Metallic iron whisker formation and growth during iron oxide reduction: basicity effect

Abstract

The effect of basicity on the metallic iron whisker growth during wüstite reduction was studied in the present investigation. Compacts of pure and CaO/SiO₂ doped wüstite were synthesised. The annealed compacts were isothermally reduced in thermogravimetric apparatus with CO gas at 800–1100°C. The course of reduction was followed by measuring the weight loss as a function of time. X-ray diffraction (XRD), scanning electron microscope (SEM), optical microscope and porosity measurements were used to characterise the annealed and reduced samples. The influence of temperature and basicity (CaO/SiO₂) on the reduction behaviour and the morphology of the annealed samples were investigated. The reduction rate increased with temperature but decreased by increasing basicity value. Metallic iron whisker shape structure was detected in the pure wüstite samples after reduction at high temperatures while in basic wüstite samples, whiskers were formed at the surface of the compacts. From the activation energy values, the reduction of pure wüstite is most likely controlled by a combined effect of gaseous diffusion and interfacial chemical reaction mechanisms. The reduction of basic wüstite compacts with 0·2 and 0·5 basicity ratios are most likely controlled by chemical reaction mechanism while for 0·8 basicity ratio, the reduction rate is most likely controlled by solid state reaction mechanism.     

Effect of Ca(II)‐ion in Wüstite on its Reduction to Iron

Fired compacts of hematite doped with different contents of CaO (mass fractions were 3.5, 4.0, 4.5, 5.0 %) were isothermally reduced to wüstite in a CO‐CO₂ gas mixture at 1173K. It was found from X‐ray diffraction investigation that the lattice parameters of wüstite increased with CaO content dissolved in wüstite crystal. The pure wüstite as well as the wüstite doped with CaO were reduced at different temperature. The results showed that the reductions were promoted with increasing CaO content. The higher the content of CaO dissolved in wüstite crystal was, the larger the lattice parameter and the interplanar distances of wüstite became. This expansion was helpful to the migration of Fe ions and enhanced the reduction. At the early stage, the reduction of pure and doped wüstite compacts was controlled by the interfacial chemical reaction step. At the latter stage, the gaseous diffusion was the rate‐determining step for both sample types.     

Blast Furnace Operating Conditions Manipulation for Reducing Coke Consumption and CO2 Emission

A comparative reduction behavior of wüstite samples prepared from iron ore sinter was investigated to find the optimum way for reducing coke consumption and CO₂ emission in blast furnace technology. A series of wüstite reduction experiments was carried out using different gas mixtures. A correlation between the experimental results and real data of blast furnaces at Egyptian Iron and Steel Company (EISCO) was demonstrated in order to optimize the coke consumption inside blast furnaces. Different theoretical models were applied on real data of blast furnaces to calculate the effect of operation parameters on the coke consumption. It was found that the wüstite reducibility can be controlled and enhanced using certain ratio of H₂ and CO gases. Such kind of enhancement decreases the remaining quantity of unreduced wüstite which descends to the high temperature region of blast furnace. The theoretical analysis of real data using certain values of H₂ and CO shows that increasing the amount of natural gas injection in blast furnace of EISCO will decrease the degree of direct reduction of iron oxide and consequently will decrease the amount of coke consumption.     

Effect of basicity on wüstite sinter reducibility under simulated blast furnace conditions

Abstract

In this study, synthetic sinters with different basicity (CaO/SiO₂?=?0, 0·5 and 2·0) were prepared at 1300°C and prereduced at 900°C using low potential reducing gas (LPRG; 20%CO, 20%CO₂, 5%H₂ and 55%N₂). The prereduced sinters were subsequently reduced to metallic iron at 950–1100°C using relatively high potential reducing gas (HPRG; 30%CO, 5%CO₂, 10%H₂ and 55%N₂). Both LPRG and HPRG were selected to simulate the gas composition in the blast furnace upper and lower shaft respectively. High pressure mercury porosimeter, X-ray phase analysis, optical and scanning electron microscope were used for the analysis of the prepared and reduced sinters. In the original basic sinter, calcium ferrite (CaFe₂O₄) and dicalcium silicate (Ca₂SiO₄) phases were identified as well as the main Fe₂O₃ phase, whereas wollastonite [Ca_2·87Fe_0·13(SiO₃)₃] and silica (SiO₂) were formed in the acidic sinter. The prereduction in sinters with LPRG at 900°C resulted in the formation of wüstite (Fe_0·902O) phase. The subsequent reduction in wüstite sinters to metallic iron using HPRG at 950–1100°C was found to be the highest for basic sinter and the least for acidic sinter. The higher reduction rate of basic sinter was attributed to the enhancement of wüstite reducibility through the formation of calcium ferrites. The lower reduction rate of wüstite in acidic sinter was attributed to the formation of hard reducible fayalite (Fe₂SiO₄) and ferrobustamite [(Ca_0·5Fe_0·5)SiO₃] phases. The rate controlling mechanism during the reduction process was estimated by the correlation between apparent activation energy calculation and microstructure investigations.     

