SEPARATION OF COPPER FROM A LEACHING SOLUTION OF PRINTED CIRCUIT BOARDS BY USING SOLVENT EXTRACTION WITH D 2 EHPA

Global generation of waste electrical and electronic equipment (WEEE) is increasing quickly. Metals from WEEE can be recovered by using unit operations of chemical engineering. This paper describes a combined hydrometallurgical route (sulfuric oxidant leaching + solvent extraction) to recover copper from printed circuit boards (PCBs). A non-magnetic fraction from comminuted PCBs was used to perform leaching tests at 75oC for 6 hours in an oxidizing media (sulfuric acid + hydrogen peroxide). In order to separate zinc, aluminum, and copper from the leaching liquor, solvent extraction tests were carried out using D2EHPA. Parameters that influence the process, such as pH, extractant concentration, and the aqueous/organic (A/O) ratio were investigated. Solvent extraction experiments were carried out in two stages: i) separation of zinc, aluminum, and residual iron, and ii) copper separation. The results showed that the leaching obtained around 60% aluminum, 94% copper, 76% zinc, 50% nickel and residual iron from the non-magnetic fraction of PCBs. With the solvent experiments, in the first stage, 100 wt.% zinc, iron and aluminum were extracted at pH 3.5, 2:1 A/O, 10 % (v/v) D2EHPA, while, in the second stage 100% of the copper was extracted at pH 3.5, 1:1 A/O, 20 % (v/v) D2EHPA.

Processes for metal extraction from WEEE based on mechanical, physical, pyrometallurgical and hydrometallurgical techniques have been developed (Chen et al., 2015;Fogarasi et al., 2013;Hagelüken, 2006).In pyrometallurgical processing, the matrix is incinerated, while target metals are smelted.Hydrometallurgical processing already consists of leaching, separation, purification and electrorefining of the metals of interest (Cui and Zhang, 2008;Tuncuk et al., 2012).Compared with pyrometallurgical processing, hydrometallurgical methods present a relatively low capital cost, and are also more easily controlled and more predictable and the recovery of metals is large for small scale applications.Pyrometallurgical processing operates at high temperatures (usually greater than 1000°C), consumes a lot of energy and produce hazardous gasses that must be treated with cleaning systems (Cui and Zhang, 2008;Rocchetti et al., 2013;Tuncuk et al., 2012).
In hydrometallurgical processing, metals from PCBs are solubilized by different methods acid leaching or bioleaching and the liquor can be purified by solvent extraction, selective precipitation and electrowinning (Ghosh et al., 2015;Tuncuk et al., 2012).Copper is the main metal present in PCBs and its leaching can be performed using sulfuric acid and hydrogen peroxide as an oxidant agent.Nickel, zinc and iron leaching can already be carried out using sulfuric acid.Several studies have evaluated the PCB leaching process and studied conditions, such as the solid-liquid ratio, acid concentration and temperature (Birloaga et al., 2013;Oh et al., 2003;Yang et al., 2011).Yang et al. (2011) analyzed copper leaching from spent PCBs.The study reported copper extraction higher than 90 wt.% using sulfuric acid, hydrogen peroxide and a solid: liquid ratio of 1:10(g/mL), after 3h at 23ºC.According to Birloaga et al. (2013) 90 wt.% of copper can be leached from PCBs after two leaching stages with sulfuric acid and hydrogen peroxide.Silvas et al. (2015) investigated copper leaching with sulfuric acid + hydrogen peroxide and 1:10 solid/liquid ratio at 75ºC over 4h.After two successive leaching stages, 100% of the copper in comminuted PCBs was leached.
The separation of metals contained in a liquor from acid leaching can be achieved by a solvent extraction technique.The extractant di-2-ethylhexyl phosphoric acid (D2EHPA) belongs to the organic phosphonic acid class and its main characteristic is the formation of a hydrogen bond between extractant molecules, causing the formation of dimeric structures (Figure 1) (Pereira, 2006).
The solvent extraction process has been widely used in hydrometallurgical metal recycling from WEEE (Kumari et al., 2016).Provazi et al., 2011) studied metal separation from a leaching liquor of spent batteries using Cyanex 272 versus selective precipitation.The results showed that solvent extraction was more efficient in metal separation than selective precipitation.Meanwhile, Oishi et al., 2007) investigated the recovery and purification of copper from printed circuit boards using LIX 26.Solvent extraction tests found that it is possible to achieve a metal extraction rate of lead, manganese, aluminum, iron, zinc and copper higher than 92% using LIX 26.
Several solvent extraction works have been developed to recover metals from WEEE (Dorella and Mansur, 2007;Kumari et al., 2016;Long Le et al., 2011;Nayl et al., 2015;Provazi et al., 2011).Table 1 shows some of the studies performed by different authors for metal recovery from WEEE using solvent extraction.
In this paper, a combined hydrometallurgical route was studied in order to recover and separate copper from discarded computer video PCBs using D2EHPA as extractant.

