Selective Recovery of Copper from the Mixed Metals Leach Liquor of E-Waste Materials by Ion-Exchange: Batch and Column Study

: Recovery of metals from e-waste forms a major focus of circular economy thinking and aligns well with the Sustainable Development Goals (SDG). While hydrometallurgical extraction from electronic printed circuit boards (PCBs) is well established, the separation of metals from the leach liquors, which are complex mixtures, remains a challenge. To achieve selective separation, ion exchange resins with chelating functional groups were employed in the present study. Batch and column studies for selective recovery of Cu 2+ from a given mixed metals leach solution were conducted using Dowex M4195 resin, and both the adsorption isotherm and kinetics were studied. The process involves three major steps: selective recovery of Cu 2+ by M4195 at low pH and elution with H 2 SO 4 ; sorption of Ni 2+ from the rafﬁnate by Dowex M4195 at pH 2 and removal of Fe 3+ from rafﬁnate. The batch experimental results showed appreciable and selective recovery of copper (51.1%) at pH 0.7 and 40.0% Ni 2+ was sorbed from rafﬁnate at pH 2.0 with co-adsorption of Fe 3+ as impurity. The batch adsorption data could be ﬁtted with both Langmuir and Freundlich isotherms and exhibited pseudo-second-order kinetics. Column studies agreed with the Yoon–Nelson model and indicated that Cu 2+ break-through time in the column decreased with an increase in ﬂowrate from 3.0 to 10.0 min/mL and decreased in sorption capacity, while it was delayed with increased bed heights from 20 to 30 mm. Complete elution of Ni 2+ was obtained with 2.0 M H 2 SO 4 after selective elution of trace impurities with dilute HCl. Iron in the rafﬁnate was removed via the addition of Ca (OH) 2 at pH 4.0 leaving Zn-Al in the solution.


Introduction
Globally, waste electrical and electronic equipment (WEEE) remains the fastest growing and most complex solid waste stream.Its annual production was 52.2 M metric tons in 2019 and is further projected to reach over 74.7 M metric tons by 2030 [1,2].
WEEE contains a significant number of valuable metals which are used for electrical contacts and conduits owing to their excellent chemical stability coupled with good conducting properties [3].Disposal of these components as waste represents both a loss of value and a potential hazard to the environment through the long-term leaching of heavy metals.Both these concerns have encouraged the recycling of e-waste to recover the metal values and safely dispose of the residues.There has been much effort towards the development of a viable approach to recycling e-waste via pyro-and hydrometallurgy routes.Acid leaching of metals from primary and secondary materials produces a liquor containing mixed metals including iron, which presents a significant problem because it has low value and at the same time interferes with the recovery of the value metals; it has been identified to be particularly problematic in electro-winning or electrochemical deposition of metals [4][5][6].Traditionally, precipitation by increasing pH to remove heavy metals is a common practice as a purification process [7][8][9][10].However, this approach is known to generate sludges that are increasingly difficult to dispose of because of potential contamination of the environment by the heavy metals contained therein, as well as significant reagent costs associated with the necessary pH control [11].Since there is always a loss of metal values when removing iron by precipitation [10,12], it is highly desirable to selectively recover the metal of interest prior to precipitation.Although solvent extraction (SX) is a common method used to purify and enrich copper from WEEE acidic leach liquors, iron would have been removed by precipitation before SX and would hence result in a loss of value metal [13][14][15][16][17][18][19].Solvent (co-)extraction of iron mostly proceeds in the pH range of 2.5-3.5 [14,17,18], and oxalic acid, which is conventionally used in its stripping, is difficult to regenerate making the process expensive.More so, numerous setbacks such as solvent flammability, crud formation and poor phase separation are associated with solvent extraction [20].Scale-up of solvent extraction requires well-designed mixer-settlers which are cost-intensive and need a skilled operation to mitigate the risk of solvent flammability and third-phase formation resulting in poor separation.Given these drawbacks, ion exchange resins, especially those with chelating properties, seem to have better loading capacity, are easy to operate and inexpensive.These resins possess specific functional groups that enable highly selective metal uptake.For example, resin functionalized with end groups such as bispicolylamine, amidoxime, 8-hydroxychinoline, dithiocarbamate, diphosphonic, iminodiacetate, aminophosphonic, sulphonic, thiol, thiourea and carboxyl resins, are commercially available [21,22].It is noteworthy that bis(2-pirydylmethyl) amine, methylglucoamine, and thiol resins are found to be most suitable for the recovery of base metal ions (such as Cu, Ni, Fe, Pb, etc.) from acidic conditions [3].Iminodiacetic and picolylamine resin have been reported to exhibit good performance for the recovery of copper and nickel in chloride and sulfate media, with the former showing preference for nickel over copper [23][24][25][26].However, optimization of parameters for selective recovery of metals from polymetallic liquor onto resins is not extensively available in the literature.
For our study, Dowex M4195, an amine-based chelating resin, was chosen due to reported advantages such as high selectivity towards copper in extremely acidic conditions [27,28].The targeted metal recovery was from sulphuric acid media, as this was the preferred method for primary WEEE leaching, owing to the low solubility of solder material (Pb and Sn) [29,30], which could interfere with uptake of metal of interest [11,31].This research focused on the selective recovery of Cu 2+ ions and removal of impurities from mixed metal sulfuric acid leach liquor of waste electronic materials using Dowex M4195 resins via fixed bed column and builds on a previous study [32].

