Article pubs.acs.org/est
A Cleaner Process for Selective Recovery of Valuable Metals from Electronic Waste of Complex Mixtures of End-of-Life Electronic Products Zhi Sun,*,† Y. Xiao,§ J. Sietsma,† H. Agterhuis,‡ and Y. Yang† †
Department of Materials Science and Engineering, Delft University of Technology, 2628 CD Delft, The Netherlands Business Development, Van Gansewinkel Groep BV, 5657 DH Eindhoven, The Netherlands § Ironmaking Department, R&D, Tata Steel, 1970 CA IJmuiden, The Netherlands ‡
S Supporting Information *
ABSTRACT: In recent years, recovery of metals from electronic waste within the European Union has become increasingly important due to potential supply risk of strategic raw material and environmental concerns. Electronic waste, especially a mixture of end-of-life electronic products from a variety of sources, is of inherently high complexity in composition, phase, and physiochemical properties. In this research, a closed-loop hydrometallurgical process was developed to recover valuable metals, i.e., copper and precious metals, from an industrially processed information and communication technology waste. A two-stage leaching design of this process was adopted in order to selectively extract copper and enrich precious metals. It was found that the recovery efficiency and extraction selectivity of copper both reached more than 95% by using ammonia-based leaching solutions. A new electrodeposition process has been proven feasible with 90% current efficiency during copper recovery, and the copper purity can reach 99.8 wt %. The residue from the first-stage leaching was screened into coarse and fine fractions. The coarse fraction was returned to be releached for further copper recovery. The fine fraction was treated in the second-stage leaching using sulfuric acid to further concentrate precious metals, which could achieve a 100% increase in their concentrations in the residue with negligible loss into the leaching solution. By a combination of different leaching steps and proper physical separation of light materials, this process can achieve closed-loop recycling of the waste with significant efficiency.
1. INTRODUCTION The development of the electronic industry is associated with an increased demand on metals, especially more precious metals and more scarce metals, and leads to increased growth of waste electrical and electronic equipment (WEEE). WEEE is considered to be one of the fastest growing solid wastes.1−3 Currently, the generation rate is around 10 million tons annually in the European Union (EU) alone, and it has become a concern in every associated country.4 The WEEE directive (2012) in Europe5 therefore distinguishes the environmental and recovery significances of different WEEE, and the minimum recovery targets have been clearly defined. It is further promoting the development of WEEE treatment technologies, in particular valuable or strategic metals recovery from electronic wastes. Concerning WEEE treatment, numerous investigations have been performed, and abundant studies including critical reviews are available.2−4,6,7 An industrial process is commonly based on high-temperature metallurgy including burning of plastics, smelting of metal and oxides, refining and further electrochemical treatment after chemical extraction.1 Comparably, hydrometallurgical processes are also © XXXX American Chemical Society
available, especially for WEEE with low calorific values, and a range of investigations have also been carried out in the field of hydrometallurgy.8−12 The HydroWEEE project has built a prototype plant using hydrometallurgical techniques to extract valuable metals from certain WEEE, including fluorescent lamps, cathode ray tubes (CRTs), Li-ion accumulators, and printed circuit boards (PCBs).13 The process is based on acidic leaching (usually sulfuric acid) and selective precipitation of metals from the leached solution. Lime is added to detoxify the wastewater in the final step. However, the research performed so far, including the HydroWEEE project, focuses mostly on monostreams or relatively “clean” WEEE, for instance, printed circuit boards (PCB),14 batteries,12 and mobile phones after disassembly.15 The leaching selectivity and effectiveness of valuable metals are always the concerns in their processes. Treating a more complex type of WEEE, usually mixtures of Received: February 26, 2015 Revised: May 19, 2015 Accepted: June 10, 2015
A
DOI: 10.1021/acs.est.5b01023 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
