Purification of nickel or cobalt ion containing effluents by electrolysis on reticulated vitreous carbon cathode

Hydrometallurgy 150 (2014) 1–8

Contents lists available at ScienceDirect

Hydrometallurgy journal homepage: www.elsevier.com/locate/hydromet

Purification of nickel or cobalt ion containing effluents by electrolysis on reticulated vitreous carbon cathode A. Dell'Era a,⁎, M. Pasquali b, C. Lupi c, F. Zaza d a

Department DME, Università degli studi, Guglielmo Marconi, Rome, Italy University Sapienza Rome, Dept. SBAI, Via del Castro Laurenziano 7, 00161 Roma, Italy University Sapienza Rome, Dept. ICMA, Via Eudossiana 18, 00184 Roma, Italy d ENEA-Casaccia R.C., Via Anguillarese 301, 00123 S. Maria di Galeria (RM), Italy b c

a r t i c l e

i n f o

Article history: Received 27 March 2014 Received in revised form 28 August 2014 Accepted 1 September 2014 Available online 18 September 2014 Keywords: Nickel and cobalt recycling Dimensionless analysis Reticulated vitreous carbon Electrochemical filter

a b s t r a c t The study is aimed at improving a methodology to purify nickel or cobalt ion containing effluents by an electrochemical filter. Consequently, a cell with reticulated vitreous carbon (RVC) cathode for Ni and Co depletion was designed to analyze, at room temperature and pH equal to 6, flowrates and cathodic potentials able to affect the electrochemical process. Starting with Ni and Co initial concentration of 150 ppm, it is possible to reach a concentration lower than 0.1 ppm for both metals in less than 1 h, with a flowrate of about 1300 ml/min and for a catholyte volume of 1000 ml, under mass transport control conditions (for Ni −1.1 V and for Co −1.2 V cathodic potential versus Standard Calomel Electrode (SCE)). Moreover, the proposed work studied the process kinetics and fluid dynamics through the use of dimensionless relations such as Reynolds and Sherwood numbers. In addition considerations on current efficiency for reduction of both metal ions were done. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Material recovery and the need not to compromise the environment are two factors that today are no longer negligible. This becomes of more relevance if resources are strategic materials and wastes, containing pollutants to the environment, are harmful to humans. The many sectors using batteries (with Ni and Co content) tend to constantly increase their demand, and strong increase in the production goes along with large amounts of exhausted batteries, which enhances the disposal problems. Indeed, being “hazardous waste”, they must be collected separately and treated appropriately before their final discharge in controlled landfills. In this field, technologies for exhausted battery recycling (Lupi et al., 2006; Pasquali and Lupi, 2003; Pistoia et al., 2001; Wang et al., 2005) assume fundamental importance, in order to both recover and reuse metals, as in producing electrochemically alloys for hydrogen evolution reaction in electrolysis processes (Lupi et al., 2009, 2011, 2013), as well as to make inert such wastes, so they can be disposed of in accordance with local regulations. The selective recovery of metals is extremely advantageous not only from the economic aspect, closely linked to the purity degree of recovered metals, but also from the environmental aspect as saving natural resources and impact on the environment and human health. Despite those treatments, the Ni and Co are not completely recovered and the ⁎ Corresponding author at: Department DME, Università degli studi, Guglielmo Marconi,Via Plinio 44, 00193 Roma, Italy. E-mail address: [email protected] (A. Dell'Era).

http://dx.doi.org/10.1016/j.hydromet.2014.09.001 0304-386X/© 2014 Elsevier B.V. All rights reserved.

process solid or liquid effluents contain metal concentration that exceeds limits allowed by environmental legislation. Electrochemical technologies have recently attracted attention, for they allow metal recovery in their most valuable form (zero-oxidation state), without requiring addition of chemicals, and thereby not generating byproducts which would later require treatment or confinement. Moreover, electro-recovered metals and waters treated using these electrochemical methods can be reused in the same process. That avoids

Fig. 1. Experimental apparatus schema.

