Refining of Secondary Pb with Retention of Sn Using Al and Ca Additions

GREEN AND LOW-CARBON EXTRACTIVE METALLURGY OF NONFERROUS METALS Refining of Secondary Pb with Retention of Sn Using Al and Ca Additions DANIEL MALECHA , 1,2,4 PAWEŁS ´WIEC, 3 ROBERT ALBRECHT, 3 PIOTR JAROSZ, 1 and STANISŁAW MAŁECKI 1 1.—Faculty of Non-Ferrous Metals, AGH University of Krakow, Al. Mickiewicza 30, 30-059 Krako ´w, Poland. 2.—Baterpol SA, ul. Obr. Westerplatte 108, 40-335 Katowice, Poland. 3.—Institute of Materials Engineering, University of Silesia, 75 Pułku Piechoty 1a, 41-500 Chorzow, Poland. 4.—e-mail: [email protected]l Lead grid from spent lead-acid batteries contains significant amounts of tin and antimony. In classical pyro-refining processes of lead, tin oxidizes and is transferred to dross, making its recovery problematic and expensive. This paper presents an innovative method of pyro-refining lead using metallic aluminum and calcium to purify the lead from contaminants while retaining a higher amount of tin than in the traditional process. The changes in the chemical composition of an impure lead alloy containing tin, under the influence of refining by adding Al and/or Ca, are discussed based on laboratory-scale studies. Microanalysis of the metallic dross formed during the process was conducted. Analyses of the metallic dross microstructures showed that lead impurities, such as Sb, As, Cu, Se, and Te, tend to accumulate in areas containing Al or Ca. The amount and form of dross produced in industrial practice indicate that its removal would be challenging. Therefore, in the second part of the study, the metallic dross was subjected to a reduction process, and the resulting products were analyzed for their chemical and phase composition. The analyses indicate that the degree of impurities return to the lead during reduction varied significantly depending on the method used. INTRODUCTION Despite the growing popularity of lithium-ion batteries and innovative energy storage technologies such as Mg-air solid-state or sodium-ion batteries, 1–3 lead–acid batteries (LAB) are still the most popular and have the highest recycling rates. 4 LABs are the primary source of electrical power in the automotive sector, not only for traditional internal combustion vehicles but also for electric cars, hybrids, electric bicycles, and scooters. 5 Additionally, these batteries are crucial for emergency power in buildings, off-road lighting, and home photovoltaic systems. 6–9 The life cycle of LABs is a model example of a closed-loop economy, with high recycling efficiency in developed countries largely due to existing legislation. 10–14 In the comprehensive recycling process of waste lead-acid battery (WLAB), the only waste produced is slag, which can be further recycled and utilized as an additive in gypsum or cement. 15,16 Despite being one of the oldest types of batteries, with over 160 years of use, the technology behind LABs continues to evolve. Innovations primarily focus on enhancing the performance and extending the service life. 17,18 A promising direction for the development of batteries is hybrid lead–carbon systems that are capable of delivering more power and provide a longer life cycle compared to traditional LABs. 19–22 The 12-V LAB market is growing every year and, according to forecasts, in 2025 the total capacity of this type of batteries will be around 664 GWh, while in 2030 it will be as much as 774 GWh. The greatest growth will occur in applications for stationary energy storage for renewable power sources and emergency installations (UPS). 23,24 (Received June 25, 2024; accepted September 16, 2024; published online October 7, 2024) JOM, Vol. 77, No. 5, 2025 https://doi.org/10.1007/s11837-024-06892-w Ó2024 The Author(s), corrected publication 2025 3249 In automotive applications, particularly for starting, lighting, and ignition, LABs typically have a 2to 4-year lifespan, leading to the disposal of hundreds of millions of these batteries globally each year. 25,26 The WLABs recycling industry is dominated by two primary technologies. The first method involves mechanically separating the battery into its components: lead paste (30–50%), metallic fraction from lead grids and terminals (24–30%), electrolyte (10–25%), polypropylene from casings (5–8%), spacers (4–7%), and scrap iron (about 2% of the battery weight). 20,27–30 The lead from grids, terminals, and battery paste is remelted and refined. Polypropylene re-granulate is produced from the battery casing, while the electrolyte can be purified or neutralized, e.g., as sodium sulfate. The second technology involves remelting the entire WLAB in shaft furnaces after removing the electrolyte. 