Tellurium is an important alloying element in free-machining steel, but its thermodynamic properties in molten iron are not well understood. In the present study, the standard Gibbs energy (ΔG°) for the dissolution of tellurium gas into molten iron was determined in the temperature (T) range from 1823 to 1873 K by a vapor-liquid equilibration technique, in which the transpiration method was used to control the vapor pressure of tellurium. This relationship is expressed by the following equations:

We quantitatively confirmed that tellurium is thermodynamically less stable than oxygen or sulfur in molten iron. We also found a correlation between the standard Gibbs energy for the dissolution of chalcogens into molten iron and the standard Gibbs energy for the formation of the respective iron-chalcogenides.

The olivine crystals and borate containing slag were effectively separated from the CaO–SiO₂–B₂O₃–MgO–Al₂O₃ system at an optimum precipitating temperature of olivine by utilizing super-gravity. The borate containing slag melt went through the filter, whereas the olivine crystals with a larger size of 300 μm–2500 μm were intercepted by the filter. Consequently, after super-gravity separation with G=900 at 1463 K for 5 minutes, the mass fraction of B₂O₃ in separated borate containing slag was 30.17 wt%, while that of the separated olivine was only 0.009 wt%, and the recovery ratio of boron in separated borate containing slag thus was up to 99.97%.

The gas-injection BF is a new iron-making technology with injecting gas instead of the traditional pulverized coal injection and recycling the BF top gas through the gasifier. Compare with traditional BF, there is a larger amount of reducing gas such as CO and H₂ in gas-injection BF. Effect of H₂ addition on reduction kinetics of iron oxides in gas-injection BF was investigated by Thermogravimetric Analysis. The result shows that the reduction rate of iron oxides rise with the increase in both temperature and H₂ content. By contrast with the H₂ content increasing from 10% to 15% at 700°C and 900°C, the improvement of reduction rate is more obvious when addition from 15% to 20% at the same temperature. For comparison, it shows an opposite result in the condition of 1000°C. In addition, the efficiency of H₂ on reduction rate can be neglected as the content is less than 5%. The reaction mechanism was obtained according to the unreacted core model. In the condition of 30% CO+10% H₂, both the diffusion rate and interfacial reaction rate reveal an increasing trend with the increase of temperature, of which the degree of improvement in diffusion rate is more significant, which leads to the interface reaction gradually being the controlling step. The same case occurs with the increase of H₂ at 900°C. Under the condition of experiment, the activation energy decreases with H₂ addition, which illustrates that the reduction of iron oxides become more easily to perform with the rich hydrogen in gas-injection BF.

This paper contrastively investigated the reduction behavior of raw and oxidized vanadium titanium magnetite ore (RVTM and OVTM) in a bubbling fluidized bed at 800°C under 75%N₂-25%H₂, and systematically revealed the relationship between the structure, phases change, MgO migration behavior and the reduction performance. Microstructure observations suggest that oxidization conditions greatly affect structural evolution, phase transformation of Fe–Ti oxides, and migration of MgO in OVTM. The reduction results indicated that the average metallization degree enhanced by 8–10% as the OVTM ore had high oxidization degree. When the RVTM ore was treated at 800°C, the reduction rate increased with the increase of oxidization degree, no significant difference was observed among these oxidized samples due to their similar internal and surface structure. However, the reduction rate of OVTM at 950°C depended on the oxidization time. The reduction rate of OVTM oxidized at 950°C for 2 h was faster than the one for 1 h, which was attributed to the better exsolution of MgO and the formation of bigger pores, even though both samples were fully oxidized and Fe₂TiO₅ formed in the oxidization process.

For bridging the huge gap of combustibility between coke breeze and charcoal in iron ore sintering, the influences of using SiO₂ and B₂O₃ as modifiers on combustion characteristics of charcoal were investigated. The results showed that due to the physical barrier function of SiO₂ powders and the surface covering function of B₂O₃ film, the porosity of modified charcoal was reduced. The combustion speed of modified charcoal was therefore slowed down and showed less differences to coke breeze than non-modified charcoal. Recommended mass percentages of SiO₂ and B₂O₃ in modified charcoal were 2% and 3% respectively. The combustion reaction activation energy of charcoal was increased from 55.54 kJ∙mol⁻¹ to 86.97 kJ∙mol⁻¹ and 92.82 kJ∙mol⁻¹ respectively after modification, which was close to 122.75 kJ∙mol⁻¹ of coke breeze. Laboratory-scale sintering tests showed that the proper replacement percentage of charcoal to coke breeze was increased from 20% to 40% after the surface modification. And the emission reduction percentages of NOx and SOx were correspondingly improved from 16.58% and 11.02% to about 28% and 37% respectively, while the average combustion efficiency was kept at the comparable level to the case using coke breeze only.

