Removal of Lead and Zinc from Blast Furnace Sludge Through High Temperature Carbochlorination

A new dual kiln processing route has been proposed and investigated for removal of heavy metals from the zinc-rich fraction of blast furnace (BF) sludge and recovery of ferrous materials. The zinc in BF sludge exists both as ZnO and ZnS. The thermodynamic analysis indicates that the carbochlorination of the sludge shows a selective heavy metal removal. Lead can be removed at a temperature below 900°C followed by the zinc removal at higher temperatures, producing an iron-rich clean residue free of heavy metals in the end feasible for recycling back to ironmaking. High temperature carbochlorination experiments have been conducted with designed mini-pellets. The experimental results prove that via the proposed dual kiln route, a valuable zinc-rich fraction and a clean iron-bearing residue can be achieved with addition of iron chloride, leaving a small stream of lead-rich residue for further processing. This newly developed recycling flowsheet provides a sequential carbochlorination in two inter-connected rotary kilns, incorporated with efficient dust-capturing systems. In total, over 98% heavy metal removal has been achieved. The Zn-rich dust contains more than 60% of zinc, and the quality of the iron-rich residue is sufficient for re-use in the integrated steel works.


Introduction
Steelmaking industry contributes greatly to the development of global economy, but at the same time generates significant amount of emissions to the environment.Today CO2 reduction and materials circularity become even more important than the economic growth.The development and application for sustainable and more efficient process technology will be crucial for the further improvement of the steelmaking industry.According to World Steel Association [1], global crude steel production has reached 1,878 million tons in 2022 with 4.2% decrease from 2021.The European Union produces 136.7 million tons of steel in 2022 with 10.5% decrease from 2021, with approximately 60% BF-BOF route and 40% EAF route at 500 steel production sites across 27 EU member states.Approximately 56% of EU steel is made from scrap, with around 100 million tonnes of scrap steel recycled every year [2].Increasing efforts have been made in recent years inside the European Union for developing CO2mitigating or even carbon-free H2-based technologies, improving resource efficiency and fostering sustainable and green transition.In 2020, the EU published a Circular Economy Action Plan, an important step in developing a truly circular economy in Europe, which helps to clarify the policies regarding recycling and the absolute minimum of waste generation from the steel production cycle.
Blast Furnace (BF) operation generates on average about 10 -15 kg/thm (hot metal) dry flue dust worldwide.It contains enriched levels of heavy metals, which can contaminate soil, water and air if not properly processed, leading to adverse environmental and ecological impacts.It is therefore essential to develop efficient methods with advanced clean-production techniques for heavy metal removal from blast furnace dusts.Tata Steel in IJmuiden of the Netherlands is one of the Europe's leading integrated steel production plants, with around 7 million tons annual rolled steel production.Two blast furnaces are in operation, with self-produced sinter, pellets, and coke.Steel scrap is recycled via steel converter with a controlled zinc input, and the steelmaking dust is recycled to BFs via sinter plant.There is a site limit for recirculating zinc in the integrated steel works.It is known that the heavy metals have a detrimental impact on the BF process stability.For example, zinc can circulate and deposit inside the furnace, and will cause scaffold in the upper shaft of blast furnace, decreasing the lining lifetime via wearing, and even creating issues during tapping.The heavy metals (Zn, Pb, Cd etc.) leave the blast furnace together with flue gas, and are enriched through a dry classifier followed by wet scrubber.The dry coarse fraction containing lower amount of zinc and lead is recycled internally in the sinter plant.The finer fraction is further processed via the hydro-cyclones into three fractions with low, middle and high zinc contents.Depending on the input limitation of zinc in the blast furnace, the low and partially the medium zinc containing fractions are recycled internally, and the high zinc fractions are currently landfilled as an external zinc exit stream.The landfill of high-zinc fractions has a higher environmental impact, and constant effort has been given in finding a more sustainable solution since 1980's at IJmuiden site, with various processing routes tested and evaluated, including both hydro-and pyrometallurgical processes [3 -6].
Chlorination is one of the effective techniques for the heavy metal extractions.It is known that zinc is present in the oxide form (ZnO) and sulfide form (ZnS), both of which can be converted into metal chloride.Metal chlorides are generally very volatile at relatively low temperatures, and often have different vapour pressure from one another [7].These characteristics lead to good selectivity in the metal separation via volatilization at a specific temperature utilizing the difference in the vapour pressure.A recent research was conducted by Hamann et al. in 2021 [8] in treating the BF sludge through direct reduction with the addition of spent pickling acid (FeCl2) to remove the heavy metals in a mixture of chlorides.A recycling solution to selectively remove the heavy metals is required, and preferably Pb could be removed from Zn at lower temperatures, leaving a zinc-rich fraction for further extraction and utilization.The small amount of Pb-rich fraction will be subjected to further processing, or landfilled after proper immobilization.In the present study, a new carbochlorination processing route was explored and tested, and the lab scale experimental results will be presented.