Sorbate and Chelating Resin Characteristics
Numerous articles have been published on the sorption of Cu 2+ from simplified and often pure solutions, whereas in practice the leach liquor would always contain multiple metals that can co-extract.Therefore, for this study, the leach liquor obtained from our previous leaching of metals from waste electrical components was used [33], with the typical chemical composition presented in Table 1.The initial pH of the liquor was 0.33 (2.0 M H 2 SO 4 ).Given the limited availability of resin for this study, the leach liquor was diluted to the desired initial concentrations prior to the sorption experiments, and pH adjustment was executed using dilute H 2 SO 4 and NaOH solutions.The reagents used (H 2 SO 4 , NaOH, Minerals 2023, 13, 1285 3 of 18 H 2 O 2 , and Ca (OH) 2 ) were all analytical-grade reagents.The selection of resin was based on the metal compositions of the leach solution, as well as its availability, cost, and physicochemical properties.Dowex M4195 is a chelating resin with a multidentate amine ligand and microporous structure.It is a polystyrene-divinylbenzene opaque bead appearing tan to dark brown or dark green [34].According to the product information given by the manufacturer, the resin functionality is based on a special chelating amine ligand which is partially quaternized by sulfuric acid.In this conjugate sulfuric acid salt form, the resin is fully swollen and hydrated, and ready for scavenging metals from acidic media even in the presence of other chelating agents such as EDTA.It is recommended for the removal of transition metals at pH < 2.0.Detailed molecular-level characterization of Dowex M4195 in the context of Cu removal has been given in a recent publication [28].

Experimental Procedure
The experimental program included the following major steps:

•
Batch experiments for copper adsorption and elution using Dowex M4195 with determination of isotherms.

•
Fixed bed adsorption and desorption studies of Cu 2+ ions.

•
Treatment of raffinate for the removal of Ni 2+ using spent Dowex M4195 and removal of Fe 3+ from solution.

Batch Experiment
Both sorption and desorption of metals onto or from Dowex M4195 resin was conducted using an orbital shaker at 200 rpm.In an Erlenmeyer flask, 200 mg of preconditioned (washed with 1.0 M H 2 SO 4 ) Dowex M4195 resin was contacted with 50 mL of sulfate leach solution.The p H of mixed metals sulfate leach solution, equilibration time and mass ratio of resin to leach solution are the parameters used to evaluate the batch recovery efficiency.After sorption, the metal content in the solution was determined by Atomic Absorption Spectrometry (AAS) using a Perkin Elmer Analyst 200 (Perkin Elmer, Rodgau, Germany).The recovery efficiency of metal from leach solution (%R) was calculated by Equation (1), and the percentage of metal stripped (%S) was determined by Equation (2).
where K 0 and K t represent the concentration of metals at times 0 and t, respectively (mg/L).
where Q s is the amount of metal stripped from the loaded resin, and Q a is the amount of metal adsorbed onto resin before stripping (mg/g).

Fixed Bed Sorption Study
After the determination of optimum pH for adsorption, selective sorption of Cu 2+ was investigated via fixed bed column experiments.The column was made from a Perspex tube with dimensions of 15 mm diameter and 205 mm length.The sampling point was located at the bottom tip of the column.The column was packed with preconditioned Dowex M4195 resin using glass wool as a supporting layer above and below the resin bed.The mixed metals pregnant leach solution at optimized pH was pumped through the column with a peristaltic pump from top to bottom.The effluent was collected at regular time intervals using a calibrated plastic sampler, and the metal ion concentration in the solution was determined by AAS.The column performance towards selective sorption of copper onto Dowex M4195 resin was investigated as a function of flow rate (3.0, 5.0, and 10.0 mL/min) and bed depth (20, 25, and 30 mm).After sorption, the column was eluted with 2.0 M H 2 SO 4 solution.
Fixed bed column parameters for the sorption of copper by Dowex M4195 resin were evaluated in terms of breakthrough time and total sorption (q t ) of the resin.The loading behavior of metals unto resin in the column was expressed in terms of concentration C t /C 0 as a function of time (t), where C 0 and C t are the concentrations of inlet mixed metals solution and outlet concentration at time t, respectively and breakthrough curves were plotted, column exhaustion time and total effluent volume (V e , mL) was calculated from the following equation: where Q (mL/min) and t total (min) are the volumetric flow rate and the time of exhaustion, respectively.With a known solution concentration and flow rate, the area of the breakthrough curve was determined via integration of the adsorbed concentration (C 0 -C t ) versus t (min) curve.The sorption parameters (Equations ( 4)-( 6)) were estimated according to [35].The total sorption q t (mg) was obtained from Equation (4): The equilibrium maximum copper sorption capacity (q e , mg/g) of the column was calculated using Equation (5), where m (g) is the weight of resin beads in the column.
The total amount of metal ions (m t ) entering column (mg) was calculated according to Equation ( 6): The Cu 2+ sorption efficiency (R %) at resin saturation was calculated based on Equation (7):