MATERIALS AND METHODS
Printed circuit boards used in this research were collected by a waste electronics recycling center in the city of São Paulo -Brazil.Initially the manual dismantling of the printed circuit boards was performed using pliers and a screwdriver for the removal of heat sinks, screws and electrolytic capacitors.With the aid of a manual guillotine, the boards were cut into smaller pieces so that they could fit into the opening of the mill.
The comminution was made primarily by grinding the PCBs in a blade mill with a 6 mm grill, followed by grinding in a hammer mill with a 2 mm grid.According to Tuncuk et al. (2012), the full release of copper happens with a particle size smaller than 2 mm as, in larger particles, there can be tunneling of Cu on the pins of the plastic components.
The magnetic separation of the sample was performed after grinding by the hammer mill.A dry drum magnetic separator was used at the following settings: magnetic roller speed of 27 rpm; vibration percentage of the feed 25%.In the present study, the material used was the non-magnetic fraction of PCBs.The sample was characterized by inductively coupled plasma optical emission spectrometry (ICP-OES) in equipment of the Agilent brand, model axial 710, in order to quantify the concentration of metals present in the liquors leached from acid digestion with aqua regia (1 HNO3: 3HCl; 24 hours; 25 ºC).The composition is shown in Table 2.

Acid leaching in an oxidizing medium
The non-magnetic fraction of PCBs was leached for 6h using 1M sulfuric acid, 1:10 solid/acid ratio (25g PCBs:250mL H 2 SO 4 ) at 75°C.An oxidizing media was induced by the addition of 10 mL of 30 % (v/v) hydrogen peroxide every 30min into the system (H 2 O 2 total added 120 mL) (Silvas et al., 2015).The liquor from the leaching process was used in solvent extraction tests.

Solvent extraction
Parameters like the pH effect, extractant concentration levels the aqueous/organic ratio (A/O) and extraction isotherms were evaluated.Temperature and time were maintained constant at 25°C and 10 min, respectively.

a) Experiments
The pH was monitored with a pH meter and specific glass electrode for solvents.Agitation was maintained constant during all the tests using a mechanical agitator.
Metal extraction (%E) was calculated from a mass balance (1) of each system studied.Equation 2 was used to calculate the extraction percentage (%E) for all A/O ratio experiments.
where C iorg and C forg are the initial and final concentrations of metal in the organic phase, respectively, while the initial and final metal concentrations in the aqueous phase are defined as C iaq and C faq .Furthermore, V aq and V org are the volumes of the aqueous and organic phases, respectively.

b) Extractant concentration and pH effect
Initially, the aqueous phase was added to the reactor and mechanical stirring was turned on.The pH was adjusted to the target value with the addition of 3M H 2 SO 4 and 5M NaOH.Then, the organic phase was mixed with the aqueous phase and the pH was maintained constant during the experiments.
Mixture separation was achieved in a separation funnel.The aqueous phase was collected and analyzed by the energy dispersive X-ray fluorescence spectrometry (EDX) technique.
In order to evaluate the extractant concentration effect, solvent extraction experiments with 10, 15 and 20% (v/v) of the extractant diluted with kerosene were performed.

c) Organic/Aqueous ratio (A/O)
To evaluate the A/O ratio, experiments using 1:1 and 2:1 A/O were realized at 25°C for 10min.A 10% (v/v) concentration of the extractant D2EHPA was used, diluted in 90% (v/v) kerosene, with pH modified to between 0.5-3.5.

d) Extraction isotherms
For determining the metal extraction isotherms, the aqueous phase was mixed with the organic phase at different A/O ratios.The extraction stage of the individual metal solutions was divided into two parts: i.Initially, zinc and aluminum were separated from the solution with copper and nickel.Tests were conducted at pH 3.5, using 10% (v/v) D2EHPA and 90% (v/v) kerosene.The A/O ratio was modified to between 1:5 and 5:1.ii.The new aqueous phase (solution with copper and nickel) was stirred with an organic phase comprising 20% (v/v) D2EHPA and 80% (v/v) kerosene.The A/O ratio varied between 1:5 and 5:1 at pH 3.5.The A/O ratio was varied between 1:5 and 5:1 to determine the number of countercurrent stages, known as McCabe-Thiele Method.