Raw materials and characterisations
The zinc rich BF dust is used in the present study, which is generated as filter cake after heavy metal enrichment via the hydro cyclones.Four mini-pellet samples, with and without FeCl2 addition, were prepared for the carbon-chlorination experiments.Table 1 lists the pellet size and its chemical compositions determined with Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) and ELTRA for carbon and sulphur analyses, and the mini-pellets are shown in Figure 1.The composition of each sample was the average of a duplicate analysis.The samples 1 and 3 have almost the same chemical compositions, and the only difference is the size.The samples 2 and 4 are with the same addition of FeCl2.4H2O(in two size ranges), calculated based on the stoichiometry required to remove lead and zinc present in the dust samples.The highest element present in the pellets is carbon (~38 -40 %), followed by iron (~25%).Zinc, lead, and cadmium content are around 2.6%, 0.5%, and 0.01%, respectively.The initial Cl level for the samples 1 & 3 is at the level of 0.13 -0.14% and it increases to the level of 2.1 -2.2% with FeCl2 addition to facilitate the heavy metal chlorination reactions.The zinc in the BF dust is mainly present as ZnO and ZnS identified with XRD, and the other possible zinc phase as zinc ferrite (ZnFe2O4) can hardly be quantified.Based on the XRD analyses, ZnO was found as zincite and ZnS was found as wurtzite and sphalerite in both samples.It is found that the zinc phases consist of approximately 1/3 of ZnO and 2/3 of ZnS.The mineralogical form of lead is not confirmed by XRD due to its very low content, but PbS was identified via Raman spectrometer previously.

Experimental
The experiments were carried out using two different furnaces.The first was the TGA furnace for the small-scale experiment focusing on mass changing during the heating and removal rate of zinc and other heavy metals.The second furnace was a horizontal tube furnace for a larger scale experiment focusing on the heavy metal removal and separation, and on generating secondary dust rich in heavy metals.The initial and the treated samples were analysed by using various analytical techniques.

Small scale TGA tests -thermal behaviour of BF dust
A Netzsch STA 409 TGA/DSC furnace was used for understanding the thermal behaviour of the mini pellets prepared from BF dust.The experiments in the TGA furnace were conducted to a maximum temperature of 800, 1000, 1200, and 1400°C for each of the four samples, respectively.The mini-pellets samples were first dried overnight at 105 °C.Each sample weighing between 100-200 mg was put in a crucible, placed next to a reference crucible inside the heating chamber and heated at 40°C/min starting from room temperature.After reaching the designated maximum temperature, the isothermal condition was maintained at the temperature level for 60 minutes.The nitrogen gas was introduced at 20 ml/min during the experiment for keeping an inert atmosphere.After the holding period of 60 minutes, the sample was then cooled through the water cooling system attached internally.Each test was repeated to ensure the reproducibility.