4,31–33 Depending on the type and purpose of the battery, lead battery grids often contain alloying additives, primarily antimony and tin, with smaller amounts of calcium, aluminum, selenium, and arsenic to improve various properties. 34–36 In recent years, there has been an increasing trend in the tin content in these batteries. 37 The metallic scrap containing lead grids and connectors is typically processed in a rotary kiln or melting furnace, which does not require an extensive smelting cycle due to the minimal presence of oxides/sulfates. 38 Nevertheless, the raw lead obtained in this process must be refined each time to meet the chemical composition requirements for lead alloys used in new batteries. Pyro-refining methods are most used in the WLAB recycling industry, where hydrometallurgical or electrolytic methods are far less common. 4,39–46 The primary pyro-refining techniques involve an oxidative process to remove major impurities such as antimony and arsenic through selective oxidation with injected oxygen or oxidizing agents like sodium nitrate and caustic soda (Harris process). 29,47,48 In both cases, tin often transfers to dross and dust, making its recovery challenging and requiring expensive hydrometallurgical processes. 49–51 The demand for tin, driven by its unique physical and chemical properties, has increased rapidly worldwide, and newer methods are being developed to recover it from lead scrap and other recycled materials. 52–55 In current lead alloys for manufacturing lead battery grids and the bridges that connect them, tin is the main alloying additive used, and its content can exceed 3%. The addition of Sn for the grid lead alloys increases its strength and reduces corrosion, in addition to improving the castability. 34,36,56 The literature describe the use of aluminum in the fire refining of lead to remove impurities such as antimony, arsenic, copper, nickel, and tellurium. 37,57–59 An overview of the studies and obtained patents on this topic is presented in in supplementary Table S-I (refer to online supplementary material). Aluminum does not form permanent compounds with lead; however, if antimony is present in the lead, aluminum combines with it and an AlSb compound is formed, with a melting point of about 1050°C. Based on the analysis of phase diagrams of aluminum with the other lead impurities mentioned, an analogous refining mechanism can be assumed. 60,61 This paper presents a novel approach to refining lead using aluminum and calcium, aiming to selectively remove unwanted impurities while preserving the tin content. The refined lead with tin and calcium will be the optimal base for producing lead alloys currently used for LAB manufacturing. The research mainly focused on the microstructure of the metallic dross formed by adding aluminum and/ or calcium was examined using electron microscopy and to determine its chemical composition. Theoretical thermodynamic calculations were conducted to list the reactions that could take place in liquid lead. The metallic dross formed after refining was further reduced to recover an additional amount of the metallic lead. The resulting dross was examined by phase analysis based on x-ray diffraction (XRD). The research carried out on a laboratory scale confirmed the validity of patents appearing in recent years based on mixtures containing Al and Ca, intended for refining lead while retaining tin. 62–65 To the authors’ best knowledge, apart from the developed patents, there are no data in the literature on the use of Ca or Al together with Ca in lead pyro-refining for the removal of impurities. There is also no literature on the metallic dross generated in such a process. The possibility of removing selenium from lead using aluminum has also not been previously presented in the literature. MATERIALS AND METHODS Materials Under laboratory conditions, three test series were carried out on lead with the identical chemical composition shown in Table I(the complete chemical composition is provided in supplementary Table S-II). This tin content in the lead was 1.46 ±0.05 wt.%, which is comparable to the tin levels typically found in Pb-Sn-Ca alloys used for manufacturing battery grids. 66 The Sb content corresponded to the average amount of this element found in typical LAB recycled lead, 37 while the content of elements such as Cu, Se, Ni, and Te was notably higher than average. Refining agents used included pure aluminum in the form of sheets (see supplementary Fig. S-1a), with a purity exceeding 99.8 wt.