Multistep Reduction Kinetics of Fine Iron Ore with Carbon Monoxide in a Micro Fluidized Bed Reaction Analyzer

The reduction kinetics of Brazilian hematite by CO is investigated in a Micro Fluidized Bed Reaction Analyzer (MFBRA) using an analyzing method based on Johnson-Mehl-Avrami (JMA) model at temperatures of 973 K (700 °C), 1023 K (750 °C), 1073 K (800 °C), and 1123 K (850 °C). The solid products at different reduction stages are evaluated by SEM/EDS and XRD technologies. Results indicate that the reduction process is better to be discussed in terms of a parallel reaction model that consists of the reactions of hematite to wüstite and wüstite to iron, rather than a stepwise route. Meanwhile, the controlling mechanism of the reduction process is found to vary with temperature and the degree of conversion. The overall process is controlled by the gas–solid reaction occurring at the iron/wüstite interface in the initial stages, and then is limited by the nucleation of wüstite, and finally shifts to diffusion control. Moreover, the reactions of hematite to wüstite and wüstite to iron take place simultaneously but with different time dependences, and the apparent activation energies of hematite to wüstite and wüstite to iron are determined as 83.61 and 80.40 KJ/mol, respectively.     

Formation and decomposition of C4F7 type calcium ferrites in superfluxed magnetite based pellets during oxidation and reduction

Abstract

In the current work the reactions of magnetite based pellets with large additions of calcite (3%CaO) during reduction have been investigated. This made it possible to use both X-ray diffraction (XRD) and scanning electron microscopy (SEM) to detect reaction phases that normally occur in very small amounts. The main binding phase in the pellets after oxidation was (CaO,MgO,FeO)₄(Fe₂O₃)₇, whereas the one commonly reported in the literature is (CaO)(Fe₂O₃)₂. During reduction at 500–700°C severe cracking occurred in these pellets, especially in the calcium ferrite phase. However, the decomposition of this phase began at 600°C, and therefore it is believed that the reason for the cracks is low strength of the phase itself, rather than weakness induced by reduction of the phase. Upon reduction of magnetite into wüstite at 800°C, the calcium began dissolving in the wüstite, and at 900°C porous calciowüstite had formed in the entire sample, except for some remaining magnetite left in the pellet cores.     

Effect of Compressive stresses on Microstructure of Scales Formed on Pure Iron Under Continuous Air Cooling

 The microstructure development of oxide scale on pure iron under the mutual effect of compressive stress and cooling conditions was investigated. Oxide-scale structure was examined by optical microscopy (OM) and scanning electron microscopy (SEM). It was found that oxide scale formed under normal cooling conditions had a structure mainly consisted of an outer magnetite and an inner wüstite layer. When a compressive stress was applied, numerous magnetite precipitates formed within wüstite layer homogeneously at starting cooling temperature of 900℃, and the wüstite layer in the scale transformed into a mixture of mostly magnetite/iron eutectoid and magnetite layer at starting cooling temperature of 700℃. The wüstite decomposition and precipitation of magnetite in wüstite under compressive stress was discussed.     

Formation of a Network Structure in the Gaseous Reduction of Magnetite Doped with Alumina

Reduction of un-doped magnetite is developed topochemically with the formation of a dense iron shell. However, the reduction of alumina-doped magnetite to wüstite proceeds with the formation of a network-like structure which consists of criss-crossed horizontal and vertical plates of wüstite. Reduction of magnetite includes the conversion of Fe³⁺ to Fe²⁺ and the movement of iron cations from the tetrahedral sites on the {400} and {220} planes of magnetite to the octahedral sites on the {200} planes of wüstite. Alumina has a negligibly small solubility in wüstite. In the reduction of magnetite doped with Al₂O₃, rejected Al³⁺ cations from wüstite diffuse to the magnetite–hercynite solid solution. Enrichment of the Fe₃O₄–FeAl₂O₄ solution with alumina in the vicinity of the reduction interface restricts the growth of {220} planes of wüstite and nucleation of {220} planes adjusted to the existing planes, preventing the merging of wüstite plates during the reduction process. Reduction of magnetite from the magnetite–hercynite solid solution practically stops when the Al³⁺ content at the interface approaches the solubility limit. Wüstite in the separated plates is reduced further to iron.     