Column Sorption Models Thomas Model
The Thomas model is applied to experimental data for the evaluation of breakthrough fitting, regardless of sorption equilibrium, and it is based on a Langmuir isotherm.The linearized form of the model is expressed as per Equation (8).
where k Th is the Thomas kinetic coefficient (mL/min/mg), t is the total flow time (min), and Q is the volumetric flow rate (mL/min).The adsorption capacity and mass of the adsorbent are denoted as q 0 (mg/g) and m (g).A plot of ln[(C 0 /C t ) − 1] versus t gives the value of k Th and q 0 .

Yoon-Nelson Model
The Yoon-Nelson model assumes that the rate of sorption of Cu 2+ ions is inversely proportional to the extent of Cu 2+ already sorbed and breakthrough on the resin.It is expressed as Equation (9).ln where τ is the time required for 50% adsorbate breakthrough (min), t is sampling time (min), and k YN is the rate constant (1/min).

Sorption of Nickel and Removal of Impurities
The raffinate obtained after copper elution was completely recovered and adjusted to pH 2.0 and the spent resin was packed and the raffinate solution was pumped through the column.Iron left in the solution was removed by precipitation with Ca(OH) 2 prior to the discharge of the solution.

Batch Sorption Study
The pH plays a significant role in the affinity of a resin for a particular ion.Hence, the effect of pH on the recovery of metals from the multi-metallic pregnant leach solution was studied to determine optimal conditions.As shown in Figure 1, Dowex M4195 resin performed excellently for the recovery of Cu 2+ in the highly acidic range.Recovery of Cu 2+ becomes less selective at pH > 0.7, at which point the recovery efficiency of other metals becomes more prominent.This is plausible because only one nitrogen atom of the bis-(2-pyridylmethyl) amine functional group is deprotonated and has a strong affinity for Cu 2+ at highly acidic conditions.On the other hand, in the mid to high pH range, two or all the nitrogen atoms of bis-(2-pyridylmethyl) amine functional groups are deprotonated and thus show increased affinity to other metals [34].

Sorption of Nickel and Removal of Impurities
The raffinate obtained after copper elution was completely recovered and adjusted to pH 2.0 and the spent resin was packed and the raffinate solution was pumped through the column.Iron left in the solution was removed by precipitation with Ca(OH)2 prior to the discharge of the solution.

Batch Sorption Study
The pH plays a significant role in the affinity of a resin for a particular ion.Hence, the effect of pH on the recovery of metals from the multi-metallic pregnant leach solution was studied to determine optimal conditions.As shown in Figure 1, Dowex M4195 resin performed excellently for the recovery of Cu 2+ in the highly acidic range.Recovery of Cu 2+ becomes less selective at pH > 0.7, at which point the recovery efficiency of other metals becomes more prominent.This is plausible because only one nitrogen atom of the bis-(2pyridylmethyl) amine functional group is deprotonated and has a strong affinity for Cu 2+ at highly acidic conditions.On the other hand, in the mid to high pH range, two or all the nitrogen atoms of bis-(2-pyridylmethyl) amine functional groups are deprotonated and thus show increased affinity to other metals [34].Therefore, pH 0.7 was found to be optimum for the selective sorption of copper.Our results were in consonance with other sources [11,36], who reported that Dowex M4195 still retains its affinity for metals in highly acidic medium.Subsequently, equilibrium recovery time was studied at the optimal pH and the result shows that the selective recovery efficiency of Cu 2+ steadily increased when the time increased from 5 to 120 min and reached equilibrium after 180 min without sorption of other metals (Figure 2).A 51.1% Therefore, pH 0.7 was found to be optimum for the selective sorption of copper.Our results were in consonance with other sources [11,36], who reported that Dowex M4195 still retains its affinity for metals in highly acidic medium.Subsequently, equilibrium recovery time was studied at the optimal pH and the result shows that the selective recovery efficiency of Cu 2+ steadily increased when the time increased from 5 to 120 min and reached equilibrium after 180 min without sorption of other metals (Figure 2).A 51.1% recovery efficiency was achieved at optimal time at the given resin loading (4.0 g/L).When the mass of resin in the solution was increased to 10.0 g/L, the recovery efficiency of Cu 2+ increased and reached 90.5% at 180 min (Figure 3).recovery efficiency was achieved at optimal time at the given resin loading (4.0 g/L).When the mass of resin in the solution was increased to 10.0 g/L, the recovery efficiency of Cu 2+ increased and reached 90.5% at 180 min (Figure 3).recovery efficiency was achieved at optimal time at the given resin loading (4.0 g/L).When the mass of resin in the solution was increased to 10.0 g/L, the recovery efficiency of Cu 2+ increased and reached 90.5% at 180 min (Figure 3).