% (the primary impurity was iron at 0.15 ±0.04 wt.%, the full chemical composition is shown in supplementary Table S-III), and calcium metal in pellet form (see supplementary Fig. S-1b), with a purity greater than 98.7 wt.%, with magnesium and aluminum as the main impurities at 0.65 ±0.08 wt.% and Malecha, S ´wiec, Albrecht, Jarosz, and Małecki3250 0.25 ±0.05 wt.% respectively (detailed chemical composition in supplementary Table S-IV). Methods For laboratory testing, 12 kg of lead with the chemical composition listed in Table Iwas used. Three lead refining trials were conducted. In the first variant, 60 g of Al was added to the lead (0.5 wt.%), in the second, 60 g of Ca (0.5 wt.%) and in the third, 40 g of Al and 20 g of Ca (0.33 wt.% and 0.17 wt.%, respectively). The lead alloy was weighed on a scale (Axis model B15), with a reading accuracy of 5 g. Al and Ca additives were weighed on a scale (Radwag model WAS 220/X), with a reading accuracy of 0.1 mg. The additives were introduced into the molten lead, maintained at 680°C using a steel basket, the temperature selected based on prior experiments to ensure the rapid dissolution of the additives. The experimental setup is illustrated in Fig. 1a and consists of a crucible (resistance) furnace (PT 12/7OST; Czylok), equipped with an automatic temperature control system an external electronic temperature gauge (EMT-111; Czaki Thermo-Product), an electric stirrer (JJ-1; Vevor) with adjustable speed, and a steel basket for inserting the refining additives. After inserting the steel basket with the refining additive into the lead, a stirrer was turned on, which rotated it at a constant speed of 150 rpm. Subsequently, the lead alloy was cooled, and its temperature was continuously monitored. Every 20 min, the stirrer was turned off to collect the metallic slime from the alloy surface using a perforated paddle, and two samples of lead were collected for chemical composition analysis. This activity was repeated six times. The total experiment time for each of the three samples was 120 min. Microstructure and chemical composition analyses were performed for the first collected metallic dross from each sample. In addition, the metallic dross collected in sample one (after adding Al) and sample two (after adding Ca) was subjected to a reduction process using sawdust, followed by XRD phase analysis. Analysis of Chemical and Phase Composition Optical emission spectroscopy with spark excitation (OES; ARL iSpark 8860 Fire Assay Analyzer; Thermo Scientific) was utilized for analyzing the chemical composition of both the metallic lead and the metallic sludge. This spectrometer, specifically dedicated to lead analysis, facilitated very highprecision measurements. Each OES analysis was performed on at least two samples, and the results were expressed as average values. Measurement uncertainty with a 95% confidence interval was calculated through the spectrometer’s dedicated OXSAS software (v.2.6). The measurement error calculated by the software includes uncertainties due to the standard deviations between the individual sub-measurements of which the result consists and the uncertainty ranges of the certified reference materials. For the analysis of the chemical composition of aluminum, calcium metal, and the calcium–aluminum alloy used in the study, as well as the metallic dross formed during the process, an inductively-coupled plasma–emission spectrometer Table I. Element composition of lead for laboratory research (wt.%) Element Sn Sb As Cu Se Ni Te Content 1.46 ±0.05 1.51 ±0.04 0.15 ±0.01 0.12 ±0.02 0.11 ±0.02 0.011 ±0.002 0.062 ±0.005 Fig. 1. The laboratory-scale research platform consisted of the following components: (1) resistance furnace with an automatic temperature control system (capacity of 15 kg of lead), (2) an external temperature sensor, (3) an electric stirrer with adjustable speed control, and 4 a steel basket into which the refining agent was placed. Refining of Secondary Pb with Retention of Sn Using Al and Ca Additions 3251 (ICP-OES; Ultima Expert; Horiba Scientific) was employed. All the materials (except metallic Al) analyzed on ICP-OES were pre-crushed on a jaw crusher (BB 100; Retsch). The next step of sample preparation was grinding it on a mortar grinder (Pulverisette 2; Fritsch) until the grain size was below 0.1 mm, and then 0.5 g ±0.001 g of the averaged sample was taken from the material thus prepared. In the case of metallic Al, a fragment was simply