2

A. Dell'Era et al. / Hydrometallurgy 150 (2014) 1–8

Fig. 2. Electrochemical cell: a) anodic plate with the silicon layer, the membrane, the first rectangular piece of RVC cathode and Pt wire, one above the other in sequence; b) cathodic plate with the second rectangular piece, before closing the cell; c) anodic side of the assembled cell; d) cathodic side of the assembled cell.

economic loss during treatment, and reduces both water consumption and environmental impact. In particular, when the metal ion concentration of the solution is very low, or at the end of electrodeposition process, the mass transport may become the limiting step, and use of electrodes with high specific area may be required. This is one of the reasons for the very large success obtained by three-dimensional electrodes, RVC, in the electrochemical treatment of wastewater (Bertazzoli et al., 1998; Czerwinski et al., 2010; Dell'Era et al., 2008; Dutra et al., 2000; Friedrich et al., 2004; Lanza and Bertazzoli, 2000; Pletcher et al., 1991; Podlaha and Fenton, 1995; Polcaro et al., 1999; Ramalan et al., 2012; Tangirala et al., 2010; Tramontina et al., 2002; Widner et al., 1998). Moreover, because of the low current of metal reduction, the reactions related to the solvent reduction, can become important and can affect the faradic yield of the process. If the main reactions are kinetically controlled, a higher current efficiency could be obtained by working at lower cathodic potential, but it must be controlled in the range at which metal reduction is under mass transport control for having the maximum value of deposition rate. In this context, the aim of this work is to emphasize how, by using electrochemical method, and in particular with the use of three-dimensional electrodes such as RVC ones, it is however possible, to obtain a satisfactory decrease of Ni and Co metal ions in solution. Specifically in this paper, starting from the results obtained in a previous work (Lupi et al., 2005), in particular by considering the after treated electrolyte, as Ni or Co containing effluents, a comparison between the electrochemical depletion of those divalent ions, by using RVC cathode has been done, highlighting the differences Table 1 Characteristics of reticulated vitreous carbon model LS190727 JV of Goodfellow. Porosity Density Specific surface Dimensions 3-d electrode volume

% g/cm3 cm2/cm3 mm3 cm3

97 0.05 27 100 × 100 × 6.5 127

and all the parameters that can affect the process such as, first and foremost, the working electrode potential and the flow rate. Furthermore a dimensionless analysis for each ionic species has been performed determining different results for the two ions.

2. Experimental 2.1. Apparatus Flow cell, especially designed, with three electrodes in flow-by configuration was used to perform experimental tests. Peristaltic pumps are used to recirculate the solution between the reservoirs and the two compartment of electrolytic cell. The electrochemical cell, made in Plexiglas, is constituted by cathodic and anodic compartments. Within the cathodic compartment there are two rectangular pieces of RVC between which is inserted a platinum wire to bring out electrical contact from the cell. The two compartments are separated by an anionic membrane 125 × 125 mm (BDH Laboratory Supplies). A cell outline is shown in Fig. 1, while in Fig. 2 the cell configuration and the flow direction are illustrated. In particular in Fig. 2a) the anodic plate, the silicon layer, the membrane, the first rectangular piece of RVC cathode and Pt wire are shown, one above the other in sequence; in Fig. 2b) the cathodic plate with the second rectangular piece is shown before closing the cell. In Fig. 2c) it is shown the anodic side of the assembled cell, the flow Table 2 Electrolysis operative conditions for Ni. Catholyte and anolyte volume [ml] Catholyte solution Anolyte solution Working potential (SCE) [V] range Flowrate [ml/min] pH

1000 ml NiSO4° 150 ppm in Na2SO4 0.1 M Na2SO4 0.5 M −1.0 – −1.3 V (mass transport control conditions) 500; 1300 ≈6

A. Dell'Era et al. / Hydrometallurgy 150 (2014) 1–8

3

Table 3 Electrolysis operative conditions for Co. Catholyte and anolyte volume [ml] Catholyte solution Anolyte solution Working potential (SCE) [V] range Flowrate [ml/min] pH

300 ml (up to 300 ml/min flowrate) 1000 ml (over 300 ml/min flowrate) CoSO4° 150 ppm in Na2SO4 0.1 M Na2SO4 0.5 M −1.1 V – −1.4 V (mass transport control conditions) 870; 1250; 1640 ≈6

direction and, orthogonally, the reference electrode; finally, in Fig. 2d) the cathodic side is illustrated. The cathodic compartment has dimensions 100 × 100 × 15 mm, while the anodic one is 100 × 95 × 15 mm. As said before, two layers of RVC (Goodfellow, LS190727 JV), whose characteristics are summarized in Table 1 were placed inside the cathodic compartment. Each carbon layer has a thickness of 6.35 mm and between the two layers a wire of platinum was placed, working as an external electrical contact to which the potentiostat (EG & G model 273) connects. The counter electrode, in the anodic compartment, is constituted by a net of platinized titanium, dipped into the solution, which is also in electrical contact with the potentiostat. Finally there is a calomel reference electrode for the measurement of the cathodic potential. The working potential is always related to the calomel electrode. The sealing of the cell is ensured by a layer of silicone interposed between the two compartments, held together by a series of through-screws. The solution concentration was analyzed on successive withdrawals, from the catholyte volume, by an atomic absorption spectrometer: AAS (Solaar series M).