Batch Sorption Isotherm
The sorption isotherm is important in the ion exchange process because it can reveal the interactive behavior between the adsorbents and ions in the solution.Therefore, the isotherm for the selective sorption of Cu 2+ using Dowex M4195 chelating resin at the given background concentrations of Ni, Fe, Zn and Al was studied.The batch sorption data obtained were fitted with Langmuir and Freundlich isotherm models.

Langmuir Isotherm
Langmuir isotherm assumes that adsorption is proportional to the fraction of the surface of the adsorbent that is open while desorption is proportional to the fraction of the adsorbent surface that is covered [37].Equation ( 10) is expressed in linear form [38].
where Q e (mg/g) is the adsorption capacity, C e (mg/L) is the equilibrium concentration of Cu 2+ ions, Q m (mg/g) is the maximum Cu 2+ ions sorption capacity, K L (L/mg) is the Langmuir constant obtained from the plot of C e /Q e vs. C e .The important parameter of the Langmuir isotherm was expressed as a dimensionless value R L , Equation (11), called the separation factor [39].
where K L is Langmuir constant (mg/g) and C 0 is the initial concentration of adsorbate (mg/g).When the R L value is in the range 0 < R L < 1, it is an indication that the sorption of Cu 2+ ions was favorable, unfavorable when R L > 1, linear when R L = 1, and irreversible when R L = 0.The sorption isotherm plot is presented in Figure 4 and the parameters calculated from fitting the plot are presented in Table 2.It is shown that the maximum sorption capacity (Q m ) is 5.89 mg/g and the separation factor R L is <1, which indicates a favorable sorption process.The correlation regression coefficient is 0.99 which indicates a good fit to the isotherm.

Batch Sorption Isotherm
The sorption isotherm is important in the ion exchange process because it can reveal the interactive behavior between the adsorbents and ions in the solution.Therefore, the isotherm for the selective sorption of Cu 2+ using Dowex M4195 chelating resin at the given background concentrations of Ni, Fe, Zn and Al was studied.The batch sorption data obtained were fitted with Langmuir and Freundlich isotherm models.

Langmuir Isotherm
Langmuir isotherm assumes that adsorption is proportional to the fraction of the surface of the adsorbent that is open while desorption is proportional to the fraction of the adsorbent surface that is covered [37].Equation (10) is expressed in linear form [38].
where Qe (mg/g) is the adsorption capacity, Ce (mg/L) is the equilibrium concentration of Cu 2+ ions, Qm (mg/g) is the maximum Cu 2+ ions sorption capacity, KL (L/mg) is the Langmuir constant obtained from the plot of Ce/Qe vs. Ce.The important parameter of the Langmuir isotherm was expressed as a dimensionless value RL, Equation ( 11), called the separation factor [39].
where KL is Langmuir constant (mg/g) and C0 is the initial concentration of adsorbate (mg/g).When the RL value is in the range 0 < RL < 1, it is an indication that the sorption of Cu 2+ ions was favorable, unfavorable when RL > 1, linear when RL = 1, and irreversible when RL = 0.The sorption isotherm plot is presented in Figure 4 and the parameters calculated from fitting the plot are presented in Table 2.It is shown that the maximum sorption capacity (Qm) is 5.89 mg/g and the separation factor RL is <1, which indicates a favorable sorption process.The correlation regression coefficient is 0.99 which indicates a good fit to the isotherm.Table 2.Estimated Isotherm Parameters.