cut out. Samples for ICP-OES tests were weighed on a precise electronic laboratory scale (WAS 220X; Radwag). All the materials were dissolved in aqua regia (HNO 3 + 3 HCl) before analysis on ICP-OES. X-ray diffraction (XRD; Empyrean; Malvern Panalytical) measurements were performed using a nickel-filtered Cu Ka 1,2 source (k= 1.5406 A ˚) operating at U= 40 kV and I= 30 mA and equipped with a PIXcell 3D ultra-fast solid-state hybrid detector. Measurements were carried out in a reflection mode, in the Bragg–Brentano geometry (h–hscan technique), within the 2hrange of 10–120°. Soller slits of 0.04 rad on both incident and reflected beam and the divergence slit of ½°in the incident beam were used. XRD phase analysis was performed using the International Centre for Diffraction Data (ICDD) PDF-4 database reference standards. Microstructure Characterization The observation of the metallographic samples was performed using a scanning electron microscope (SEM; JSM-6480; JEOL) operating at 20-kV voltage, equipped with an energy-dispersive x-ray spectroscopy (EDS; IXRF) detector. Images were captured in secondary electrons (SE) mode at magnifications ranging from 91000 to 92000. This setup facilitated detailed observation of the microstructure and precipitates. RESULTS AND DISCUSSION Thermodynamic Calculation of Main Reactions Al-Pb and Al-Sn binary systems do not form intermetallic compounds that enable permanent connections. 29 Thus, adding aluminum to a lead– tin alloy bath at temperatures above 660°C results in its complete dissolution. On the other hand, adding aluminum will react to form solid compounds with impurities found in lead, such as Sb, As, Se, Ni, and Te. The reactions that should then take place can be described as: Al þSb ¼AlSb ð1Þ Al þAs ¼AlAs ð2Þ 2Al þ3Se ¼Al2Se3ð3Þ 3Al þNi ¼Al3Ni ð4Þ 3Al þ2Ni ¼Al3Ni2ð5Þ 2Al þ3Te ¼Al2Te3ð6Þ 3Ca þ2Sb ¼Ca3Sb2ð7Þ The melting points of these compounds are significantly higher than the maximum processing temperature of 690°C, 51 so they will remain in the solid state. To investigate the thermodynamic feasibility of the above reactions, HSC Chemistry software was employed to calculate standard Gibbs free energy changes. The results shown in Fig. 2reveal that the DG values for all the reactions are negative and increase with increasing temperature. Notably, the probability of reaction 7, combining Ca with Sb, is the highest, and the probability of reaction 1, combining Al with Sb, is the lowest. This sequence indicates a hierarchy in the reactivity of aluminum with the impurities: Se, Te, Ni, As, Sb. Use of Al and Ca in Laboratory-Scale Pb Refining Change in the Chemical Composition of Lead Under the Influence of Al and Ca Additives Throughout the refining studies, metal samples were periodically collected and analyzed to assess the concentration of basic impurities in lead. Figure 3a illustrates the changes in the concentrations of several impurities, antimony, arsenic, copper, selenium, nickel, tellurium, and tin, as well as the temperature variations of the lead bath during the refining process with aluminum. Fig. 2. Relationship between DG and temperature of the reactions 1–6. The data presented in the graph were obtained using HSC Chemistry v.6.1 software. Malecha, S ´wiec, Albrecht, Jarosz, and Małecki3252 The lead alloy refining rate using aluminum was significantly higher than that of classical oxidation methods. 37 The expected effect was that the tin concentration would remain constant. According to the Gibbs free energy changes calculated (Fig. 2), selenium was removed from the lead most rapidly and efficiently, while antimony was the slowest and least efficiently removed. These results are consistent with previous industrial-scale studies employing aluminum in lead refining. 37 With the addition of aluminum and a decrease in temperature from 681°C to 360°C over 120 min, the content of impurities changed as follows: antimony decreased from 1.51 ±0.04 wt.% to 0.96 ±0.04 wt.%, arsenic from 0.15 ±0.01 wt.% to 2 ±1 ppm, copper from 0.12 ±0.01 wt.% to 29 ±3 ppm, selenium from 0.11 ±0.02 wt.% to 0.1 ±0.2 ppm, nickel from 0.011 ±0.002 wt.% to 0.3 ±0.2 ppm, and tellurium from 0.062 ±0.005 wt.