from the fine fraction of residue I. Residue II was then obtained, with mainly silicate, ceramics, and enriched with precious metals. 2.2. Experimental Setup and Procedures. The asreceived ICT waste was directly used as the feedstock for metals extraction. In each laboratory-scale experiment, the sample size was taken as ∼50 g which is a representative sample size of the bulk composition according to our previous investigations. The leaching experiments were carried out in a glass reactor (500 mL volume), and the leaching solution is usually 250 mL (liquid to solid ratio of 5 mL/1g). A roundshaped reactor was used to ensure better mixing conditions. Air was purged into the bottom of the reactor, and the gas bubbles were distributed by the agitator. The gas flow rate was controlled by a flow meter. During leaching, the system temperature fluctuates slightly because of the exothermic reactions.21 A heating bath with water was used to control the temperature, and an additional thermometer was used to monitor the temperature change during leaching (temperature control is from room temperature to 100 °C). The temperature was found to stay within ±2.5 °C of the set temperature. The pH values were detected by a pH meter throughout the leaching process (for ammonia-based solutions, the starting pH is around 10.6−10.8). Prior to the leaching, the ammonium carbonate was first dissolved into the ammonia solution, and demineralized water was added to reach a liquid volume of the required liquid to solid ratio. The solution was kept in the reactor for a period of time to reach the required temperature before the ICT waste was added. Liquid samples were taken at certain time intervals to track the leaching behavior of different metals in the ICT waste. The sample size was 1−2 mL, which is considered to have an insignificant effect on the liquid to solid ratio. Since the waste exhibits high heterogeneity, the residue after leaching was also analyzed to diminish any discrepancy in composition. The liquid solution was analyzed with inductively coupled plasma−optical emission spectrometry (ICP-OES, PerkinElmer Optima 3000DV), and the residue was analyzed by an X-ray fluorescence spectrometer (XRF, PANalytical Axios). The extracted fraction/recovery of different metals is therefore calculated from
various kinds of end-of-life products or waste from different streams, is still, however, a great challenge. Additionally, such types of WEEE is becoming one of the main WEEE feedstocks.5,16,17 In previous research, the metals can be extracted selectively from a highly complex industrial information and communication technology (ICT) waste by using a hydrometallurgical method.18 Copper in the ICT waste could be selectively extracted by using ammonia−ammonium carbonate solution. In the downstream to recovery of copper from the leached solution, direct copper electrowinning from the ammonia-based leaching solutions usually ends with low current efficiency.19 Extensive research has been carried out to control the concentration of Cu+ in order to develop an energy-saving copper electrowinning process.20 However, it is difficult in practice to keep a high concentration of Cu+ in the solution since Cu+ is easily oxidized into Cu2+ even at low oxygen partial pressure. Under both conditions, solvent extraction is frequently required in order to minimize the effect of impurities during copper recovery. With this research, we demonstrate a closed-loop process for valuable metals extraction, i.e., copper and precious metals, from a highly complex industrial information and communication technology (ICT) waste. Copper is selectively extracted by an ammonia-based solution. With an in-house designed reactor, two residue streams are obtained−coarse fraction and fine fraction. The fine fraction is further processed for acid or alkaline leaching to concentrate precious metals, while the coarse fraction is returned to the ammonia leaching. Finally, a solution with copper and a solid concentrate of minimized volume with precious metals is obtained. Electrowinning of copper is optimized in order to minimize the energy consumption, which has been a challenge for ammonia-based solutions.19 After electrowinning, ammoniabased solution is regenerated to be reused as the leaching reagent of the ICT waste. The precious metal concentrate is treated as an intermediate product that will be further processed using traditional metallurgical technologies. A schematic illustration of different steps in the process is given in Figure A1 (Supporting Information). The main objective of the current research is to recover valuable or strategic metals from a complex WEEE by integrating fundamental hydrometallurgy principles with the aim of minimized environmental effects, less energy consumption, and high extraction efficiency.