Uneven charging of burden in circumferential direction has a negative impact to stable operation of ironmaking blast furnaces. To visualize and prevent uneven charging, particle flow in bell-less top with parallel hoppers was simulated with different tilting angles of rotating chute by Discrete Element Method (DEM). The simulation revealed that the particles flowed decentering in the vertical chute, which led to circumferential imbalance in the trajectory from the rotating chute. The tendency of imbalance against the circumferential direction also differed by the chute titling angle. This tendency was also confirmed in the scaled down experimental work. As a counter measure, DEM simulations showed that after tapered vertical chute installed was improved the circumferential imbalance significantly.

Hydrogen sulfide (H₂S) removal and catalytic ammonia (NH₃) decomposition performance of limonite in the presence of coke oven gas (COG) components has been studied in a cylindrical quartz reactor at 300–850°C under a high space velocity of 51000 h⁻¹ to develop a novel hot gas cleanup method. The H₂S removal behavior in 50% H₂/He depends on the temperature, with high performance observed at lower temperature. An investigation of the removal behavior of H₂S in the presence of COG components (CH₄, CO, CO₂ and H₂O) at 400°C reveals that CH₄ does not affect the removal performance. On the other hand, the coexistence of CO drastically decreases the H₂S removal performance. However, the addition of 5% H₂O to 50% H₂/30% CH₄/5% CO/He dramatically improves the H₂S removal performance, whereas the performance is low at 5% CO₂ with 50% H₂/30% CH₄/5% CO/He. In addition, the H₂S breakthrough curve strongly depends on the space velocity.

The limonite catalyst achieves almost complete decomposition of NH₃ in He at 850°C until 240 min. When the decomposition run is performed in the presence of COG components, the coexistence of 30% CH₄ deactivates limonite with significant formation of deposited carbon. On the other hand, the addition of 5% CO₂, 5% H₂O or 5% CO₂/5% H₂O to 50% H₂/30% CH₄/5% CO improves the catalytic activity without carbon deposition, and >99% conversion of NH₃ to N₂ is maintained until 240 min.

The reductions of titania-ferrous solution ore (TFSO) ranging from 800°C to 1100°C temperature and 10 to 100 vol% H₂–Ar gas mixture were conducted to account for the optimization of reduction parameters, the phase transition behaviors and solid solubility changes with Fe and Ti elements segregation and enrichment, and the micro morphology features in reduction process. 40 vol% of hydrogen in reducing gas, as well as 900°C reaction temperature was the suitable parameters when simultaneously considering the growth of reduction rate and the reasonable economy cost. The reduction of Fe₃O₄ contained in titanomagnetite (TTM) was proceeding earlier than that of ulvöspinel (Fe₂TiO₄), even though Fe₂TiO₄ was also the ingredient of TTM. The initial ilmenite was reduced quickly at primary stage, after that, there were also newly born ilmenite which experienced accumulation and vanishment, along with transformation of Fe₂TiO₄ and ferropseudobrookite (FeTi₂O₅). Although TiO₂ was appeared as early as 5 min, the low valence titanium (Ti₃O₅ and Ti₂O₃) were just observed at the end of reduction, and it was mainly because of the existence of iron oxide, resulting in the weak reduction potential to the titanium oxide. For the non-homogeneous TFSO particles, the lamellas showed a harder reducibility than the substrates and a different priority reduction positions in terms of the central and the edge. For the homogeneous TFSO particles, the reduction was carried out in an outside-in way, which was preferential along some interphase strips that had a lower gangue elements distribution.

The reduction and melting separation experiment of iron ore containing Al₂O₃ was performed. The results showed that Al₂O₃ obviously affected the melting separation behaviors and the carbon content of the iron nugget. When the Al₂O₃ content in iron ore was 4.59 wt%, the carbon content in the iron nugget was as low as 0.80 wt% and was much lower than the iron ores with lower Al₂O₃ content, and increased with the increasing of slag basicity. The melting separation and carburiz ation mechanism of high Al₂O₃ iron ore/coal composite pellet was deduced.