% to 42 ±5 ppm (as shown in Fig. 3a). The changes in the chemical composition of lead indicate that the predicted reactions 1–6are likely to occur. The aluminum bonded with the contaminants, creating solid compounds insoluble in lead, as, for example, AlSb, AlAs, Al 2 Se, etc. The obtained relationships confirm the conclusions resulting from thermodynamic calculations. The order in which metals are removed from lead is practically confirmed. High removal rates are observed in the case of selenium, nickel, copper, and arsenic. In the case of arsenic, this may be due to the fact that arsenic removal is promoted by the presence of copper. The rate of tellurium removal is slightly lower, but it is also eliminated to low contents. Antimony is the last to be removed, but may be due to the much higher concentration of this element than other impurities. Thermodynamic calculations also indicate the probability of its removal being last (Fig. 2). During the period of 2 h analyzed in the graph (Fig. 4a), the average antimony content was 1.3 wt.%, while the level of its removal was 0.27 wt.%/h. In the studies carried out on an industrial scale, where Sb was the main pollutant, a curve was determined for the rate of Sb removal using Al depending on its concentration in the lead. For the content of Sb 1.3 wt.% analyzed in the experiment, the antimony removal rate was 0.16 wt.%/h. 37 Despite obvious differences resulting from the scale of the study, the lead cooling speed (which was much slower on an industrial scale) and the influence of other contaminants, the result of the Sb removal rate was similar. A second trial using calcium instead of aluminum involved adding 60 g of Ca (with a purity above 99.7%). This process was slower and less effective. With the addition of Ca and a temperature decrease from 680°C to 356°C over 120 min, the impurity levels were reduced as follows: antimony from 1.51 ±0.04 wt.% to 1.12 ±0.04 wt.%, arsenic from 0.15 ±0.04 wt.% to 0.066 ±0.005 wt.%, copper from 0.12 ±0.01 wt.% to 0.087 ±0.009 wt.%, selenium from 0.11 ±0.02 wt.% to 67 + 10 ppm, nickel from 0.011 ±0.002 wt.% to 42 ±5 ppm, as shown in Fig. 3b. The changes in element concentrations differed somewhat from those observed with aluminum usage. The gradual reduction in the levels of arsenic, selenium, copper, antimony, and nickel can be attributed to both the reactive interactions of Ca with these elements and the temperature reduction, leading to the formation of metal compounds. The precipitation of copper at temperatures close to lead solidification, similar to the drossing (liquation) stage in classical fire refining, resulted in copper dross formation. 47 The tellurium content decreased Fig. 3. (a) Changes in the concentration of elements, temperature, and mixing speed during the refining test using Al. (b) Changes in the concentration of elements, temperature, and mixing speed during the refining test using Ca. Refining of Secondary Pb with Retention of Sn Using Al and Ca Additions 3253 minimally, from 0.062 ±0.005 wt.% to 0.054 ±0.005 wt.%, with the measured difference falling within the range of measurement uncertainty. Changes in the content of the analyzed elements in lead are of a different nature than when using aluminum. Similar relationships occur only in the case of antimony (reaction 7) and selenium. Which, in the case of antimony, was expected according to its phase system with calcium. 67 Arsenic and copper are removed more slowly and to lower concentrations than when aluminum is used. The reduction in the copper content in lead is probably caused by the formation of Ca-Cu intermetallic compounds. 68 The concentrations of tellurium and nickel do not change much despite the possibility of their intermetallic compounds forming with calcium. 69,70 The change in concentrations in connection with temperature allows us to conclude that the elimination of most of the analyzed elements is determined by the change in their solubility in lead with temperature. It is important to maintain a constant level of tin in lead, despite the presence of tin–calcium intermetallic compounds. 71 This is probably because the tin is only removed after the remaining contaminants have been removed to a low level. This may be confirmed by a much higher value of the Gibbs free energy (DG) for the reaction of Ca with Sn (DG= – 128 kJ/mol Ca calculated for 700°C in the HSC 6.1 program) than, for example, for reaction 7whose DG expired in Fig. 2. In the third test series, a combination of 40 g of Al and 20 g of Ca was utilized. In this test, it was decided to check whether adding Al and Ca at the same time would bring any additional measurable refining effects. The premise for this thesis were patents that appeared in recent years regarding lead refining while preserving tin by using a mixture of Al with Ca and coke breeze. 