RESULTS AND DISCUSSION
After hydrometallurgical processing, it was possible to leach zinc, aluminum, copper, nickel and residual iron from video PCBs.Also, with the solvent extraction experiments variables such as the extractant concentration, pH effect, and A/O ratio were evaluated.

Acid leaching in an oxidative medium
In order to leach copper, other authors (Birloaga et al., 2013;Jha et al., 2011;Yang et al., 2011) also used sulfuric acid and hydrogen peroxide to leach metals from PCBs.The operational conditions used in the present study obtained around 60% aluminum, 94% copper, 76% zinc, 50% nickel and residual iron from the non-magnetic fraction, while the solution pH and ORP after the test were 0.5 and 673mV, respectively.Table 3 shows the liquor concentration obtained after the video PCB leaching process.After the leaching test, a solid fraction can be used to extract other metals or materials and then it has to be disposed of in landfills.In a study developed by Muniyandi et al. (2013) the non-metallic fraction was used as a filler material in recycled High-Density Polyethylene (HDPE), recycled high-density polyethylene (rHDPE) and in the production of rHDPE/ PCBs composites.The environmental regulation (balance in strength, stiffness, and toughness) for Separation of copper from a leaching solution of printed circuit boards by using solvent extraction with D2EHPA composite materials was attained with the incorporation of a 30 wt% non-metallic fraction PCBs and 6 phr MAPE (maleic anhydride modified linear low-density polyethylene) compatibilizer (Muniyandi et al., 2013).

Solvent extraction
Concentration of the synthetic aqueous feed solution is shown in Table 3.The initial pH of the synthetic solution was adjusted to 0.5 with additions of sulfuric acid.

e) Extractant concentration and the pH effect.
The extractant concentration effect has been investigated by several authors to analyze metal extraction efficiency from sulfate solutions (Lee et al., 2010;Mohapatra et al., 2007;Vahidi et al., 2009).In this paper, the influence of the D2EHPA concentration was examined in solvent extraction.
As shown in Figure 2, zinc, residual iron, copper and aluminum extractions increased upon increasing the extractant concentration.Also, metal pH isotherms were displaced to the left with the increase in extractant concentration.As can be seen in Figure 2 (a), the higher metal extractions were achieved when a mixture of 20% (v/v) D2EHPA and 80% (v/v) kerosene was used.This reflects the fact that the system with 20% (v/v) D2EHPA has more extractant molecules than systems with 10% (v/v) and 15% (v/v) D2EHPA.The metal extraction order found for this system was: Fe>Zn>Al>Cu>Ni, which is consistent with the results achieved by Sole and Hiskey (1992).
It is known that acid extractants release hydrogen ions during metal extraction and, when metallic cations are extracted, the extractant (D2EHPA) releases hydrogen ions.Metal extraction reactions, equilibrium constants (Keq) and distribution coefficients (D) for divalent and trivalent ions are described in Table 4 (Mansur et al., 2002;Mohapatra et al., 2007).
For divalent and trivalent metals, the distribution coefficient can be calculated by Equation 7. Therefore, considering that the extractant concentration is constant, it follows that:  The values of log K eq and n are related to the extraction of the metals.Thus, only the metals extracted were considered for the determination of both parameters.Then, the values of log K eq and n were calculated for aluminum, zinc and iron in the systems that used 10%v/v D2EHPA, 15% v/v D2EHPA, and 20%v/v D2EHPA (see Table 5).Since copper only Romano Espinosa

Brazilian Journal of Chemical Engineering
showed increasing extraction with 20% v/v D2EHPA, the equilibrium values were only calculated for this system (Table 5).
The results showed that pH can directly influence metal extraction capacity.An estimated number of hydrogen ions generated by each metal molecule extracted could be calculated with the experimental data of D and pH.Equation 4shows that metal extraction produces hydrogens ions; therefore, the pH decreases and the metal extraction percentage also decreases.
Figure 3 shows the pH variation effect in the distribution coefficient of aluminum, copper, iron and zinc.

f) Aqueous/organic ratio (A/O)
In solvent extraction tests with D2EHPA at an A/O ratio of 1:1 extractions of zinc, aluminum and copper higher than an A/O ratio of 2:1 (Figure 4) were achieved.However, aluminum and zinc were the metal cations most affected in the solution.For example, at pH 3.5 and an A/O ratio of 1:1, the extraction percentages for zinc and aluminum were 100% and 78%, respectively, while at pH 3.5 and an A/O ratio of 2:1 the E% decreased to 84% and 59%, respectively.