XM = mt /(mf + mr ) × 100%
(1)
where mt and mf are the metal mass in the leaching solution at time t and the final solution, respectively; mr is in the metal mass in the final residue after leaching. After ammonia leaching, the residue (residue I) was sieved into two fractions: fine fraction (smaller than 1 mm) and coarse fraction (larger than 1 mm). The fine fraction of residue I was leached by an acidic solution or caustic solution to further concentrate precious metals. Then vacuum filter was used to separate the leached solution and the second residue (residue II). The solution was returned to leach the fresh fine fraction of residue I. When the salt concentrations of, e.g., Fe3+, Al3+, and Si4+, are high enough, the solution will be purified by precipitation. Acid or base is added from time to time to compensate the consumption during leaching. Residue II is physically processed to obtain a precious metal concentrate. In the second-stage leaching, the same reactor as ammonia leaching was used, while a magnetic stirrer was introduced instead of mechanical stirring. After leaching, due to the relatively high concentration of copper in the ammonia leaching solution (about 20−70 g/L), copper electrowinning directly from the solution without solvent extraction was used for
2. EXPERIMENTAL SECTION 2.1. Materials. The ICT waste (smaller than 8 mm) was provided by Van Gansewinkel Groep (VGG). The material was pretreated and concentrated physically and shredded into sizes smaller than 8 mm. This size is defined to be a limit since further shredding will significantly increase the processing cost. The material was of high heterogeneity from observation and contained a large amount of sand, stones, glass, ceramics, and plastic particles. Metallic materials are either trapped in nonmetallic components or present as metal wires or scrap. The general morphology is given in Figure A2 (Supporting Information). Ammonia (NH3, 25% in water, Alfa Aesar) and ammonium carbonate ((NH4)2CO3, 99.0%, analytical grade, Alfa Aesar) are used during selective leaching of copper, and the ICT waste becomes residue I (see Figure A1). Hydrochloric acid (HCl, 37%, analytical grade, Alfa Aesar), sulfuric acid (H2SO4, 98%, analytical grade, Alfa Aesar), and hydrogen peroxide (H2O2, 30%, analytical grade, Alfa Aesar) are used in the second-stage leaching of less valuable base metals extraction B
DOI: 10.1021/acs.est.5b01023 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 1. Selectivity of leached metals into different leaching solutions.
recycling,13,23 the current process aims to maximize the extraction efficiency and minimize the environmental impact. 3.1. Selective Recovery of Copper. In order to selectively extract copper from the waste, a leaching reagent ideally needs to only react with copper, while other metals are inert to the leaching solution. However, it is difficult to find such reagents available on the market. Practically, it is possible to find a leaching reagent that is only effectively reacting with copper and other metals extracted into the solution and has minor effect on further copper recovery from the solution. In this research, three types of leaching solutions were selected, based on our preliminary investigations: (i) ammonium (7.55 wt %) with 196 g/L ammonia carbonate, air flow rate of 70 L/h at room temperature for 4 h; the solid to liquid ratio is 1g/5 mL; (ii) sulfuric acid (25 wt %) with 1.2 times the stoichiometric amount of H2O2 at 80 °C for 4 h; the solid to liquid ratio is 1g/5 mL; (iii) salt−aluminum chloride (25 wt %) with 1.2 times the stoichiometric amount of H2O2 at 80 °C for 4 h; the solid to liquid ratio is 1g/5 mL. A parameter “selectivity” (SM), that is the concentration of the metal (M) of interest divided by the sum of all leached metals in the solution, is defined to evaluate the effectiveness of different leaching solutions as given by
copper extraction. The solution was charged into a single compartment of an electrolysis cell, with a width 10 mm × height 10 mm copper cathode and a graphite anode of 40 mm × 10 mm. Ag/AgCl electrode was used as the reference electrode (see Figure A7 for the setup). Cylindrical electrodes with the same materials (by keeping the same area ratio) were also used in order to improve the current efficiency. At room temperature, a constant DC current was supplied with a Parstat 4000 electrochemical interface (Princeton), and the process was computer controlled with automatic recording of the current and cell voltage. After the electrowinning cycle, the copper deposit on the cathode was rinsed with demineralized water, naturally dried, and measured for the cathode weight change. The deposit was observed with an optical microscope for product morphology, and the copper particles deposited on the cathode were removed from the cathode by scratching for X-ray diffraction (XRD) and XRF analyses. The solution of the electrolyte was sampled and analyzed with ICP before and after electrowinning. 2.3. Characterization. XRF was used for compositional analyses of the bulk composition and the composition of the residue after leaching. The sample was prepared according to the procedures in ref 22, and the powder was pressed into pellets for XRF analyses. Dissolution of solid residue using aqua regia was also carried out and analyzed using ICP-OES, and the remaining solid was analyzed using XRF.21 The morphology of the residue was characterized with both digital camera and scanning electron microscope (SEM, JEOL JSM 6500F) with energy-dispersive spectroscopy (EDS). The sample was carbon coated when necessary. All liquid samples were diluted and prepared for analyses with ICP-OES.