62–65 Figure 4a illustrates the changes in the concentrations of antimony, arsenic, copper, selenium, nickel, tellurium, and tin, as well as the temperature variations in the lead bath during the refining process. By adding Al and Ca and lowering the temperature from 680°C to 351°C over a period of 120 min, significant reductions in impurity concentrations were achieved: antimony decreased from 1.51 ±0.04 wt.% to 0.52 ±0.02 wt.%, arsenic from 0.15 ±0.01 wt.% to 0.6 ±0.05 ppm, copper from 0.12 ±0.01 wt.% to 16 ±3 ppm, selenium from 0.11 ±0.02 wt.% to 0.1 ±0.2 ppm, nickel from 0.011 ±0.002 wt.% to 0.1 ±0.1 ppm, and tellurium from 0.062 ±0.005 wt.% to 31 ±5 ppm. These changes suggest likely reactions between the impurities in the lead and both aluminum and calcium. The results of this measurement series indicate that the interactions of aluminum and calcium add up. This is determined by both the formation of intermetallic compounds and the decrease in their concentration in lead with lowering temperature. This results in an increase in the speed of the process of removing lead contaminants and their deeper removal. In each of the three samples conducted, the tin content remained constant, and the differences in concentration were within the range of measurement uncertainty. Figure 4b summarizes all three tests, displaying the percentage removal of lead contaminants, including Sb, As, Cu, Se, Ni, and Te. Temperature measurements at successive intervals for all three samples were very similar, and Fig. 4b shows the calculated average temperature value, demonstrating that the highest degree of lead refining was Fig. 4. (a) Changes in the concentration of elements, temperature, and mixing speed during the refining test using Al and Ca. (b) Contaminant removal rate for each of the three trials. Malecha, S ´wiec, Albrecht, Jarosz, and Małecki3254 achieved in test three using 40 g of Al and 20 g of Ca, in which 73.1 % of the impurities were removed in 120 min. In comparison, using only Al resulted in 50.5 % impurity removal, while using only Ca achieved 31.9 % impurity removal. These relationships confirm the conclusions resulting from the analysis of subsequent studies using various refining agents. Microstructure of Metallic Dross In each of the three samples, metallic dross was collected from the surface of the lead alloy before sampling for chemical composition analysis (supplementary Fig. S-2). The most significant removal of impurities in each sample occurred within the first 20 min of the experiment (Figs. 3,4); therefore, the dross collected during this period was considered optimal for the analysis of the intermetallic phases present. Figure 5a and b shows the microstructure of the collected metallic dross in the sample using aluminum. These dross exhibited a very heterogeneous appearance (supplementary Fig. S-2a), which corresponded with their chemical composition analysis and a high standard deviation of the measurements. The chemical composition of the metallic dross is detailed in supplementary Table S-V. Lead was the primary component, with Sb content at 3.5 ±0.4 wt.%, As 3.2 ±0.3 wt.% and Cu 1.6 ±0.2 wt.%. The content of elements such as Se, Te, and Ni was approximately 0.2 wt.%. It is important that the tin content was at the level of 1.0 ±0.1 wt.%, i.e., lower than the lead alloy from which the dross came, where the tin content was at the level of 1.46 ±0.05 wt.%. SEM analysis revealed areas enriched in aluminum. At points where Al was present, clusters of elements such as Sb and Te were also visible (Fig. 5a), indicating that Fig. 5. (a) SEM image and the corresponding EDS element mappings of Pb, Al, Sn, Sb, and Te for a sample of metallic dross after using Al as a refining additive. (b) SEM image and the corresponding EDS element mappings of Pb, Al, and Cu for a sample of metallic dross after using Al as a refining additive. Refining of Secondary Pb with Retention of Sn Using Al and Ca Additions 3255 these elements had formed intermetallic compounds. In another area of the same analyzed sample cross-section, at the same magnification, it was possible to observe areas also rich in aluminum but containing mainly copper and an evidently increased concentration of selenium (Fig. 5b). Interestingly, the aluminum areas containing antimony and tellurium shown in Fig. 5a were poor in Cu and Se. On the other hand, the aluminum areas containing copper and selenium in Fig. 5b were poor in Sb and Te. This may indicate different times of their formation during the refining process. Figure 6shows the microstructure of metallic dross collected in a sample using calcium. The chemical composition is shown in supplementary Table S-VI. The primary component of the metallic dross was Pb, with Sb at 1.6 ±0.3 wt.