Selective heavy metal removal from BF dust in a horizontal tube furnace
The heavy metal removal experiments were carried out in a horizontal tube furnace, with an installed glass filter to capture the flue dust from the off gas, as illustrated in Figure 2. The stage-wise tests were designed to investigate the feasibility for selective removing and separating the heavy metals (in particular zinc and lead) under carbochlorination conditions.Samples 1 and 2 of smaller sized minipellets without and with FeCl2 addition were selected for the experiments, and about 40 grams of each sample was put in a stainless steel boat, and the boat was placed into the constant temperature zone of the tube furnace.The furnace was closed gas tight and flushed with dry nitrogen.
Two sets of experiments were designed.Table 2 and Table 3 give an overview of the stage-wise experimental design with the samples and the achieved conditions for both sample 1 (series A, without FeCl2 addition) and sample 2 (B series, with FeCl2 addition).For tests No. 1-8 (A1-A8, B1-B8), the same mini-pellets sample (A or B) went through 8 stage-wise heating, with a local holding of 30 minutes, at the temperature intervals specified in Table 2.Only filter dust samples were collected stage-wise following thermal treatment profile, with just one last residue pellet sample in the boat.In each stage a new glass filter was used.For the test No. 9-16 in Table 3, 8 individual experiments were conducted for both sample 1 and sample 2, each started at 150 o C and ended at the pre-set temperatures for 30 minutes.Each experiment will generate an accumulative dust and the residue (after reaction) samples.
Prior to the experiments the pellet samples were dried at 150°C for one hour in the same furnace.The glass filter was applied at the exit of the off-gas pipe and held at the position by load.At a given temperature, the sample was reacted for 30 minutes, then cooled quickly at the low temperature zone of the furnace by pulling out of the hot zone The residues after reactions and the captured secondary flue dust on the filters were subjected for further analyses.The residue samples and the captured secondary flue dusts were analysed with various analytical methods including ICP-OES, X-ray Fluorescence (XRF), X-ray Diffraction (XRD), Scanning Electron Microscopy and Energy Dispersive X-ray Spectroscopy (SEM/EDS), ELTRA for carbon, as well as wet chemical analysis.

Impact of chloride addition on the thermal behavior of BF dust -TGA experiments
TGA experiments were conducted at four different temperatures of 800°C, 1000°C, 1200°C and 1400°C with a constant heating rate of 40 °C/min in an inert atmosphere..When the temperature was reached, the isothermal heating was held for another 60 mins.The results of the mass change during the nonisothermal and isothermal heating periods for samples of 1-4 are compared and shown in Figure 3.It can be seen that the significant mass change occurred during the non-isothermal heating period.In addition, for all the samples heated up to 800°C (Figure 3(a)), there were continued considerable mass loss within the isothermal heating period.On the other hand, for all the samples heated up to 1000, 1200, and 1400°C, the kinetics of the carbochlorination reactions in the blast furnace sludge were sufficiently fast to be almost completed within the non-isothermal heating period.As the maximum temperature increases, the sample's residual mass becomes smaller, which means that more elements were converted into the gas phase.It is expected that the most generated gases were the mixtures of carbon monoxide and carbon dioxide during processing at temperature range from 800°C to 1400°C, due to the high presence of carbon and iron oxide in the dust samples.Another contribution would arise from the volatilization of zinc, lead, and cadmium compounds which could account for up to 5% when heated up to 1400°C, based on FactSage simulation.
The largest influence of chloride addition is shown by the samples heated to the maximum temperature of 800°C and 1000°C, with further increase in the maximum temperature, the influence of chloride on the mass loss is less significant.The samples with chloride addition (samples 2 and 4) have higher mass loss than those without chloride addition (samples 1 and 3).This difference is caused by the vaporization of crystal water in FeCl2 and by the chlorination reactions which promote the removal of heavy metals from the blast furnace dust with faster reaction kinetics.In this study, the difference of mass loss between sample 1 and 3, and between sample 2 and 4 are insignificant which reflects the insignificant effect of the pellets size.
Figure 4 illustrates the zinc and lead removal efficiency from the TGA experiments.The results show that all the sample reacted at temperatures higher than 1000°C led to a high removal rate of Zn and Pb, up to 98 and 99%, respectively.Furthermore, it is interesting to note that up to 98% of Pb has already been removed at 800°C from the samples with iron chloride addition (samples 2 and 4), however, only less than 40% Pb was removed from the samples without chloride addition (samples 1 and 3).This indicates that the carbochlorination reactions could lower the temperature required for the complete removal of Pb.Thermodynamic analyses of this reaction system were conducted with the FactSage simulations to support the experimental results, which is shown in Figure 5.The thermodynamic calculation results show a consistent agreement with the TGA results, and both show the chloride addition leads to a relatively lower temperature of lead removal, subsequently followed by zinc removal at higher temperatures.No significant impact from the pellet size could be observed by comparing the differences from samples 1 with 3 and 2 with 4, regardless of the effect of chloride addition.Usually the sample size would influence the removal rate, especially when the sample is heated at a high temperature.In this case, the heating temperature is sufficient to trigger the removal, and the sample of less than 5 mm is sufficiently small not to hinder the diffusion of the zinc and heavy metal volatiles.It demonstrates that zinc oxide, zinc sulfide, lead oxide/sulfide are thermodynamically favourable to be converted into chloride gas at a temperature higher than 800°C.A small sample (diameter <10 mm) could increase the diffusion rate from the inner core to the bulk, increasing the removal rate of heavy metals.Moreover, Zhu (2021) [12] reported a reduced removal rate of zinc when the sample size was increased from 12 mm to 18 mm, which is far greater than the diameter of the pellets used in this study.