Fig. 4. ln(c/c0) vs. time at − 1.1 V constant potential and different flow rates for Ni depletion.

2.2. Ni and Co electrolysis operative conditions After the Ni and Co recovery process studied in a previous work (Lupi et al., 2005), in this paper the depletion process of the two ions in solution has been considered. The experimental baths containing Ni+2 and Co+2 ions have been prepared starting from cathodic material (LiCoO2, or LiNiO2) of lab lithium ion batteries. It was dissolved with H2SO4/H2O2 solution having appropriate ratio, as described in the previous work (Lupi et al., 2005), and the solution obtained was diluted until the desired initial concentration. Therefore the solution bath had 150 ppm of Ni+2 or Co+2, to which Na2SO4 was added as supporting electrolyte. Before each electrolysis test, the catholyte solution has been deoxygenated by using argon for 20 min, with solution circulating into the cell. Tables 2 and 3 report electrolysis operative conditions respectively for nickel and cobalt ions in solution. The catholyte pH value was maintained at 6 by adding a dilute solution of sulphuric acid. The working potential was in the range of − 1.1 and − 1.4 V for Ni depletion and −1.0 and −1.3 V for Co depletion in order to determine the conditions of mass transport control. The rate of electrolysis can be expressed in terms of reactant concentration decrease that, for a batch reactor containing a three-dimensional electrode, can be assumed as:  ct ¼ c0  exp

K m Ae V r Vt

−

 t

ð1Þ

for the global process. Eq. (1) is used for PFR (Plug Flow Reactor) as that used in our experiments and reported in literature (Bertazzoli et al., 1998; Dutra et al., 2000; Pletcher et al., 1991; Polcaro et al., 1999; Ponce de Leon et al. 1996). Where -

ct is the concentration at time t; [s] Vr is the 3-d electrode volume; [cm3] Ae is the 3-d electrode specific surface; [cm−1] Km is the mass transport coefficient; [cm/s]

Table 4 Slope values at different flowrates and cathodic potentials for Ni depletion.

Fig. 3. Polarization curves for the electrochemical reduction on RVC of a) Ni+2 ion; b) Co+2 ion.

Nickel

Slopes

Flowrate (ml/min)

−1.0 V

−1.1 V

−1.2 V

−1.3 V

500 750 1000 1300

−7 × 10−4 −1 × 10−3 −1.3 × 10−3 −1.5 × 10−3

−1 × 10−3 −1.3 × 10−3 −1.6 × 10−3 −2.2 × 10−3

−1 × 10−3 −1.2 × 10−3 −1.7 × 10−3 −2.1 × 10−3

−1 × 10−3 −1.3 × 10−3 −1.7 × 10−3 −2.1 × 10−3

4

A. Dell'Era et al. / Hydrometallurgy 150 (2014) 1–8

Table 5 AeKm values at different catholyte flowrates for Ni depletion. Nickel Flowrate (ml/min)

Linear velocity (cm/s)

AeKm

500 750 1000 1300

0.66 1.00 1.30 1.70

0.008 0.010 0.013 0.017

- Vt is the solution volume equal to 300 cm3 for low flows and 1000 cm3 for high flows Making explicit the relationship (1) as ln (c/c0) versus time, which interpolates the experimental points, the slope m of the straight line is obtained: m¼

Ae  K m  V r Vt

ð2Þ

The value of AeKm product can be derived from the Eq. (2) and then Km, required for calculating the dimensionless Sherwood, may be determined. Preliminary polarization tests have been performed (using a flowrate of about 1300 ml/min) to determine the mass transport control conditions in terms of electrode potential for Ni and Co electrodeposition on RVC cathode. Fig. 3a) and b) show that at about − 1.1 V and −1.2 V respectively for Ni and Co the mass transport control conditions are reached. Besides for Co the limiting current value is higher than that of Ni, there will also be a higher AeKm value. By increasing the electrode potential toward more cathodic values, the hydrogen evolution reaction becomes evident in both depletion reactions. 3. Results and discussion 3.1. Nickel depletion Electrolysis tests were performed at constant potential and by varying flowrates for each potential used. Experiments were repeated several times in order to test data repeatability and result validity. Fig. 4 shows the results obtained by performing Ni reduction at − 1.1 V cathodic potential, using four different values of flow rate, while Table 4 highlights slope values obtained in similar experiments performed respectively at −1.0 V, −1.1 V, −1.2 V and −1.3 V cathodic potential. As it can be observed, the slope values, for each different flow rates, remain constant, confirming that, even operating at − 1.1 V cathodic potential, it is working under mass transport control at conditions of limit current. Thus for each flow rate value, changing the electrode potential, from − 1.1 V toward more cathodic potentials, the