%, As at 1.5 ±0.2 wt.%, Ca at 0.8 ±0.2 wt.%, Cu at 1.3 ±0.2 wt.%, Se at 0.11 ±0.05 wt.%, Te at 0.06 ±0.03 wt.%, and Ni at approximately 0.01 wt.%. The tin content was 1.3 ±0.1 wt.%, which was close to the value in the refined alloy (1.46 ±0.05 wt.%). SEM analysis identified areas rich in calcium. At points where Ca was present, clusters of elements such as Sb, As, Se, Cu, and Te were also visible, suggesting that these elements had combined into complex compounds. Figure 7reveals the microstructure of the metallic dross collected in the sample using 40 g of Al and 20 g of Ca. Their chemical composition is detailed in supplementary Table S-VII. The main component was Pb, with Sb at 2.1 ±0.3 wt.%, As at 0.9 ±0.1 wt.%, Cu at 0.65 ±0.09 wt.%, Se at 0.4 ±0.2 wt.%, Te at 0.21 ±0.06 wt.%, and Ni at approximately 0.02% wt.%. The tin content in the dross from the third test was 1.1 ±0.1 wt.%, which was also lower than the value in the refined alloy. Analysis using SEM showed areas with high concentrations of both Al and Ca. The points where Ca was predominantly found contained elements such as Sb and Te, while As and Se were evenly distributed in both the Ca and Al areas. Cu appeared only in certain portions of the Ca-containing areas and in certain portions of the Al-containing areas. In the areas analyzed with SEM (Figs. 5,6), it was noted that the tin is evenly distributed in the lead, whereas significantly less tin was found in areas containing Al or Ca. This phenomenon was also observed in the third sample with Al and Ca (seen in supplementary Fig. S-3). This distribution of tin results from the fact that it does not form intermetallic compounds with aluminum. However, in the case of the addition of calcium, it binds earlier with other lead impurities and there was probably no thermodynamic possibility for tin to bind with calcium. Chemical analyses performed on the metallic dross formed in the experiment (supplementary Tables S-V–S-VII) indicated that the content of elements such as Sb, As, Cu, Se, Ni, and Te is Fig. 6. SEM images and the corresponding EDS element mappings of Pb, Sn, Ca, Sb, As, Cu, Se, and Te for a sample of metallic dross after using Ca as a refining additive. Malecha, S ´wiec, Albrecht, Jarosz, and Małecki3256 significantly higher than in the base material, while elements such as Bi and Ag, for instance, remain at the same levels as in the base material (Table Iand supplementary Table S-II). Reduction of Metallic Dross and Analysis of the Resulting Products Under industrial conditions, removing metallic dross formed during processing proves to be problematic and cost-ineffective. Therefore, this is reduced to using coke to extract metal lead from the dross. To simulate this process under laboratory conditions, metallic dross taken from a sample with Al as a refiner was placed into a laboratory furnace, heated to 600°C, and mixed with gradually added sawdust (20 g of sawdust was used for a total of 555 g of metallic dross). The process yielded 309 g of dross (seen in supplementary Fig. S-4a) with the composition detailed in supplementary Table SVIII, and 246 g of metallic lead, containing total impurities (Sb, As, Se, Cu, Ni, Te) of 0.59 wt.% (refer to supplementary Table S-IX for a detailed lead analysis). Phase analysis of the resulting dross is shown in Fig. 8. The obtained dusty dross contained compounds such as lead, antimony, and arsenic oxides, as well as sulfur compounds with copper and antimony. Similarly, the metallic drosses refined with Ca were reduced. These were heated to 600°C and mixed with gradually added sawdust (15 g of sawdust for 314 g of dross). This resulted in 49 g of dross (shown in supplementary Fig. S-4b) with the composition presented in supplementary Table S-X, and 265 g of metallic lead with the composition shown in supplementary Table S-XI. When reducing the metallic dross from refining with Ca, the rate of impurities return to the lead was significantly higher than that using Al, with total impurities of 1.67 wt.% (Sb, As, Se, Cu, Ni, Te). The resulting dross was given to phase analysis (XRD), Fig. 7. SEM images and the corresponding EDS element mappings of Al, Ca, Sb, Se, As, Cu, and Te for a sample of metallic dross after using Al-Ca mixture as a refining additive. Fig. 8. X-ray diffraction pattern (XRD) of the dross resulting from the reduction of metallic sludge using Al with sawdust. Refining of Secondary Pb with Retention of Sn Using Al and Ca Additions 3257