The diffusivity of sulfur in wüstite and its solubility in wüstite and γ iron in equilibrium with each other and a liquid oxysulfide

The solubility of sulfur in wüstite in equilibrium with γ iron and liquid oxysulfide was found to be 0.011 wt pct at 1050°C. The sulfur solubility in γ iron in equilibrium with wüstite and liquid oxysulfide was also determined at 1050°, 1150°, and 1250°C and found to be 135, 165, and 160 ppm respectively. These values are considerably lower than the sulfur-solubility in y iron in the binary Fe-S system saturated with pyrrhotite. The diffusivity of sulfur in wüstite was determined by oxidizing Fe-S alloys in mixtures of CO and CO₂, and analyzing the entire sample for sulfur afterwards. From the amount of sulfur diffused through the growing wüstite layer into the gas phase, the diffusivity of sulfur in wüstite was evaluated, and found to be 4.1 × 10⁻⁸ and 9.6 × 10⁻⁷ cm²/s at 1050° and 1250°C respectively. These values are of the same order as the self-diffusivity of iron in wüstite in equilibrium with iron at the same temperatures.     

Influence of Manganese Oxide and Silica on the Morphological Structure of Hematite Compacts

Dry compacts of pure Fe₂O₃ and Fe₂O₃ doped with either (2–6 mass%) MnO₂, (2.5–7.5 mass%) SiO₂ or with both (2–6% MnO₂ + 7.5% SiO₂) were indurated at 1373 K for 6 hours and physically and chemically characterized. The fired compacts were isothermally reduced with pure CO gas at 1073–1373 K. The O₂‐weight loss was continuously recorded as a function of time using TGA technique. The external volume of pure and doped compacts was measured at different reduction conditions and the volume change was calculated. The structural changes accompanying the reduction process were visually and microscopically examined and the different phases were identified by X‐ray diffraction analysis. After firing, manganese ferrite (MnFe₂O₄) phase was identified in MnO₂‐doped compacts. In pure Fe₂O₃ compacts, the external volume of compacts was increased with reduction temperature, showing a maximum swelling value at 1198 K. Catastrophic swelling was observed in MnO₂‐doped Fe₂O₃ compacts, the volume change increased with MnO₂ content showing catastrophic swelling in compacts containing 6%MnO₂ at 1248 K. The catastrophic swelling was attributed to the formation of dense metallic iron whiskers and plates in a highly porous structure. Unlike in MnO₂‐doped samples, no considerable volume changes were detected in SiO₂‐doped Fe₂O₃ and (MnO₂ + SiO₂)‐doped Fe₂O₃ compacts where the presence of silica greatly hindered the swelling phenomenon at all reduction temperatures.     

Isothermal reduction behaviour of MnO2 doped Fe2O3 compacts with H2 at 1073–1373 K

Abstract

Pure Fe₂O₃ and Fe₂O₃ doped with 2, 4, and 6 mass% of MnO₂ (>99%) compacts annealed at 1473 K for 6 h were isothermally reduced with H₂ at 1073–1373 K. The O₂ weight loss resulted from the reduction of compacts was continuously recorded as a function of time using thermogravimetric analysis (TGA). High pressure mercury porosimeter, optical and scanning electron microscopes, X-ray phase analysis and vibrating sample magnetometer were used to characterise both the annealed and reduced samples. In MnO₂ containing samples, manganese ferrite (MnFe₂O₄) was identified. The rate of reduction of pure and doped compacts increased with temperature and decreased with the increase in MnO₂ content. Unlike in pure compacts, the reduction of MnO₂ containing samples was not completed and stopped at different extents depending on MnO₂ (mass%). At initial reduction stages, the decrease in the rate was due to the presence of poorly reducible manganese ferrite (MnFe₂O₄) phase which was partially reduced to iron manganese oxide (FeO_0.899, MnO_0.101) at the final stages. The reduction mechanism was predicted from the correlation between the reduction kinetics and the structure of partially reduced samples at different temperatures. The reduction of pure and doped samples was controlled by a combined effect of interfacial chemical reaction and gaseous diffusion mechanism at their initial stages. At final stages, the interfacial chemical reaction was the rate controlling mechanism.     