Isotherm Langmuir Model
Freundlich Model

Freundlich Isotherm
This isotherm describes the surface heterogeneity and the exponential distribution of active sites and their energies [39].The linear form of this isotherm is presented in Equation ( 12) [40].
where K F is the sorption capacity (L/mg), and 1/n is the sorption intensity; and useful to predict heterogeneity of the adsorbate sites which are obtained from a plot of ln Q e vs. lnC e (Figure 5).The value of 1/n obtained shows that at high concentrations, the relative adsorption decreases, which is indicative of the saturation of adsorption sites available.
This isotherm describes the surface heterogeneity and the exponential distribution of active sites and their energies [39].The linear form of this isotherm is presented in Equation (12) [40].
where KF is the sorption capacity (L/mg), and 1/n is the sorption intensity; and useful to predict heterogeneity of the adsorbate sites which are obtained from a plot of ln Qe vs. lnCe (Figure 5).The value of 1/n obtained shows that at high concentrations, the relative adsorption decreases, which is indicative of the saturation of adsorption sites available.However, the smaller value of 1/n and larger KF values determined (as shown in Table 2) indicate that the sorbent has a higher sorption capacity and a strong affinity towards Cu 2+ ions.The slight deviation from the linear at the high concentration end in the Freundlich plot indicates saturation and thus multi-layer sorption for Cu 2+ is the less likely mechanism, despite the high correlation coefficient (R 2 ) for either isotherm model in the range studied.

Sorption Kinetics
The kinetics for the selective sorption of Cu 2+ from the multi-element leach solution by Dowex M4195 were evaluated.The modeling of experimental data was performed using pseudo-first and second-order kinetics.The calculated results obtained from pseudofirst and second-order Equations ( 13) and ( 14), adapted from [41], are given in Table 3.However, the smaller value of 1/n and larger K F values determined (as shown in Table 2) indicate that the sorbent has a higher sorption capacity and a strong affinity towards Cu 2+ ions.The slight deviation from the linear at the high concentration end in the Freundlich plot indicates saturation and thus multi-layer sorption for Cu 2+ is the less likely mechanism, despite the high correlation coefficient (R 2 ) for either isotherm model in the range studied.

Sorption Kinetics
The kinetics for the selective sorption of Cu 2+ from the multi-element leach solution by Dowex M4195 were evaluated.The modeling of experimental data was performed using pseudo-first and second-order kinetics.The calculated results obtained from pseudo-first and second-order Equations ( 13) and ( 14), adapted from [41], are given in Table 3.
where Q e (mg/g) and Q t (mg/g) are the sorption capacity at equilibrium and at time t (min), respectively; k 1 (1/min) and k 2 (g/mg/min) are the sorption rate constants of pseudo-first and second-order models.The coefficients (R 2 ) obtained from the plotted data for the two-kinetics models showed that the data fitted pseudo-second-order kinetics well, but a Minerals 2023, 13, 1285 9 of 18 poor fit was achieved for the pseudo-first-order plot (Figures 6 and 7).This is an indication that chemisorption occurred.
where Qe (mg/g) and Qt (mg/g) are the sorption capacity at equilibrium and at time t (min), respectively; k1 (1/min) and k2 (g/mg/min) are the sorption rate constants of pseudo-first and second-order models.The coefficients (R 2 ) obtained from the plotted data for the twokinetics models showed that the data fitted pseudo-second-order kinetics well, but a poor fit was achieved for the pseudo-first-order plot (Figures 6 and 7).This is an indication that chemisorption occurred.

Column Sorption of Copper
Based on optimal conditions obtained from the batch sorption study, a fresh stock of mixed metal leach solution was carefully diluted to give the target pH of 0.7.The solution was then pumped into the column at a flow rate in the range of 3-10 mL/min and for varying bed height (20-30 mm), and the performance of Dowex M4195 resin for selective sorption of Cu 2+ through the column was then evaluated.The loading behavior of metals

Column Sorption of Copper
Based on optimal conditions obtained from the batch sorption study, a fresh stock of mixed metal leach solution was carefully diluted to give the target pH of 0.7.The solution was then pumped into the column at a flow rate in the range of 3-10 mL/min and for varying bed height (20-30 mm), and the performance of Dowex M4195 resin for selective sorption of Cu 2+ through the column was then evaluated.The loading behavior of metals onto the resin was expressed in terms of the concentration C t /C 0 as a function of time (t), where C 0 and C t are the concentrations of metals in the inlet and outlet at time t, respectively.

Effect of Flow Rate on Breakthrough Curve
The plots of breakthrough curves are presented in Figure 8.It is evident that the breakthrough of Cu 2+ (i.e., its detection in the effluent) occurs almost instantaneously due to the short length of the resin bed (20 mm).The time for Cu 2+ adsorption to cease completely (resin saturation) decreased with increases in flow rate from 3 to 10 mL/min, as expected.Thus, Cu 2+ ion sorption attained saturation at 180 min with a flow rate of 3 mL/min, whereas when increased to 5 and 10 mL/min, saturation was noticed at 120 and 60 min, respectively.Further, the effect of flow rate on the breakthrough of other metal impurities was studied.The results presented in Figure 9 reveal that small amounts of Ni 2+ were adsorbed initially but were observed to have become saturated in less than 10 min, and this time becomes even shorter as the flow rate increases (Figure 9a-c).This behavior was due to some limited interaction of Ni 2+ ions with binding sites available on Dowex M4195 at highly acidic conditions.It is also observed that other impurities breakthrough immediately, showing minimal to no retention.