Selective heavy metal removal via carbochlorination in the horizontal furnace
The pre-mixed mini-pellets were first dried in the same horizontal furnace and then reacted at given temperature for 30 minutes in N2 atmosphere, following the schemes described in Table 2. Due to the limited amount of secondary flue dust captured on the glass filters from the stage-wise experiments (Table 2), it is difficult to accurately quantify the total chemical compositions.Instead the flue dust deposited on the filters were dissolved following the certified procedures and analyzed for the relative ratio of Pb to Zn in mass with ICP-OES, and the results are given in Figure 6. Figure 7 shows the impact of chloride additions on the Pb to Zn ratio in the captured flue dust and solid residues following the scheme in Table 3, which illustrates the potential of an earlier removal of Pb starting from the temperature of 600 °C.At 900 °C, over 97% of Pb has been removed from the solid residue with FeCl2 addition, where the majority of zinc still remains in the solid residues, as seen in Figure 8. Figure 9 (af) shows the composition changes of various elements of Zn, Pb, Fe, V, Cd and Sn in the remaining residue as a function of temperature, respectively.It can be seen that the samples with chloride addition have a slight higher zinc removal rate than the samples without chloride addition (Figure 9a), probably due to the presence of ZnS as thermodynamically predicted in Figure 5.For lead removal, more considerable difference can be seen as the samples with chloride addition lead to significant lead decrease in the residue starting at 500 o C (Figure 9b), and have a removal efficiency of higher than 97% already at 900°C while zinc loss from the residues is insignificant (Figure 8).In contrast, the samples without chloride addition have only a lead removal efficiency of 42%, which also agrees with the results from the TGA measurements.At higher temperatures at 1100°C, the addition of chloride is no longer of influence on the removal efficiency of lead and zinc, and nearly all lead and most of the zinc have been removed from the residue.
Figures 9(c) and 9(d) show the variations of iron and vanadium in the residue with slightly increased concentrations at higher temperatures, and this is mainly due to the consumptions of carbon and the loss of heavy metals.However, some amount of iron chloride might be lost to the gas during chlorination, based on the thermodynamic assessment, but this requires further quantifications.Adding chlorides shows no impact on cadmium removal since its removal rate has already reached nearly 100% at 800°C for all samples, as shown in Figure 9(e).Figure 9(f) shows the similar behaviour of tin to that of lead, and this may create an option for Sn coated scrap recycling in the integrated steel works.
Upscaling the experiments with larger samples is essential to mimic the industrial case, especially to have a more representative composition of the generated secondary dust.In this study, the highest content of Pb in the lead-rich dust is about 30%, and Zn in zinc-rich dust is 68%.This quality is a baseline for the further utilization of the by-products from the recycling process.The ferrous residue after the thermal treatment contains very low heavy metals and can re-enter into the ironmaking route.
The metallization of iron was also confirmed by the wet-chemical analyses of the reacted samples (the residue), and the highest metallization degree of 44% was achieved for the samples without chloride addition.With increasing temperature, the iron metallization increases and stabilized after 1100 °C.The samples with chloride addition has a relative lower metallization degree.It is clear that addition of ferrous chloride from pickling liquor as source of chloride can lower the onset temperature and increase the lead removal through chlorination reactions, which is consistent with the thermodynamic simulations.

