technology Research Project

Award Identifier / Grant number: BJ2017050

Funding source: Graduate Student Innovation Fund of North China University of Science and Technology

Award Identifier / Grant number: 2019S01

Acknowledgments

The study is financially supported by the National Natural Science Foundation of China (No. 51674120), Hebei Province Higher Education Science and Technology Research Project (No. BJ2017050) and Graduate Student Innovation Fund of North China University of Science and Technology (No. 2019S01).

  1. Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.

  2. Research funding: None declared.

  3. Conflict of interest statement: The authors declare no conflicts of interest regarding this article.

References

Allanore, A., H. Lavelaine, G. Valentin, J. P. Birat, P. Delcroix, and F. Lapicque. 2010. “Observation and Modeling of the Reduction of Hematite Particles to Metal in Alkaline Solution by Electrolysis.” Electrochimica Acta 55 (12): 4007–13, https://doi.org/10.1016/j.electacta.2010.02.040.Search in Google Scholar

Allanore, A., H. Lavelaine, G. Valentin, J. P. Birat, and F. Lapicque. 2007. “Electrodeposition of Metal Iron from Dissolved Species in Alkaline Media.” Journal of the Electrochemical Society 154 (12): E187, https://doi.org/10.1149/1.2790285.Search in Google Scholar

Chen, G., and D. Fray. 2004. Understanding the Electro-Reduction of Metal Oxides in Molten Salts. Warrendale, PA, USA: TMS.Search in Google Scholar

Chen, G., D. Fray, and T. Farthing. 2000. “Direct Electrochemical Reduction of Titanium Dioxide to Titanium in Molten Calcium Chloride.” Nature 407 (6802): 361–4, https://doi.org/10.1038/35030069.Search in Google Scholar PubMed

Dolganov, A., M. T. Bishop, M. Tomatis, G. Z. Chen, and D. Hu. 2020. “Environmental Assessment of the Near-Net-Shape Electrochemical Metallisation Process and the Kroll – Electron Beam Melting Process for Titanium Manufacture.” Green Chemistry 22 (6): 1952–67.10.1039/C9GC04036FSearch in Google Scholar

Gao, H. P., X. Jin, S. W. Zou, F. Z. Ling, J. J. Peng, and G. Chen. 2013. “Liquid Diffusion of the Instantaneously Released Oxygen Ion in the Electrolytic Porous Fe from Solid Iron Oxide in Molten CaCl2.” Electrochimica Acta 107: 261–8.10.1016/j.electacta.2013.06.013Search in Google Scholar

Haarberg, G. M., and B. Yuan. 2014. “Direct Electrochemical Reduction of Hematite Pellets in Alkaline Solutions.” ECS Transactions 58 (20): 19–28, https://doi.org/10.1149/05820.0019ecst.Search in Google Scholar

Hu, H. B., Y. M. Gao, Y. G. Lao, Q. W. Qin, G. Q. Li, and G. Z. Chen. 2018. “Yttria-Stabilized Zirconia Aided Electrochemical Investigation on Ferric Ions in Mixed Molten Calcium and Sodium Chlorides.” Metallurgical and Materials Transactions B 49 (1): 2794–808, https://doi.org/10.1007/s11663-018-1371-z.Search in Google Scholar

Jiao, S.Q., H.D. Jiao, W.L. Song, M.Y. Wang, and J.G. Tu. 2020. “A Review on Liquid Metal as Cathode for Molten Salt/oxide Electrolysis.” Int. J. Miner. Metall. Mater 27 (5): 1–31.10.1007/s12613-020-1971-xSearch in Google Scholar

Li, C., W. Han, M. Li, W. Wang, and Yang XiaoGuang. 2018. “Electrochemistry of Zr (IV) in Molten LiCl-KCl-K2ZrF6 System.” International Journal of Electrochemical Science 13 (13): 11795–807, https://doi.org/10.20964/2018.12.04.Search in Google Scholar

Li, G. M., D. H. Wang, and Z. Chen. 2009. “Direct Reduction Of Solid Fe2O3 in Molten CaCl2 by Potentially Green Process.” Journal of Materials Sciences and Technology 25 (06): 767–71.Search in Google Scholar

Li, H., L. Jia, J. Wang, J. L. Liang, H. Y. Yan, Z. Y. Cai, and L. Wang. 2020. “Electrochemical Reduction Mechanism of Several Oxides of Refractory Metals in FClNaKmelts.” High Temperature Materials and Processes 39 (2020): 1–9, https://doi.org/10.1515/htmp-2020-0008.Search in Google Scholar

Li, H., L. S. Zhang, J. L. Liang, R. G. Reddy, H. Y. Yan, and Y. h. Yin. 2019. “Electrochemical Behavior of Fe3O4 in NaCl-CaCl2 Melts.” International Journal of Chemical Reactor Engineering 17 (10), https://doi.org/10.1515/ijcre-2019-0017.Search in Google Scholar