Effect of Bed Height on Breakthrough Curve
Conversely, an increase in the mass of resin (bed height) slightly delayed both the breakthrough point of Cu 2+ through the column and the time to reach saturation (Figure 10).When the bed height was increased from 20 to 25 mm, Cu 2+ saturation points were noticed at 90 and 120 min.Further increasing the bed height from 2.5 to 3.0 cm led to further lengthening of breakthrough time to 180 min.A limited number of different bed heights were used due to the limited quantity of resin available for the test work.

Effect of Bed Height on Breakthrough Curve
Conversely, an increase in the mass of resin (bed height) slightly delayed both the breakthrough point of Cu 2+ through the column and the time to reach saturation (Figure 10).When the bed height was increased from 20 to 25 mm, Cu 2+ saturation points were noticed at 90 and 120 min.Further increasing the bed height from 2.5 to 3.0 cm led to further lengthening of breakthrough time to 180 min.A limited number of different bed heights were used due to the limited quantity of resin available for the test work.

Effect of Bed Height on Breakthrough Curve
Conversely, an increase in the mass of resin (bed height) slightly delayed both the breakthrough point of Cu 2+ through the column and the time to reach saturation (Figure 10).When the bed height was increased from 20 to 25 mm, Cu 2+ saturation points were noticed at 90 and 120 min.Further increasing the bed height from 2.5 to 3.0 cm led to further lengthening of breakthrough time to 180 min.A limited number of different bed heights were used due to the limited quantity of resin available for the test work.

Determination of Column Parameters
Sorption experimental data obtained based on the flow rate and bed height were substituted into Equations ( 3)- (7), and the evaluated results are presented in Table 4.It is shown that the sorption capacity (q e ) decreases with increases in flow rate and slightly increases when bed height (mass of resin) is increased.The decrease in the amount of Cu 2+ ions uptake by the resin (q t ) as the flow rate increases from 3 to 10 mL/min is a result of a low retention or contact time for Cu 2+ ions to interact with charged resin, coupled with slower diffusion of Cu 2+ ions into the resin binding sites.Similar results have been observed in the literature albeit under different experimental conditions in terms of sorbents, solutions, and concentrations [42][43][44][45][46][47].However, this behavior was observed to affect the sorption efficiency of Cu 2+ which decreases with higher flow rates, because desorption of already adsorbed Cu 2+ ions onto resin is possible at higher flow rates.The result, however, revealed that the best way to increase the residence time is to decrease the flow rate and increase the bed height.This is proven as total sorption (q t ) and the sorption efficiency (% R) for Cu 2+ increased as bed height increased from 2.0 to 3.0 cm and decreased with flow rate from 10 to 5 mL/min.C 0 -influent concentration (mg/L), Q-flow rate (mL/min), H-bed height (mm), q t -adsorption at breakthrough (mg g −1 ), q eq -adsorption at saturation (mg/g), m t -total amount of Cu 2+ ions, %R-total Cu 2+ sorption at saturation (%).

Batch and Column Desorption
In the literature, the NH 3 solution was reported to be effective for the desorption of Cu 2+ from Dowex M4195 [31].However, the complexation of NH 3 and Cu 2+ in the form of cupric amine ([Cu(NH 3 ) 4 ] 2+ ) is unsuitable for subsequent processing in direct electrowinning to achieve a quality cathode product.Better current density and cathode product purity can be achieved from sulfate media, and this therefore makes sulfuric acid a much superior choice for copper desorption [11,48,49].For this work, concentrated H 2 SO 4 was used for the desorption of Cu 2+ ions from Dowex M4195 resin and regenerated.Initially, batch elution studies were performed, and the results are presented in Figure 11.They show that 90% desorption of Cu 2+ ions was achieved in 60 min using 2.0 M H 2 SO 4 .Almost complete desorption of Cu 2+ was also achieved when the strength of the desorbing solution increased further.Based on the batch elution profile, column desorption studies were conducted using 2.0 M H 2 SO 4 and the result is presented in Figure 12.Instant breakthrough of the trapped impurities was observed, while the desorption rate for Cu 2+ increased when increasing the flow rate from 5 to 10 mL/min (Figure 12a,b).This implies that the concentration of eluted copper was 1.0 g/L over the first 5 min at 5.0 mL/min and decreased to 0.27 g/L when the flow rate was changed to 10 mL/min.Cu 2+ was completely desorbed when eluting with 2.0 molar H 2 SO 4 at 10 mL/min after 120 min.The resin could be regenerated with distilled water and reused.