SM = CM /∑ C i × 100% i
(2)
where CM is the concentration of metal which is of interest in the leaching solution and Ci is the concentration of any metal which has been leached into the solution. As shown in Figure 1, the leaching selectivity of the ICT waste using three leaching solutions shows very different characteristics. The ammonia solution has the highest copper selectivity of 95.5%, while it becomes 26.6% in sulfuric acid solution and only 0.2% in the salt solution. Both ammonia solution and salt solution have zero selectivity for silver. Therefore, it is possible to selectively leach copper prior to other metals or selectively leach other metals than copper. However, recovery of the metals leached into the salt solution is difficult due to many metals being coleached into the solution. In this process, ammonia solution is adopted to selectively recover copper from the waste. In industrial practice, it is suggested that copper concentration and the ionic state of copper in the solution are very
3. RESULTS AND DISCUSSION According to our previous research,21 the ICT waste is very complex and contains a range of metals, including copper, zinc, nickel, iron, aluminum, lead, tin, and precious metals (silver and gold), with significant amounts of sand, glass, ceramics, and plastics/polymers (see Figure A2 for a typical image). The typical composition of metals in the ICT waste is given in Table A1. It can be found that around half of the waste is copper, mostly in the presence of wire, coating layer, and plate. The waste contains significant amounts of precious metals, silver at a concentration of 640 ppm and gold at 120 ppm. Considering the values of different metals, the extraction of copper and precious metals is important in the recycling of this waste. On the basis of environmental policies and the philosophy of C
DOI: 10.1021/acs.est.5b01023 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 2. Copper concentration and pH evolution with leaching time in ammonia leaching solution.
precious metals need to be finally recovered or enriched, a second-stage leaching to remove other metals is necessary. To make an effective process design, the distribution of metals, especially precious metals (Ag and Au), must be evaluated. Table A2 gives the typical metal content and distribution in residue I. It is found that the concentration of precious metals increases due to the mass reduction. Figure 3 shows the
important to evaluate the feasibility of the following step for copper recovery, e.g., for electrowinning the copper concentration to be higher than 10−20 g/L.21,24−26 Understanding the copper reaction mechanisms in the solution is essential to controling the copper leaching behavior. Figure 2 shows the copper concentration and pH evolution with leaching time during ammonia leaching. In the current leaching conditions, it is possible to obtain a solution with more than 70 g/L copper, which is high enough for electrowinning. Copper leaching is a gas−liquid−solid multiphase process. It involves oxidation of copper with oxygen and amine complex formation. In the literature, it is suggested that copper leaching follows the following procedures:24 2Cu + O2 → 2CuO
(3)
CuO + NH3· H 2O → Cu(NH3)12 + + 2OH−
(4)
Cu(NH3)12 + + 2NH3· H 2O → Cu(NH3)24 + + 2H 2O
(5)
It was declared that copper is first oxidized into CuO by the dissolved oxygen in the solution before it forms the amine complex24 near the metal surface. However, according to the Eh−pH diagram (Figure A3), in the pH range of 9−11 it crosses a region with Cu+ before copper is converted into amine complex (Cu(NH3)42+). In order to confirm the mechanisms of copper conversion, intermediate solid samples during leaching were taken for analyses. As shown in Figure A4, the morphology of the surface of copper wire before and after ammonia leaching is compared. The composition was analyzed using energy-dispersed spectroscopy in the SEM. It can be found that the surface of a copper wire is covered partially by a layer of CuO due to the oxidation in air. After leaching, a layer of Cu2O is found, and this layer should be formed during Cu oxidation by dissolved oxygen in the ammonia solution during leaching. The thickness of the Cu2O layer is found to be around 10 μm. With an enlarged view of the copper area, it is noticed that the surface is doped by Cu2O particles which are surrounded by copper metal. Compared with the oxygen line in Figure A3, the oxygen partial pressure or oxygen activity near to the metal surface should be much lower than unity with the formation of Cu2O. It means that the dissolution of oxygen plays an important role in copper leaching into the ammonia based solution. 3.2. Extraction of Other Base Metals. After ammonia leaching with selective copper extraction, a residue containing other materials from the ICT waste is obtained (residue I). As
Figure 3. Distribution ratio of different metals in the original ICT waste and residue I (the original ICT waste was 49.98 g and residue I is 30.12 g).