Li, W., X. Jin, F. Huang, and G. Chen. 2010. “Metal-to-Oxide Molar Volume Ratio: The Overlooked Barrier to Solid-State Electroreduction and a “Green” Bypass through Recyclable NH4HCO3.” Angewandte Chemie 49: 3203–6, https://doi.org/10.1002/anie.200906833.Search in Google Scholar

Liu, Z. W., H. L. Zhang, L. L. Pei, Y. L. Shi, Z. H. Cai, H. B. Xu, and Y. Zhang. 2018. “Direct Electrolytic Preparation of Chromium Metal in CaCl2–NaCl Eutectic Salt.” Transactions of Nonferrous Metals Society of China 28 (2): 376–84, https://doi.org/10.1016/s1003-6326(18)64671-0.Search in Google Scholar

Rong, L. B., R. He, Z. Y. Wang, J. J. Peng, X. B. Jin, and G. Z. Chen. 2014. “Investigation of Electrochemical Reduction of GeO2 to Ge in Molten CaCl2-NaCl.” Electrochimica Acta 147: 352–9, https://doi.org/10.1016/j.electacta.2014.09.107.Search in Google Scholar

Tang, D. Y., H. Y. Yin, X. H. Cheng, W. Xiao, and D. H. Wang. 2016. “Green Production of Nickel Powder by Electro-Reduction of NiO in Molten Na2CO3–K2CO3.” International Journal of Hydrogen Energy 41 (41): 18699–705, https://doi.org/10.1016/j.ijhydene.2016.06.078.Search in Google Scholar

Tang, D. Y., H. Y. Yin, W. Xiao, H. Zhu, X. H. Mao, and D. H. Wang. 2013. “Reduction Mechanism and Carbon Content Investigation for Electrolytic Production of Iron from Solid Fe2O3 in Molten K2CO3–Na2CO3 Using an Inert Anode.” Journal of Electroanalytical Chemistry 689: 109–16, https://doi.org/10.1016/j.jelechem.2012.11.027.Search in Google Scholar

Weng, W., G. Gong, Z. Wang, D. Wang, and Z. C. Guo. 2017. “Electrochemical Reduction Behavior of Soluble CaTiO3 in Na3AlF6-AlF3 Melt for the Preparation of Metal Titanium.” Journal of the Electrochemical Society 164 (9): D551–7, https://doi.org/10.1149/2.0611709jes.Search in Google Scholar

Weng, W., M. Wang, X. Gong, Z. Wang, D. Wang, and Z. Guo. 2016. “Direct Electro-Deposition of Metallic Chromium from K2CrO4 in the Equimolar CaCl2-KCl Molten Salt and its Reduction Mechanism.” Electrochimica Acta 212: 162–70, https://doi.org/10.1016/j.electacta.2016.06.142.Search in Google Scholar

Wu, Y.K., S Chen, and L.J. Wang. 2020. “Electrochemistry of Hf (IV) in NaCl-KCl-NaF-K2HfF6 Molten Salts.” Int. J. Miner. Metall. Mater 27 (5): 1–19.10.1007/s12613-020-2083-3Search in Google Scholar

Yin, H., D. Tang, H. Zhu, Y. Zhang, and D. Wang. 2011. “Production of Iron and Oxygen in Molten K2CO3–Na2CO3 by Electrochemically Splitting Fe2O3 Using a Cost Affordable Inert Anode.” Electrochemistry Communications 13 (12): 1521–4, https://doi.org/10.1016/j.elecom.2011.10.009.Search in Google Scholar

Zhang, D.Y., X Ma, H.W. Xie, X Chen, J.K. Qu, Q.S. Song, and H. Y. Yin. 2020. “Electrochemical Derusting in Molten Na2CO3-K2CO3.” Int. J. Miner. Metall. Mater 27 (5): 1–19, https://doi.org/10.1007/s12613-020-2189-7.Search in Google Scholar

Zou, X. L., S. L. Gu, X. L. Xie, C. Y. Lu, Z. F. Zhou, and W. Z. Ding. 2015. “Electroreduction of Iron(III) Oxide Pellets to Iron in Alkaline Media: A Typical Shrinking-Core Reaction Process.” Metallurgical and Materials Transactions B 46 (3): 1262–74, https://doi.org/10.1007/s11663-015-0336-8.Search in Google Scholar

Received: 2020-09-18
Accepted: 2020-12-12
Published Online: 2020-12-25

© 2020 Walter de Gruyter GmbH, Berlin/Boston

Downloaded on 26.4.2024 from https://www.degruyter.com/document/doi/10.1515/ijcre-2020-0179/html
Scroll to top button