Ion Extraction Reaction Equations
Divalent metal (Adapted from Mansur et al., 2002) Trivalent metal (Adapted from Mohapatra et al., 2007) Table 5. Equilibrium value for different organic concentrations (contact time = 10 min, A/O ratio = 1:1 and temperature = 25 °C)  During the solvent extraction process, the fraction of extractant molecules available has to be enough to ensure the extraction of target metal ions (Ritcey and Ashbrook, 1984).When the A/O ratio was changed from 1:1 to 2:1, the volume of aqueous solution was two times higher, while the extractant capacity was still the same; therefore, the metal extraction was decreased.

Separation parameters
In solvent extraction, the separation factor is a measure of the selectivity of a solvent for one component over another one.A separation factor higher than one indicates that the separation is favored and can occur.
The separation factors of aluminum, zinc, and residual iron over copper are listed in Table 6.It was found that the selectivity of the solvent for Al, Fe, and Zn over copper increased with increasing pH.At the same time, it was observed that the separation factor of Fe and Zn increased with increasing extractant concentration as a result of the strongest affinity of D2EHPA for iron and nickel over the other metals in solution.
After evaluation of the separation factors, the variables chosen to carry out the separation of zinc, aluminum and residual iron from the aqueous phase were: pH of 3.5, an A/O ratio of 2:1 and 10% (v/v) D2EHPA diluted with kerosene.
The solvent selectivity for copper over nickel is given in Table 7.It was observed that the extraction of copper over nickel was improved with the increase in the pH and in the D2EHPA concentration.For copper extraction, the parameters adopted were: pH of 3.5, A/O ratio of 1:1 and 20% (v/v) D2EHPA.

McCabe-Thiele isotherms
To plot the extraction isotherms of metals, experiments were performed in two steps: i) the extraction of zinc, residual iron and aluminum, and ii) solvent extraction trials to separate copper from nickel.

g) Zinc, residual iron and aluminum extraction
The synthetic solution simulating the oxidant acid liquor (Table 2) was put in contact with a solution of 10% (v/v) D2EHPA diluted with kerosene and using A/O ratios between 1:5 and 5:1.To construct McCabe-Thiele diagrams for aluminum and zinc extraction (see Figure 5), the pH was maintained constant at 3.5.It was possible to verify that both zinc and aluminum can be extracted with two theoretical stages, although zinc, residual iron and aluminum cannot be separated from each other, as shown in Figure 2(a).Extraction of residual iron was achieved with one countercurrent stage.Therefore, in this case, a McCabe-Thiele diagram is not necessary.
Figure 5 (a) shows that two theoretical stages were needed to separate zinc from a synthetic solution using a 2:1 A/O ratio, and Figure 5 (b) shows that two theoretical stages were needed to extract aluminum from the aqueous phase using 2:1 A/O.
D2EHPA is commonly used to separate zinc from solutions that contain metals.Pereira et al., 2007) investigated the extraction of 11.9g/L of zinc from a solution with other metals like iron, magnesium, manganese and calcium.The work reported that zinc extraction using 20% (v/v) D2EHPA and an A/O ratio was possible with two theoretical stages.
Furthermore, D2EHPA can separate aluminum from solutions with other metals.Mohapatra et al. (2007) studied aluminum extraction from a solution by 0.3M D2EHPA.The investigation found that two theoretical stages were required to extract aluminum from the aqueous phase.

h) Copper extraction
To determine the number of theoretical stages to separate copper from a solution with nickel, 20% (v/v) D2EHPA diluted with 80% (v/v) kerosene was mixed with the aqueous phase at pH 3.5.As shown in Figure 6, at an A/O ratio of 1:1, with two theoretical stages, copper was completely separated from nickel.
Similar to a study developed by Pranolo et al., 2010), in this study, it was observed that, in the organic phase, the copper concentration increased with the increase in the A/O ratio.
While copper is extracted by the organic phase (20% (v/v) D2EHPA diluted with kerosene), nickel remained in the aqueous phase and can be recovered by an electrolysis process.The organic solution with copper must be forwarded to the stripping stage before copper recovery by the electrowinning technique.Total copper extraction from a liquor produced in the leaching process of the non-magnetic fraction of PCBs was carried out using the present process.

Table 1 .
Solvent extraction studies for WEEE treatment.

Table 2 .
Characterization of PCBs non-magnetic fraction

Table 6 .
Separation factors of aluminum, iron, and zinc over copper at different pH values, A/O ratios and at various extractant concentrations.

Table 7 .
Separation factors of copper over nickel at different pH values, A/O ratios and at various extractant concentrations.