Fig. 6. XRD diffractogram of 3d reticulated vitreous carbon cathode presenting Ni deposit.

slope remains constant (reading along the lines). Instead, changing the flowrate, the slopes also change (reading along the columns), because with changing the flowrate, the AKm values and then the m values of the Eq. (2) also change. Measuring the ion concentration during the electrolysis process by atomic absorption method and plotting the neperian logarithm of normalized concentration, C/Co, versus time (Eq. (1)), the straight lines, shown in Fig. 4, are obtained and by slope calculation (Eq. (2)) it is possible to determine the AeKm values shown in Table 5 The values of AeKm, are the product of the specific surface area and the mass transport coefficient, obtained for different flow regimes. This value increases with the flow rate as expected. Knowing the cell geometry and the RVC porosity, the flow passage cross section (Table 1) has been calculated and therefore, for each flow value, the linear speed has been evaluated and reported in Table 5. The deposit, obtained by reduction process of nickel ions on the three-dimensional electrode surface, has been also observed by SEM (scanning electron microscope); it can be seen (Fig. 5 a) and b)) that the deposit is homogeneous, compact and uniformly distributed on the electrode; moreover, the electrode pores are not clogged. X-ray diffraction (XRD) analysis (Cu Kα source: λ = 1,5418 Å) has been performed, confirming that the deposit on RVC is pure Ni (fcc structure and Fm3m space group) as shown in Fig. 6 where the experimental diffractogram is compared with the main reflection peaks of Ni, by using as source powder diffraction file.

Fig. 5. SEM micrograph of 3d reticulated vitreous carbon cathode presenting Ni deposit.

A. Dell'Era et al. / Hydrometallurgy 150 (2014) 1–8

5

Table 6 Slope values at different flowrates and cathodic potentials for Co depletion. Cobalt

Slopes

Flowrate (ml/min)

Vc [ml]

−1.1 V

100 200 300 870 1250 1650

300 300 300 1000 1000 1000

−2.5 −2.5 −3.5 −1.5 −1.7 −1.9

× × × × × ×

−1.2 V 10−3 10−3 10−3 10−3 10−3 10−3

−3.0 −4.9 −6.2 −1.9 −2.2 −2.4

× × × × × ×

−1.3 V 10−3 10−3 10−3 10−3 10−3 10−3

−2.9 −4.9 −6.2 −2.0 −2.2 −2.5

× × × × × ×

−1.4 V 10−3 10−3 10−3 10−3 10−3 10−3

−3.0 −5.0 −6.2 −2.0 −2.2 −2.5

× × × × × ×

10−3 10−3 10−3 10−3 10−3 10−3

3.2. Cobalt depletion Electrolyses were performed at different potentials, varying flow rates at each potential. Whenever the potential for which the reaction is controlled by mass transport was reached, the slope of the straight lines ln(c/c0) versus time remained constant, even decreasing the potential. An increasing of the catholyte flow in the hydraulic circuit, leads to the increase of turbulence that accentuates the discharge process. Reporting the trend of ln(c/c0) versus time (Eq. (1)) and by fitting experimental points, it can be clearly observed that by increasing the flow rate, the slope of straight lines increases, indicating a better metal depletion. Table 6 shows calculated slope values, using Eq. (2), at −1.1 V, −1.2 V, −1.3 V and −1.4 V for different flow rates. As before, also for Co+ 2 depletion, starting from − 1.2 V toward more cathodic electrode potentials, (then decreasing the electrode potential value) the slopes, for each flow rate value, remain constant, because the process is under mass transport control condition. Indeed, reading along the lines of Table 6, it is possible to see that from − 1.2 V to more cathodic potential (− 1.3 V, − 1.4 V) the slopes are very similar and sometimes equals (then constants). Therefore, as for nickel, changing the flowrate, the slopes also change (reading along the columns).The graphs of Figs. 7 and 8 show this trend in various flow conditions, maintaining the cathodic potential at −1.2 V (reading along the fourth column of Table 6 related to − 1.2 V). The used catholyte volume was 300 ml up to 300 ml/min flow rate (Fig. 7), while it was 1000 ml when higher flow rates were used (Fig. 8). Even if the depletion processes in Fig. 7 are made by using lower flowrate values (and then lower AKm values) respect those used in experiments of Fig. 8, the time required to reach the final concentration is lower because of lesser catholyte volume used. Table 7 shows the AeKm values for Co electrodeposition; as for Ni, also Co depletion increases with increasing flow rate and linear speed. SEM observations performed on Co deposits show that (Fig. 9a) and b)), unlike Ni, a dendritic aspect is present; thus using RVC cathode in electrolysis, as time goes by, its pores are completely clogged. XRD analysis confirmed the depletion of pure Co ions on RVC, as shown in Fig. 10, where, as for Ni, the experimental Co