Modelling of Column Data
The sorption data were fitted to both the Thomas and the Yoon-Nelson model, thereafter the relevant parameters were evaluated from a plot of ln [Co/Ct − 1] against t as shown in Equations ( 8) and (9).For the Thomas model, the value of sorption capacity (q0) and rate constant (KTh) were obtained from the intercept and slope.Results presented in Table 5 show that 0 values increased with the flow rate, and slightly increased with the bed depth.Similar trends were found in the results of other scholars [50,51].The correlation coefficient (R 2 ) value is approximately 0.90 which indicates the experimental data fitted the Thomas model reasonably well.

Modelling of Column Data
The sorption data were fitted to both the Thomas and the Yoon-Nelson model, thereafter the relevant parameters were evaluated from a plot of ln [Co/Ct − 1] against t as shown in Equations ( 8) and (9).For the Thomas model, the value of sorption capacity (q0) and rate constant (KTh) were obtained from the intercept and slope.Results presented in Table 5 show that 0 values increased with the flow rate, and slightly increased with the bed depth.Similar trends were found in the results of other scholars [50,51].The correlation coefficient (R 2 ) value is approximately 0.90 which indicates the experimental data fitted the Thomas model reasonably well.

Modelling of Column Data
The sorption data were fitted to both the Thomas and the Yoon-Nelson model, thereafter the relevant parameters were evaluated from a plot of ln [C 0 /C t − 1] against t as shown in Equations ( 8) and (9).For the Thomas model, the value of sorption capacity (q 0 ) and rate constant (K Th ) were obtained from the intercept and slope.Results presented in Table 5 show that q 0 values increased with the flow rate, and slightly increased with the bed depth.Similar trends were found in the results of other scholars [50,51].The correlation coefficient (R 2 ) value is approximately 0.90 which indicates the experimental data fitted the Thomas model reasonably well.
Values of the rate constant (k YN ), and time required for half Cu 2+ ions concentration to break through the column (τ) were estimated from the slope and intercept of the Yoon-Nelson plots at different flow rates and bed heights as shown in Table 5.The values of k YN increased with increasing flow rate and bed height.Values of τ decreased as the flow rate increased but slightly increased as bed height increased.The decrease in τ as the flow rate increases indicated that the resin bed exhausted more quickly as the flow rate increases, this behavior is evident in most sorption processes [50,52].The value of τ (min) in Table 5 represents the time for the Cu concentration in the column effluent to reach 50% of its final level.However, a higher value of τ (min) signifies better performance of the column as similarly reported in the literature [53,54].Therefore, these experimental data were found to be well-fitted with the Yoon-Nelson model with R 2 values greater than 0.94.This is an indication that internal mass transfer may be the limiting step.

Treatment of Raffinate
After complete sorption of Cu 2+ ions from the mixed metal sulfate liquor through multiple contacts with the resin, the pH of the raffinate was adjusted (pH 2.0) and nickel was sorbed using Dowex M4195 resin.The test work followed a similar methodology to Cu adsorption.The result obtained from batch experiments shows an increase in sorption of Ni 2+ from 40.1 to 48.1% as the solution pH changed from 2.0 to 3.0.More so, the result presented in Figure 13 shows that Dowex M4195 sorbed Ni 2+ with traces of Fe 3+ which reached saturation at 120 and 20 min, respectively.A similar trend was reported in the literature, where Dowex M4195 showed strong affinity towards Ni 2+ , and Fe 3+ broke through the column due to its low affinity and concentration at pH 2.0 [31].Hydrochloric acid was effective for the removal of iron sorbed onto Dowex M4195.As presented in Figure 14, different concentrations of HCl solution were tested, and 99.9% of the sorbed trace Fe 3+ ion was selectively removed from the column by 0.2 M HCl, leaving behind the Ni 2+ on the resin which was later easily eluted with 2.0 M H 2 SO 4 .As the final step of the process, the remaining Fe 3+ in the raffinate can be easily precipitated using Ca (OH) 2 , leaving behind traces of zinc and aluminum in the liquor which can be purified further.The overall process flow for the leaching of PCBs and selective recovery of Cu and Ni and rejecting minor impurities from the leach liquor is shown in Figure 15.