An numerical model for the bubble behavior of disintegration and dispersion is developed by using the model of Euler two-phase flows coupled with the bubble population balance model (BPBM). The gas volume fraction, and bubble size distribution are investigated under different stirring modes and gas flow rates. The results show that bubble behavior of disintegration and dispersion is the best under the clockwise-anticlockwise stirring mode (C-ARM), where effects of impeller rotating speed on the bubble behavior of disintegration and dispersion are different under different stirring modes. The excessive increasing in gas flow rate is not helpful for bubble behavior of disintegration and dispersion, while the gas holdup increases gradually and the mixing time increases rapidly with increasing the gas flow rate. The impeller under C-ARM with 2 rad/s for 2 s switching and the gas flow rate not exceeding 2.0 Nm³h⁻¹ are recommended to use for the present system, which can optimize the bubble behavior of disintegration and dispersion in the most space of ladle except for the bottom one. Moreover, numerical results show a good agreement against experimental ones.

A study has been carried out to measure the oxygen contents in molten Ni in equilibrium with NiO in the top slag, contained in a magnesia crucible, in the temperature range higher than 1500°C. The slag was composed of CaO–SiO₂–MgO satd.-NiO system with (mass% CaO)/(mass% SiO₂) = 1 and 8 mass% NiO. Firstly, it was found that oxygen content rapidly increased after adding the slag of CaO–SiO₂–MgO–NiO system into molten Ni and that it took 30 minutes to attain equilibrium. The measured oxygen contents were lower than those in equilibrium with pure solid NiO and increased with an increase in temperature. Thermodynamic analysis was made on the basis of the above results. It was found that activity of NiO in the present slag of CaO–SiO₂–MgO satd.-NiO system was positively deviated from ideal solution. In addition, Henry constant of NiO in this oxide system was obtained as the following relation under the assumption that Henry’s law is available:

under the slag of CaO–SiO₂–MgO satd.-NiO with (mass% CaO)/(mass% SiO₂) = 1.

O content could be estimated as follows by applying obtained in this study:

under the slag of CaO–SiO₂–MgO satd.-NiO with (mass% CaO)/(mass% SiO₂) = 1.

This paper presents the experimental study of an electromagnetically stirred mould flow using a 1:3 scale acrylic glass model of the round bloom caster from voestalpine Stahl Donawitz GmbH. An electromagnetic stirrer was installed at the strand producing a rotating magnetic field (RMF). Flow measurements were performed in the eutectic alloy GaInSn at room temperature by means of the ultrasound Doppler velocimetry (UDV). Up to 10 ultrasonic transducers were employed simultaneously in order to obtain a two-dimensional reconstruction of the flow structure. The experiments contribute to a better understanding of electromagnetically stirred mould flows and provide an extensive and valuable data base for the validation of numerical methods. The flow measurements reveal a distinct influence of the secondary flow on the distribution of the angular velocity in various regions of the mould. The submerged jet intensifies this secondary motion in the upper part of the mould and thus causes a strong deformation of the free surface of the melt. The jet is deflected, bent and rotates around the strand axis.

Control of slag Cr₂O₃ content is essential in stainless steelmaking electric arc furnace (EAF). Excessive Cr₂O₃ content of slag lead to high Cr₂O₃ content after the EAF tapping, since the solid precipitates forming in high Cr₂O₃ contents are not reduced significantly during the tapping procedure. In this work the slag Cr₂O₃ content during the EAF process was analysed by measuring the optical emission spectrum of the electric arc. The measurements were conducted in a pilot EAF situated in Aachen, Germany. Cr₂O₃ content of the slag was increased with periodical additions of Cr₂O₃ powder. The line ratios calculated from the optical emission spectra were compared to the results of the X-ray fluorescence (XRF) analysis of the slag samples taken from the furnace. The results indicate that best accuracy in a pilot scale can be obtained by using Ca I, Fe I or Mn I lines as reference for Cr I lines. By combining the most accurate line ratios of these three components, the Cr₂O₃ composition of the slag could be measured with an average absolute error of 0.62%-points and a standard deviation of 0.49%-points. The results suggest that Fe I and Mn I lines are the most promising reference lines for analysing Cr₂O₃ content of industrial EAF slag.