Fig. 8. ln(c/c0) vs. time at −1.2 V constant potential and different flow rates (catholite volume: 1000 ml) for Co depletion.

diffractogram is compared with the higher peaks of pure Co (hcp structure and P6 3 /mmc space group), by using as source powder diffraction file. Differently from Ni the Co dendritic deposit results to be quite amorphous. 3.3. Dimensionless analysis A dimensionless analysis with a voltage value that guarantees a mass control regime for different flow rates, thus summarizing all the results, was performed, by considering: 27 cm2/cm3 specific surface area Ae, 0.65 × 10− 5 cm2/s diffusion coefficient D for both Ni and Co ions (Leaist (1989)), 97% RVC void volume ε and 1.1 × 10−2 cm2/s solution kinematic viscosity ν. Particularly, log(Sh) versus log(Re) was considered and reported in Fig. 11 for Ni and in Fig. 12 for Co depletion, in which the Sherwood number Sh = Kmε/(ΑεD) is, therefore, related to the Reynolds number Re = v/(ν Ae) where v is the linear speed. The following equations were derived from the above data, for Ni and Co ions respectively: LogðShÞ ¼ 0:780  Log ðRe Þ−0:036

ð4Þ

LogðShÞ ¼ 0:293  Log ðRe Þ þ 0:392

ð5Þ

It is thus possible to write relationships between Reynolds, Sherwood and Schmidt numbers, whereby Sc is the Schmidt number (ν/D). Sh ¼ 0:077  Re

0:78

Sh ¼ 0:207  Re

0:293

Sc

1=3

ð6Þ

1=3

ð7Þ

Sc

Such dimensionless equations are very useful to design these electrochemical systems and to perform a possible scale-up. In Figs. 11 and 12 different trends of Sh number versus Re number are shown respectively for Ni and Co ion depletion. Considering Co depletion the curve has a minor slope value and this is due to the

Table 7 AeKm values at different catholyte flowrates for Co depletion. Cobalt

Fig. 7. ln(c/c0) vs. time at −1.2 V constant potential and different flowrates (catholite volume: 300 ml) for Co depletion.

Flowrate (ml/min)

Linear velocity (cm/s)

AeKm

100 200 300 870 1250 1650

0.13 0.27 0.39 1.15 1.65 2.17

0.007 0.012 0.014 0.016 0.018 0.020

6

A. Dell'Era et al. / Hydrometallurgy 150 (2014) 1–8

Fig. 9. SEM micrograph of 3d reticulated vitreous carbon cathode with Co deposit.

deposition nature. Generally, the flow rate can affect the thickness of diffusive boundary layer, thus affecting Km value: if considering the dendritic aspect of Co deposit, the transport coefficient Km is higher (more thin boundary layer) but less influenced by flow rate condition, that is represented from Re number and consequently the slope is quite flat. 3.4. Current efficiency analysis The current efficiency has been calculated for every interval, with referring at two successive samplings, by the ratio between the Coulombs (QΔC) relative to the ion concentration change ΔC (obtained from AAS) and the total Coulombs (Qtot) really passed through the cell: Y ¼ Q ΔC =Q tot

ð8Þ

Q total is obtained from the potentiostat and Q ΔC is: Q ΔC ¼

ΔC  V  n  F 1000  PM

ð9Þ

where ΔC is the difference between the solution concentration at time t

Fig. 10. XRD diffractogram of 3d reticulated vitreous carbon cathode presenting Co deposit.