Untersuchungen zum Hydratationsverhalten von synthetischen Magnesiowüstiten

Eine ausreichende Raumbeständigkeit ist die notwendige Voraussetzung für die Verwendung von Stahlwerksschlacken als Baustoff mit höheren qualitativen Anforderungen im Straßenbau. Die Raumbeständigkeit kann durch Mineralphasen gefährdet werden, die mit Feuchtigkeit unter Volumenzunahme hydratisieren. Zu den hydratisierbaren Mineralphasen zählen Freikalk und in MgO-reichen Schlacken auch bestimmte MgO-Phasen. Mikroskopische Untersuchungen an MgO-reichen LD-Schlacken zeigen, daß das MgO größtenteils in Form von Magnesiowüstiten – Mischkristallen zwischen MgO und FeO/MnO – gebunden vorliegt. Während reines MgO hydratisierbar ist, verhalten sich FeO und MnO inert gegenüber Feuchtigkeit. Anhand von synthetisch hergestellten Magnesiowüstiten unterschiedlicher Zusammensetzung wurde geprüft, welche Mischkri-stalle als hydratisierbar einzustufen sind. Dazu wurden Hydratationsversuche bei unterschiedlichen Temperatur-/Druck-Bedin-gungen durchgeführt. Bei 84°C/0,53 bar war zur Stabilisierung der Magnesiowüstite ein (FeO + MnO)-Gehalt von 30% notwendig, bei 215°C/21 bar lag dieser Gehalt bei 60%. Da sich die Bedingungen bei der Prüfung im Labor von denen in der Straße deutlich unterscheiden, ist die Festlegung eines (FeO + MnO)-Grenzgehalts, durch den Magnesiowüstite unter den in einer Straße herrschenden Bedingungen stabilisiert werden, zur Zeit noch nicht möglich. Solange die Übertragbarkeit der Laborergebnisse auf die Praxis noch nicht nachgewiesen ist, wird der unter den sehr scharfen Autoklavbedingungen ermittelte MgO-Grenzgehalt von 40% zugrunde gelegt. Als hydratisierbares MgO – oder auch MgO_frei in Analogie zu CaO_frei – werden somit ungebundenes MgO und Magnesiowüstite mit > 40% MgO bezeichnet.     

Reduction swelling behaviour of haematite/magnetite agglomerates with addition of MgO and CaO

Abstract

The influence of three kinds of CaO and MgO additives (dolomite, burnt lime and serpentine) on the reduction swelling behaviour of haematite–magnetite (H–M) concentrates pellets was studied. Burnt lime and dolomite increased the reduction swelling index of H–M oxidised pellets, while the reduction swelling index was able to be reduced when serpentine was added. CaO accelerated the formation and growth of metallic iron whiskers and led to abnormal swelling of the magnetite briquettes, while MgO was able to be dissolved in wüstite and reduced the migration rate of Fe²⁺; therefore, there was no catastrophic swelling in either the haematite or magnetite briquettes. As far as H–M concentrate pellets were concerned, because the solubility of CaO in magnetite was greater than that in the primary haematite and the secondary haematite generated from magnetite during the oxidation was easy to be reduced to wüstite, there was abnormal swelling in the reduced H–M pellets with CaO addition.     

Kinetics of dense magnetite formation during oxidation of wüstite and reduction of hematite in CO/CO2 gas mixtures

Thermogravimetric study of the wüstite oxidation. Influence of the cation diffusion in magnetite and of the oxygen transfer at the phase boundary magnetite/gas on the wüstite oxidation and on the hematite reduction. Linear and parabolic growth of dense magnetite.     

碱金属对铁氧化物气-固还原行为的影响

 利用热重分析的方法研究了Fe2O3在竖式电阻炉内被CO/CO2混合气体还原,Na2O对其还原行为的影响,共设计了4个成分,分别为纯Fe2O3、Fe2O3-1％Na2O、Fe2O3-3％Na2O和Fe2O3-5％Na2O。在还原过程中由于还原失氧引起的试验小饼失重量被实时记录下来,根据这些数据可以计算出还原反应速率以及化学反应的速率常数,以此来判断Na2O对Fe2O3还原性的影响。同时利用扫描电镜观察了还原后试样的微观形貌变化。研究发现：Na2O的存在将阻碍Fe2O3的还原。通过对还原后小饼的显微形貌观察,结合FeO-Na2O二元相图,发现这种负面影响主要是由于在还原过程中产生液相引起的。另外,还检测了还原前后试验小饼的孔隙率和体积变化。检测结果确认：还原后含Na2O的试验小饼的孔隙率和体积都有所下降。分析认为这种现象同样归因于还原过程中液相的产生。     

Effect of FeO and Al2O3 on the MgO Solubility in CaO–SiO2–FeO–Al2O3–MgO Slag System at 1823 K