comparison of the metal distribution ratios at different particle sizes in the original ICT waste and in residue I. The distribution ratio was calculated on the basis of the content of each metal in the materials of different sizes. It is clear that the distribution ratio of precious metals increases in the fine fraction (smaller than 1 mm) of the residue after ammonia leaching, while aluminum, nickel, and chromium distribute more to coarse fraction (larger than 1 mm). From observation, it is noticed that some thick copper wires were still not fully dissolved, and the total weight of the coarse fraction occupies more than half of the weight of residue I. In the second-stage leaching to further enrich precious metals, the coarse fraction and fine fraction of residue I were separated and treated separately. Residue I particles larger than 1 mm are returned to ammonia leaching, while residue I particles smaller than 1 mm are treated in the second-stage leaching to enrich precious metals. On the basis of preliminary investigations, the second-stage leaching was performed using sulfuric acid with residue I (smaller than 1 mm). The leaching is to maximize the D
DOI: 10.1021/acs.est.5b01023 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 4. Extraction efficiency of different metals in the second-stage leaching by sulfuric acid (25 wt %) at 80 °C without oxidant for 6 h (residue II is 19.58 g).
influence the precious metal concentrations in residue II. Figure A6 shows the effect of different sulfuric acid concentration and addition of oxidant (H2O2) on the silver concentration in residue II. In the cases of no oxidant, i.e., H2O2, the silver concentration in the residue increases considerably with the concentration of acid (wt %). This is mainly because of the enhanced leaching of other metals than precious metals to the solution and the total dissolution ratio of residue I (smaller than 1 mm) is found to be increased from around ∼14% in acid concentration of 17.6 wt % to be ∼18% in acid concentration of 38.6%. In a less concentrated sulfuric acid, silver metal is rather inert when oxidant is not present. The possibility of leaching silver increases in a sulfuric acid solution with higher acid concentration at elevated temperature, e.g., 80 °C, and sulfuric acid may react with silver as covalent molecules according to
extraction of less valuable metals, e.g., Fe, Al, while minimizing the loss of precious metals into the solution. A typical composition of the leached solution is given in Figure 4. It can be found that the extraction efficiency is high for Fe, Ni, Cr, Al, and remaining Cu. Sn and Pb however have low extraction rates indicating that they remain in the residue II together with precious metals. The total metal content is around 19 g/L which is already significant for metals recovery in traditional waste acid treatment which can be processed industrially.27 After treatment, the sulfuric acid is recycled to be used again in the second-stage leaching. However, additional acid is always required to compensate its consumption during leaching of residue I. The loss of precious metals to the leaching solution is around 3 wt % (2.58 for Ag and 3.17 for Au), as given in Figure 4, which may accumulate when the leached solution is recycled. When the concentrations of precious metals reach a level of profit, they will be separated following traditional wastewater treatment procedures.27,28 Figure A5 shows the SEM image with compositional analysis with EDS. It is clear that a large fraction of silica-containing phases exist and metals are present as alloys with or without partial oxidation. Silver is found as alloy with copper and tin, which represents the content of lead free solder, while gold-rich phase could not be found with SEM analyses. Very frequently, Au is used as coating layer above copper in electronic components as gold-plated copper in order to improve their performance especially at higher temperature.29 During firststage ammonia leaching and second-stage acidic leaching, copper is leached into the solution while gold remains in the residue. However, the gold layer is normally very thin, e.g., less than 1 μm. This is most probably the reason that gold-rich phase is difficult to be observed under SEM. To have a better understanding of the leaching behavior of different metals, the