Conclusions
In this study, recovery of base metals from mixed metals leach liquor of electronic waste materials via an ion exchange process using Dowex M4195 resin is shown to be feasible in principle.Batch and fixed bed sorption of Cu 2+ ions on Dowex M4195 resin and removal of Ni 2+ from raffinate using the spent resin at higher pH were investigated.The results showed that the resin selectively sorbed Cu 2+ ions at highly acidic conditions (pH < 1.0).Desorption of Cu 2+ from resin was achieved with 2.0 M H2SO4.A higher flow rate was found to speed up the breakthrough curve but lead to an overall decrease in Cu 2+ ion uptake while increasing the bed height delayed the breakthrough curve and more Cu 2+

Conclusions
In this study, recovery of base metals from mixed metals leach liquor of electronic waste materials via an ion exchange process using Dowex M4195 resin is shown to be feasible in principle.Batch and fixed bed sorption of Cu 2+ ions on Dowex M4195 resin and removal of Ni 2+ from raffinate using the spent resin at higher pH were investigated.
The results showed that the resin selectively sorbed Cu 2+ ions at highly acidic conditions (pH < 1.0).Desorption of Cu 2+ from resin was achieved with 2.0 M H 2 SO 4 .A higher flow rate was found to speed up the breakthrough curve but lead to an overall decrease in Cu 2+ ion uptake while increasing the bed height delayed the breakthrough curve and more Cu 2+ uptake was obtained.Adsorption followed both Langmuir and Freundlich isotherm models as well as pseudo-second-order kinetics.The column data modeled indicates that the sorption of Cu 2+ fitted the Yoon-Nelson model.It was also demonstrated that spent Dowex M4195 could be used further towards Ni 2+ sorption when contacted with the raffinate from the Cu adsorption step at pH 2, while dilute HCl facilitated selective elution of iron co-sorbed with Ni from the resin, before eluting the Ni again with 2.0 M H 2 SO 4 .The Fe 3+ remaining after Cu and Ni extraction can be precipitated with Ca (OH) 2 leaving a final raffinate free from iron, nickel, and copper for further processing.

Figure 4 .
Figure 4. Langmuir plot for the sorption of Cu 2+ ion from multi-metal leach solution by Dowex M4195.

Figure 4 .
Figure 4. Langmuir plot for the sorption of Cu 2+ ion from multi-metal leach solution by Dowex M4195.

Figure 5 .
Figure 5. Freundlich plot for the sorption of Cu 2+ ion from multi-metal leach solution by Dowex M4195.

Figure 5 .
Figure 5. Freundlich plot for the sorption of Cu 2+ ion from multi-metal leach solution by Dowex M4195.

Figure 7 .
Figure 7. Pseudo second order plot for the sorption of Cu 2+ ion from multi-metal leach solution by Dowex M4195.

Figure 11 .Figure 12 .
Figure 11.Batch desorption profile of Cu 2+ ions from Dowex M4195 using different concentrations of H2SO4 at 25 °C for 60 min using 4.0 g/L resin/solution ratio.

Figure 11 .
Figure 11.Batch desorption profile of Cu 2+ ions from Dowex M4195 using different concentrations of H 2 SO 4 at 25 • C for 60 min using 4.0 g/L resin/solution ratio.

Figure 11 .Figure 12 .
Figure 11.Batch desorption profile of Cu 2+ ions from Dowex M4195 using different concentrations of H2SO4 at 25 °C for 60 min using 4.0 g/L resin/solution ratio.

Figure 15 .
Figure 15.The proposed schematic flow sheet for the sequential recovery of copper, nickel, and zinc from the polymetallic sulfate leach solution of WEEE.

Figure 15 .
Figure 15.The proposed schematic flow sheet for the sequential recovery of copper, nickel, and zinc from the polymetallic sulfate leach solution of WEEE.

Table 1 .
Composition of sulfate leach liquor of waste electrical materials. )

Table 3 .
Estimated kinetics parameters for the selective sorption of Cu 2+ by Dowex M4195.

Table 3 .
Estimated kinetics parameters for the selective sorption of Cu 2+ by Dowex M4195.
Figure 7. Pseudo second order plot for the sorption of Cu 2+ ion from multi-metal leach solution by Dowex M4195.

Table 4 .
Calculated column parameters for the recovery of copper by Dowex M4195.

Table 5 .
Column parameters predicted by Thomas and Yoon-Nelson model for Cu 2+ by Dowex M4195.

Table 5 .
Column parameters predicted by Thomas and Yoon-Nelson model for Cu 2+ by Dowex M4195.

Table 5 .
Column parameters predicted by Thomas and Yoon-Nelson model for Cu 2+ by Dowex M4195.= bed height (cm), C 0 is initial concentration of Cu (mg/L), Q is the flow rate (mL/min), K TH is the Thomas constant (mL/mg/min), q o is the adsorption capacity (mg/g), k YN is the Yoon-Nelson rate constant (1/min), τ is the half-breakthrough time (min) and R is the regression coefficient. H