With the use of low-grade iron ores with high P content, slag with high P₂O₅ content generated after dephosphorization is considered a great potential source of P. Because of the solubility difference between the solid solution and matrix phase, it is possible to extract the P-rich solid solution selectively from slag by leaching. The soluble P obtained is suitable to produce phosphate fertilizers. To achieve selective leaching of P and increase its dissolution ratio, the effects of the cooling rate and acid on dissolution of the slag in aqueous solutions at pH 5 and 7 were investigated. This study found that during solidification, slow cooling facilitates coarsening of the solid solution and formation of the magnesioferrite phase. The solid solution was dissolved preferentially. At pH 7, the air-cooled slag showed the highest dissolution ratio of P. When the pH was decreased to 5, slag dissolution was significantly promoted. As the cooling rate decreased, the dissolution ratio of P increased. Slow cooling not only enhanced dissolution of the solid solution but also suppressed dissolution of the matrix phase. Citric acid performed better in promoting dissolution of slag. At pH 5, almost all of the solid solution was dissolved from the furnace-cooled slag. However, the dissolution ratio of Fe was also high. When nitric acid was used, 66.8% of the solid solution was dissolved, without dissolution of large amounts of the matrix phase. After leaching, the P₂O₅ content in the residue reduced and the Fe₂O₃ content increased.

The isothermal hot compression tests of 43CrNi steel was carried out at temperatures 800°C to 1050°C at an interval of 50°C and at constant strain rates of 0.01, 0.1, 1.0 and 10 s⁻¹ for total true strain of 0.7. The values of experimental stress were corrected for adiabatic heating which was used for flow stress modeling, and further its predictability was verified with experimental results. The true stress-true strain curves exhibit peak stresses followed by softening due to occurrence of DRX. The activation energy was calculated based on sinh type equation and found to be increasing with increase in strain. The obtained stress exponent 5.4, suggests that the mechanism of hot deformation is dislocation glide controlled by dislocation climb. Strain rate sensitivity maps and processing maps were developed using dynamic materials model (DMM), modified DMM with different instability criteria. The instability parameter, κ_j, is directly related to the dissipative function which is associated to microstructural changes, J, whereas the instability parameters, ξ and κ, are associated with the strain rate sensitivity parameter, m. Therefore, the instability parameter, κ_j, is more mathematically precise as it eliminates m from its calculations. The microstructures of deformed specimens were studied and correlated with the domains of processing maps to validate the workability region.

Development of a strain in oxide scale during phase transformation of FeO was investigated by a flexure method. A pure iron foil was oxidized on only one side, and the formed FeO was transformed at 400 or 500°C. In-situ observation of the bending behavior of the foil-formed specimen during the phase transformation of FeO was performed. Compressive strain develops in the scale in the early stage of the phase transformation, following which a significantly large expansion strain develops. Such a strain development corresponds to the two-step phase transformation of FeO and can be explained by the volume change of the scale considering the non-stoichiometry of FeO.

This study presents a computational framework for an efficient distortion analysis based on coupled process-mechanics modeling of metal active gas (MAG) welding. For coupled process-mechanics analysis, an integrated simulation model between bead formation, thermal conduction, and thermo-mechanical behaviors during MAG welding was developed. A combined bead formation and thermal conduction model provided bead surface and temperature profiles during welding from welding process conditions. Then, heat source parameters estimated based on welding process conditions enabled a predictive simulation of MAG welding process. The bead surface and temperature profiles obtained were used for a large deformation thermal elastic–plastic analysis of weld angular distortion. The calculations were compared with the measurements, which were obtained in a previous study, under the same welding process conditions to validate the developed analysis model. From the comparison results, we concluded that the developed model has great potential to be efficient distortion analysis that estimate weld angular distortion from welding process conditions with a high degree of accuracy. Additionally, the effect of welding process conditions on weld bead morphology, temperature profiles, and angular distortion were discussed through the developed analysis model to obtain a more detailed understanding of dominant factor influencing the weld angular distortion of MAG welded structural steel plates.