The MgO solubility in CaO–SiO₂–FeO–Al₂O₃–MgO quinary slag system at 1823 K was measured to evaluate the effect of FeO and Al₂O₃ on the MgO solubility. It was found that the MgO solubility was decreased with higher optical basicity, FeO concentration, and increased with higher Al₂O₃ concentration. The MgO solubility was affected by activity coefficient of Mg²⁺ ($\gamma _{{\rm Mg}_{{\rm 2 + }} } $ ). Increase of the activity coefficient of Mg²⁺ ($\gamma _{{\rm Mg}_{{\rm 2 + }} } $ ) with higher FeO or lower Al₂O₃ decreased the MgO solubility. The increment in MgO solubility is remarkably reduced beyond a critical $X_{{\rm Al}_{2} {\rm O}_{{\rm 3}} } /(X_{{\rm Al}_{2} {\rm O}_{{\rm 3}} } + X_{{\rm FeO}} )$ ratio. The significant decrease of the increment of MgO solubility is caused by the change of the molten slag structure. The excess stability function of Al₂O₃ and the Fourier transform infrared (FT‐IR) analysis were applied to indirectly verify the existence of the spinel structure in the CaO–SiO₂–FeO–Al₂O₃–MgO slag system.     

Effect of Na2O on the Reduction of Fe2O3 Compacts with CO/CO2

Compacts of Fe₂O₃ and Fe₂O₃ doped with varying amounts of Na₂O were isothermally reduced at several temperatures, using CO/CO₂ mixed gas in a vertical resistance furnace. To determine the effect of Na₂O on the reduction of Fe₂O₃ compacts, the mass loss due to oxygen removal was continuously recorded, from which the reduction rate and rate constant were obtained. Na₂O was found to retard the reduction of Fe₂O₃ compacts. The apparent activation energy (E _a) of reaction and the mathematical relationship for pore gas diffusion suggested that the reduction behavior at the initial stages was controlled by a combination of pore gas diffusion and interfacial chemical reaction. At the intermediate and late stages of reduction, pore gas diffusion was the sole contributing factor. Morphological examination of the reduced compacts showed the formation of a liquid phase during the reduction process, which appeared to lower the rate of reaction.     

Effects of FeO and CaO/Al<Subscript>2</Subscript>O<Subscript>3</Subscript> Ratio in Slag on the Cleanliness of Al-Killed Steel

The slag composition plays a critical role in the formation of inclusions and the cleanliness of steel. In this study, the effects of FeO content and the C/A (CaO/Al₂O₃) ratio in the slag on the formation of inclusions were investigated based on a 10-minute slag–steel reaction in a MgO crucible. The FeO content in the top slag was shown to have a significant effect on the formation of MgO·Al₂O₃ spinel inclusions, and critical content exists; when the initial FeO content in the slag was less than 2 pct, MgO·Al₂O₃ spinel inclusions formed, and the T.O (total oxygen) was 20 ppm; when the initial FeO content in the slag was more than 4 pct, only Al₂O₃ inclusions were observed and the T.O was 50 ppm. It was clarified that the main source of Mg for the MgO·Al₂O₃ spinel inclusion formation was the top slag rather than the MgO crucible. In addition, the cleanliness of the steel increased as the initial FeO content in the top slag decreased. As regards the effects of the C/A ratio, the MgO amount in the observed inclusions gradually increased, whereas the T.O content decreased gradually with the increasing C/A ratio. Slag with a composition close to the CaO-saturated region had the best effect on the inclusion absorption.     

The mechanism of whisker growth in the reduction of wüstite

The growth of iron whiskers on wüstite is explained on the basis of C. Wagner’s mechanism for the reduction of nonstoichiometric oxides. The first metal nucleus develops into a whisker if the iron accumulation in the supersaturated wustite particle is nearly uniform, a condition which is favored by chemical control. A mathematical model is proposed in which the “filling ratio” of wüstite by iron prior to nucleation is expressed in terms of the reaction constant, the chemical diffusion coefficient of iron, the particle radius, the Fe/o ratio at equilibrium with the gas, and the critical Fe/o ratio for nucleation at the most favorable site. The morphology of iron in the early stages of reduction, which goes from a single cylindrical whisker, through sponge iron to a smooth layer, can be predicted by comparing at all times the map of the actual Fe/o ratio at the surface with the map of the local Fe/O ratio for nucleation. A satisfactory test of the theory is obtained through a survey of the experimental evidence available about the effect of several physico-chemical factors of reduction on the iron morphology,e.g. surface characteristics, particle size, composition of the reducing gas, solute cations in wüstite, temperature, and gas transport. Formerly with the Ecole Centrale des Arts et Manufactures, Chătenay-Malabry