metal content of Residue II is analyzed and given in Table A3. Copper has been reduced from originally ∼49 wt % in the ICT waste to below 1 wt % in residue II. Both precious metals are significantly enriched, i.e. silver from originally 600 ppm in the ICT waste to 1800 ppm and gold from 120 to 240 ppm in residue II. In the second-stage leaching, the enrichment of precious metals is highly dependent on the leaching conditions, e.g., acid concentration, temperature, reaction time, and oxidant (H2O2). According to current research, it is found that during second-stage leaching, acid concentration and oxidant (H2O2) in the leaching solution are two important factors that will
2Ag + 2H 2SO4 → Ag 2SO4 + 2H 2O + SO2
(6)
This redox reaction is more easily happening in concentrated sulfuric acid (e.g., 96%). When silver is already oxidized or oxidant is present during leaching. Silver is reacting with sulfuric acid flowing alternative procedures as 2Ag + H 2O2 → Ag 2O + H 2O
(7)
Ag 2O + H 2SO4 → Ag 2SO4 + H 2O
(8)
Since the solubility product Ksp for silver sulfate is 1.2 × 10−5, which is relatively high, and its solubility is 8.3 g/L in water at 25 °C.30 When silver sulfate is formed, silver can be significantly leached by the solution resulting in low silver concentration in residue II. According to the morphology and compositional analyses, it can be found that a large fraction (∼50 wt %) of the residue II is silicate-containing particles. In addition, around 10 ± 4 wt % of plastics/light materials is present. Since the densities of silicate and plastics are much lower than the alloys with heavy metals, it is suggested that physical separation can be introduced to further enrich precious metals. The precious metal concentrate is considered as one of the products in current process. 3.3. Flow of Precious Metals. In order to effectively recover precious metals from the ICT waste, it is important to understand how the precious metals flow into different material streams, which is a key decisive factor in evaluating the applicability of the process. However, since the ICT waste is of E
DOI: 10.1021/acs.est.5b01023 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 5. Concentration and flow of precious metals (silver and gold) during different leaching stages (leaching I, first-stage ammonia leaching of ICT waste; leaching II, second-stage leaching of residue I; the percentage indicates the content of precious metals in a residue comparing with their original content in the ICT waste; screening is a step to separate residue I into a coarse fraction and a fine fraction, and the fine fraction is treated in leaching II).
sulfate, without a step of solvent extraction. As shown in Table A4, the current efficiency of both solutions is quite low at ∼60%. In order to improve the current efficiency, a new electrodeposition configuration was used, changing from plate electrodes into cylindrical electrodes, while the area ratio of anode to cathode was kept the same. As shown in Figure A7, it basically changes the electrical field configuration from nonsymmetric into symmetric. An obvious improvement in the current efficiency could be reached. Table 1 shows that at
large heterogeneity, it is difficult to have accurate quantitative analyses on its composition as well as the composition of the following residue after leaching. In order to minimize the discrepancies on compositional analyses, the ICT waste was analyzed according to the well-evaluated procedures,22 and five samples (each with 10 g) taken from the bulk were treated. A baseline of the composition is therefore determined where the precious metal content is set as 100%. In a typical leaching experiment, the sample size was taken as 50 g; after ammonia leaching with conditions in section 3.1 and acidic leaching with conditions in section 3.2, the residues at different stages were analyzed to understand the flow of precious metals. As given in Figure 5, both silver and gold can be concentrated 2−3 times compared with the original ICT waste in one leaching without considering recycling of the coarse fraction of residue I (larger than 1 mm). After leaching I, 96% silver and 88% gold are concentrated into residue I. Screening of residue I results in a fine fraction (1 mm). It is shown that 70% silver and 40% gold are concentrated into the fine fraction (residue I