The coarse grained heat affected zone (CGHAZ) of T23 steel was assessed by strain-to-fracture (STF) tests during post-weld heat treatment (PWHT) at 750°C to investigate the microstructure changes and corresponding hot ductility. Scanning electron microscope (SEM), transmission electron microscope (TEM), and small angle x-ray scattering (SAXS) combining with JmatPro was used to study the characteristics of carbides, dislocation density, and lath size. Voids and micro-cracks were also analyzed by SEM. The results showed that sample with 1 min PWHT lost its hot ductility and intergranular cracking occurred. The hot ductility recovered with increasing PWHT time. However, the mechanisms were quite distinct during different time periods. Within PWHT time range from 0.5 h to 1.5 h, the recovery of hot ductility resulted from decreasing dislocation density. The hot ductility exhibited an opposite tendency with hardness. By contrast, the hot ductility and the hardness increased together after 2 h PWHT. It was resulted from the formation of lots of V-rich MC carbides in grain and coherent/semi-coherent smaller M₂₃C₆ carbides at the prior austenite grain boundaries (PAGBs). The strength of the grain interior and PAGB increased together. Plastic deformation was accommodated only by the cracking or shearing of low strength blocks. The fracture surface exhibited quasi-cleavage cracking together with dimples, which represented of good hot ductility. It could be concluded that the drastic changes of microstructure altered the manner of plastic deformation accommodation during stress relief, which led to the hot ductility loss and recovery during PWHT at 750°C.

The size and distribution of interphase precipitates in micro-alloyed steels is a crucial micro-structural feature to control for obtaining the necessary strength in low-cost automotive sheets. In order to optimize both alloy chemistry and thermal processing an enhanced understanding of the interphase precipitation mechanism is required. It is proposed that the evolution of inter-sheet spacing of MC carbides during the γ→α+MC transformation can be explained considering the interfacial segregation and the corresponding dissipation of Gibbs energy inside the moving interphase boundary. The inter-sheet spacing of interphase precipitates is controlled by a complex interplay between the interfacial energy and interfacial segregation, this is presented in form of an analytical model. It is shown that the general trend of refining inter-sheet spacing with growing ferrite half-thickness can be well predicted by the proposed model.

Microstructure evolution during the reverse transformation of low alloy steel consisting of lath martensite and chromium carbide was examined using in-situ electron backscatter diffraction at high temperatures. Austenite grains nucleated during the reverse transformation were categorized into two types: austenite grains nucleated along the lath boundaries with almost the same crystal orientation as the prior austenite (type A), and austenite grains nucleated at the prior austenite grain boundaries or inside the prior austenite grains with a different crystal orientation (type B). After the reverse transformation was finished, the prior austenite grains were reconstructed by the rapid growth and coalescence of type-A grains, which was called the “austenite memory phenomenon.” Here, misorientation remained in the reconstructed austenite grains. Hence, upon heating to a higher temperature, type-A grains were invaded and were eventually replaced by type-B grains, resulting in a new fine-grained microstructure; this was similar to recrystallization. Therefore, these results showed that the austenite memory phenomenon occurred when the nucleation and growth of type-A grains was more dominant than those of type-B grains and that the degree of grain refinement depended on the nucleation and growth rate of the type-B grains.

The changes in the dislocation substructure and internal stresses in a tempered 9% Cr heat resistant steel during creep at 923 K were studied. The mean lath size gradually increased from 330 nm for the initial tempered state to 740 nm once the specimen had crept to failure under a nominal stress of 118 MPa after 1271 hours. Correspondingly, the dislocation density within the lath decreased from 6.2 × 10¹⁴ m⁻² to ~10¹⁴ m⁻². The tempered structure of the martensite lath was characterised by large lattice curvatures, which were attributed to a high density of dislocations with like signs and the long-range stress fields originating from the martensite lath boundaries. An internal stress of 49 MPa, as evaluated by measuring the lattice curvature within individual laths in specimens crept to 1% (just after transient creep stage), is comparable to the threshold stress of 51 MPa estimated from the creep rate and stress relationship. The improved creep resistance of advanced 9% Cr martensitic steel results from both dispersion strengthening and the internal stresses of the martensite lath substructure.

Improvements in the high-temperature strength are desired for the ferritic Ni–Cr–Mo steels used for the brake discs in mechanical brake systems on railway. The objective of this paper is to construct a constitutive equation for the high-temperature deformation in Ni–Cr–Mo steels containing small amounts of V. The examined steels were obtained by varying the addition of V from 0 mass% to 0.27 mass% in the steel. The flow stress of Ni–Cr–Mo steels was increased significantly by the addition of V. It was found that the creep mechanism of examined Ni–Cr–Mo steels was climb-controlled dislocation creep controlled by the lattice diffusion. Owing to the increase in the number of vanadium carbide (VC) particles with increasing the concentration of V, the threshold stress increased due to the Orowan mechanism and as a result, the high-temperature strength also increased. The constitutive equation for the high-temperature deformation of Ni–Cr–Mo steels containing small amounts of V was constructed. The constructed equation is useful in modeling the high-temperature strength under similar heat conditions in brake discs.