A flexible dual kiln route and perspectives
Compared to the hydrometallurgical routes, the pyrometallurgical processes have been used in the industry to treat zinc-bearing wastes, including blast furnace sludge.It offers a high levels of zinc and heavy metals removal through roasting, direct reduction, fuming or smelting.The pyrometallurgical route can remove zinc compounds, including the phases resistant to leaching (e.g., ZnS, ZnFe2O4), usually being a hindrance to the hydrometallurgical route [6,9].The direct reduction of the sludge/dust results often in good quality products for direct use in the relevant industry, consisting of an iron-rich fraction as the residue and a zinc-rich fraction as flue dust (containing heavy metals) from off-gas filtering.The high carbon content of ironmaking flue dust, especially from the blast furnace could promote self-reducing behaviour for direct reduction without the need for reducing agent for both iron and heavy metal constituents.The limitation lies within the minimum zinc content based on the economic perspective.Some pyrometallurgical techniques require a high content of zinc input with specific input tonnage.For example, Rotary/Waelz Kiln is mainly used to recycle the dust with more than 15% of zinc, commonly contained in the EAF dust.Moreover, pyrometallurgy involving a high temperature furnace requires a substantial initial capital investment, commonly used to extract highgrade ores.Blast furnace sludge with low zinc content (1.5 -8%) is limited to particular pyrometallurgical techniques to achieve a positive economic benefit.The established smelting-based processes consume significant amount of energy to melt the input materials entirely.Slag is then generated as a by-product that needs additional processing for utilization.
Most pyrometallurgical techniques can remove over 90% of zinc.The direct reduced iron or the ironrich residue contains a high amount of metallic iron which can be directly sent to the sinter plant or even blast furnace after briquetting.As discussed, BF sludge contains a large amount of carbon compared to the dust from EAF and BOF steelmaking.The carbon within the dust or sludge can lead to self-reduction of the materials without the need for coke or reducing gas (CO) which has been shown effective in producing high-grade (partly) metallic iron and zinc-rich secondary dust.
Direct reduction in the rotary kiln has been well established in the industry, with a high removal rate of zinc up to 94% [10 -13].Around 83% of EAF dust is recycled through this process, with the capacity in the range of 100 -350 kt/year.[10].The typical input for Rotary Kiln or Waelz Kiln is pelletized mixture of dust, binder, pulverized coal, and water.At a temperature of 1100 to 1200°C, the reduction of metal oxides starts to occur while zinc and alkali metals will be selectively volatilized.The reactions in the Waelz kiln occur in the bottom area for reduction and upper area for oxidation, with a countercurrent movement of the solid feed and injected hot air in a rotating chamber.Rotary kiln has the advantages of the relatively straightforward process of low environmental pollution and has been widely used in treating EAF dust with high zinc content (>16%) [10].The problem of the rotary or Waelz kiln is the solid ring formation on the furnace lining due to the adhesion of elements with a low melting point as the kiln rotates.Compared to the rotary kiln mainly used for EAF dust (high zinc) processing, RHF technology has been widely used to treat BF and BOF dust with the final product of DRI containing high Fe, below 0.004% of Zn with almost no lead [10].However it requires higher capex and has no selectivity for heavy metals.
Recycling of blast furnace sludge has been theoretically feasible and proved by the high removal efficiency of the zinc and other heavy metals selectively through the present experimental works.The proposed flowsheet as shown in Figure 10 is based on a dual-kiln selective heavy metal removal process, which launches a new recycling route with higher flexibility in waste processing.The recycling process starts with raw materials preparation using a designed recipe for pelletizing.The addition of chlorination agent together with bentonite as binder promotes heavy metal removal at lower temperatures, in particular for lead compounds.The carbochlorination reaction creates the selective removal of Pb at 900 o C or below (together with Cd), before zinc is substantially removed.Thus the preliminary direct reduction takes place at 800 -900°C in the first reactor for removing the environmental harmful elements like Pb and Cd, the dust from which requires further investigations for safe disposal.The solid residue will be transported hot to the second reactor for further processing at 1100°C to capture zinc as a valuable secondary raw material for the zinc smelter.The maximum operating temperature will be preferably lower than the softening temperature of the residue material to reduce the liquid formation.Finally, nearly heavy metals-free residue can be recycled back to the ironmaking process internally.The selection and design of a new process, such as the proposed dual kiln process, for waste minimization in an integrated steel works offers several benefits.One of the key advantages is potential for a significant reduction in waste generation for external treatment or landfill.The rotary kiln process is known for its ability to efficiently process various types of waste materials, by converting the waste streams into useful resources.The dual kilns increase the flexibility and selectivity in the process operation and targets control.This minimizes the amount of waste sent to landfills, therefore reducing environmental pollution and the associated costs of waste management and disposal.Additionally, the use of this new process can enhance the overall waste management and reduce the risk of spread of environmental harmful elements.On the other hand, implementing a new technology like the dual kiln process involves substantial upfront investment and operational costs.The installation and maintenance of the required equipment along with the operational investment can be expensive and time consuming.Moreover, transition to a new green production route requires modifications to the existing infrastructure and production system, may cause potential disruptions to the existing operations and create risks to the new market situation.
It is therefore important to consider the potential risks associated with the new process, such as emission control related to the carbochlorination, the influence of the added chloride.The final processing of the heavy metal chlorides requires further investigations in the follow-up research.Careful evaluation and planning are necessary to ensure that the benefits outweigh the drawbacks and that the chosen process aligns with the long-term goals of the business and the local regulatory requirement.