A novel low-carbon high-silicon martensitic steel, 22SiMnCrNiMo, whose comprehensive mechanical properties approach the 00Ni18Co9Mo4Ti maraging steel, was investigated. Microstructure characterizations revealed that the steel was composed by lath martensite, retained austenite films and ε-carbides after different heat treatment processes. And the 22SiMnCrNiMo steel obtained optimal mechanical properties of a yield strength of 1261 MPa, a tensile strength of 1548 MPa, an impact toughness of 120 J/cm², a fracture toughness (K_IC) of 94.8 MPa·m^1/2 and a threshold value (ΔK_th) of 7.1 MPa·m^1/2 when the steel was austenitised at 900°C for 1 h followed by water quenched and then tempered at 320°C for 1 h. The 22SiMnCrNiMo steel exhibited excellent mechanical properties similar to those of the 00Ni18Co9Mo4Ti maraging steel as well as high resistance to fatigue crack initiation and growth.

To investigate hydrogen embrittlement susceptibility of an individual grain boundary, microcantilevers were fabricated on Ni–Cr bialloy surfaces by focused ion beam, and microbending tests were conducted during hydrogen charging by electrochemical nanoindentation. For microcantilevers fabricated across a twin boundary and inside a grain, no crack was formed by the bending both in air and during hydrogen charging. On the other hand, for microcantilevers fabricated across a random grain boundary, a crack was formed by the bending during hydrogen charging, whereas no crack was formed in air. Thus, it was identified that, for Ni–Cr bialloy, random grain boundaries are more susceptible to hydrogen than twin boundaries and crystalline planes. To validate these experimental results, slow strain-rate tensile tests were also performed, and subcracks of the fractured specimens were analyzed by electron beam back-scattering diffraction. In accordance with the microbending test results, subcracks were predominantly formed along random grain boundaries, but never along twin boundaries and inside grains. Higher hydrogen susceptibility of the random grain boundaries with lower cohesive energy (i.e., work for separation), smooth fracture surfaces without dimples or tear ridges, and no correlation between strain accumulation and subcrack formation indicate that hydrogen embrittlement of Ni–Cr bialloy is caused by decohesion mechanism, where hydrogen atoms lower cohesive energy at grain boundaries.

Mechanical properties of ferrite/cementite lamellar structure model in pearlite steel with the Pitsch-Petch orientation relationship are examined by a strain gradient crystal plasticity analysis. Dislocations’ behavior in the lamellar structure is especially considered in the analysis and characteristic lengths of the structure used in its constitutive equations are approximately determined for each slip system. From the analysis results for different lamellar thicknesses, a scale effect on the yield stress of the pearlite model is smaller than that in case of uniform estimation of the characteristic length by the lamellar thickness because of a reduction of the Orowan stress. The yield stress shows better agreement with an experimental result. On the other hands, a scale effect on the strain hardening rate is larger than that of the uniformly estimation. This is attributed to obvious change of active slip systems in the ferrite layer associated with the thickness. It is found that while slip systems with the largest Schmid factor act in case of large thickness, slip systems which have small critical resolved shear stress act in case of small thickness due to the increase of slip anisotropy. And the change of active slip systems arises large shear incompatibility in each layer and makes the stress fields in the layer non-uniform for the model with Pitsch-Petch relationship.

The recovery of antimony from high arsenic-bearing flue dusts was carried out by a selective oxidation roasting process using MnO₂, in which the arsenic was removed through a volatilization and antimony was oxidized to Sb₂O₄ staying in the roasted products. In a certain range, the MnO₂ additive has an active effective on the arsenic volatilization for the reason that structures of some complicated As-Sb phases were destroyed after the Sb₂O₃ being oxidized to Sb₂O₄ and this part of arsenic continued to volatile. About 90.06% arsenic and only about 6.89% antimony went into smoke under the condition of roasting temperature of 723.0 K, MnO₂ amount of 20.5%, and roasting time of 90 min. The MnO₂ selective oxidation provided a good separation of arsenic from high arsenic-bearing flue dusts. The left antimony existing in the roasted products can be reclaimed through a process of reduction roasting and dust collection.