Conclusions
The present study aims to investigate the efficiency and viability of selective removal of the heavy metals in the BF flue dust or sludge via carbochlorination reactions.Through this research, the following key findings have been obtained.
(1) Carbochlorination offers a new route to selectively remove the heavy metals from BF dust at relative lower temperatures.The experimental results reveal a high removal rate of heavy metals (Zn, Pb, Cd) at a relatively lower temperature and could produce nearly a heavy metal-free ferrous residue along with the partial self-reduction of iron oxides.(2) Adding chloride reduces the onset reaction temperature and thus lowers the temperature of the carbochlorination reactions.High removal efficiency of lead is achieved at the temperatures of below 1000°C.The addition of chlorides improves the removal kinetics of the heavy metals, completed at 800°C (mainly for Cd and Pb), leaving a cleaner Zn-rich secondary dust and a ferrous residue nearly free of heavy metals.(3) A high Pb/Zn ratio in the secondary flue dust during early removal stage at 800 -900 °C has been achieved, which requires further investigation for proper processing and disposal.In contrast, zinc is concentrated in the dust generated at a distinctively higher temperature of around 1000°C.Zinc concentration of up to 60% with very low Pb was captured in the secondary dust generated at 1100°C, which can be extracted more efficiently to close the zinc cycle.(4) The proposed flowsheet provides a sequential double-step reduction in two inter-connected kilns or other reactors with separately attached dust capturing system, resulting in the final products of iron-rich ferrous residue and zinc-rich dust as new secondary resources.With using a chlorination agent such as FeCl2, the selective removal of lead and cadmium (even tin) over zinc at lower temperatures of 800 -900 o C create a strong business case for easier zinc recovery as secondary resource.
In conclusion, the recycling of blast furnace sludge with a novel double-step direct reduction and carbochlorination is a favourable and attractive route.The developed process could be implemented with low risks, subject to a further economic feasibility study.

Figure 2 :
Figure 2: Schematic diagram of horizontal tube furnace for carbochlorination tests

Figure 4 :
Figure 4: Removal of zinc and lead from the BF dust during the TGA experiments: effect of maximum temperature.

Figure 5 :
Figure 5: Equilibrium compositions of zinc and lead in the solid residues based on the thermodynamic analyses with FactSage's equilibrium module.

Figure 6 :Figure 7 :
Figure 6: Impact of FeCl2 addition on the removal of Zn and Pb from the BF dust at different temperature range (following experimental schemeTable 2-A)

Figure 8 : 9 (
Figure 8: Zinc and lead in the remaining solid residue in the presence of FeCl2 addition

Figure 9 :
Figure 9: Impact of chloride addition on the composition changes of various elements in the residue

Figure 10 :
Figure 10: Proposed flowsheet for BF dust processing and heavy metal separation

Table 1 .
Chemical compositions of the prepared mini-pellets mm Cl, wt% C

Table 2 :
An overview of experiment design and the samples collected for analyses (A series for sample 2, and B-series for sample 1): stage-wise experiments

Table 3 :
An overview of experiment design and the samples collected for analyses (A series for sample 2, and B-series for sample 1): continuous experiments