and the concentration at the time of next sampling; V is the solution volume; 1000 is the conversion factor from mg to grams; PM is the ion atomic mass; n is the valence ion; and F is the Faraday constant: 96500 C. Considering the current efficiency behavior versus ΔC% as shown in Figs. 13 and 14 for Co and Fig. 15 for Ni depletion, related to tests performed with potential in mass transport control conditions and by varying the flow to get an idea of the values on which current efficiency attests, in all the cases it can be asserted that, increasing flowrate, increases the current efficiency with which the same ΔC% is obtained; furthermore the time required to get the same percentage reduction decreases (Figs. 4, 7 and 8). It can be seen how the current efficiencies, in different flow conditions, decrease with concentration decreasing. The nickel depletion process is still on average more efficient than that of cobalt, as demonstrated by the fact that the efficiency of nickel depletion, obtained as ratio between total theoretical coulomb and actual coulomb passed up to reaching the final concentration, is 39%, while for cobalt is 15%. This difference is, also, attributed to the different deposit morphology: Co deposit has a dendritic aspect, while Ni deposit is quite smooth; therefore on dendritic deposit with higher specific surface area the hydrogen evolution reaction is easy, thus decreasing the current efficiency of the process.

Fig. 11. log(Sh) vs. log(Re) for Ni electrodeposition.

A. Dell'Era et al. / Hydrometallurgy 150 (2014) 1–8

7

Fig. 12. log(Sh) vs. log(Re) for Co electrodeposition.

Fig. 14. Efficiency vs percentage variation of the Co ion concentration for high flow values.

Fig. 13. Efficiency vs percentage variation of the Co ion concentration for low flow values.

4. Conclusion The study is aimed at improving a methodology to purify nickel or cobalt ion containing effluents. The treated effluent comes from battery hydrometallurgical recovery process. In this context, the objective of this work was to emphasize how, by using electrochemical methodology, particularly with the use of RVC three-dimensional electrodes, it is possible to obtain, in the most convenient way, a satisfactory decrease of metals in wastewater. Indeed, an electrolytic cell, equipped with a RVC three-dimensional cathode working at controlled potential, was set up as a true electrochemical filter, useful for metal depletion, that allows to reach in the solution a metal concentration below 0.1 ppm in less than 1 h, with a flowrate of about 1300 ml/min and for a catholyte volume of 1000 ml, under mass transport control conditions (for Ni −1.1 V and for Co −1.2 V cathodic potential versus standard calomel electrode (SCE)). All the parameters that could affect the process such as, above all, the electrode working potential as well as the flow rate were analyzed. The deposit morphology was also considered. The proposed work also studied the process kinetics and fluid dynamics determining dimensionless relationships for both metals, useful to perform an electrochemical system design and a potential scale-up. Energy analysis shows that the average current efficiency for the removal of Ni+2 and Co+2 ions is respectively 39% and 15%. References Bertazzoli, R., Rodrigues, C.A., Dallan, E.J., Fukunaga, M.T., Lanza, M.R.V., Leme, R.R., Widner, R.C., 1998. Mass transport properties of a flow-through electrolytic reactor

Fig. 15. Efficiency vs percentage variation of the Ni ion concentration.

using a porous electrode: performance and Figures of merit for Pb+2 removal. Braz. J. Chem. Eng. 15, 4. http://dx.doi.org/10.1590/S0104-66321998000400008. Czerwinski, A., Obrebowski, S., Kotowski, J., Rogulski, Z., Skowronski, J.M., Krawczyk, P., Rozmanowski, T., Bajsert, M., Przystałowski, M., Buczkowska-Binieck, M., Jankowska, E., Baraniak, M., 2010. Electrochemical behavior of negative electrode of lead-acid cells based on reticulated vitreous carbon carrier. J. Power Sources 195, 7524–7529. Dell'Era, A., Lupi, C., Pasquali, M., 2008. Electrochemical purification of cobalt polluted effluents, REWAS 2008: Global Symposium on Recycling, Waste Treatment and Clean Technology. In: Mishra, B., Ludwig, C., Das, S. (Eds.), pp. 1413–1418. Dutra, A.J.B., Espinola, A., Borges, P.P., 2000. Cadmium removal from diluted aqueous solutions by electrowinning in a flow by cell. Miner. Eng. 13 (10), 1139–1148. Friedrich, J.M., Ponce-de-Leon, C., Reade, G.W., Walsh, F.C., 2004. Reticulated vitreous carbon as an electrode material. J. Electroanal. Chem. 561, 203–217. Lanza, M.R.V., Bertazzoli, R., 2000. Removal of Zn(II) from chloride medium using a porous electrode: current penetration within the cathode. J. Appl. Electrochem. 30, 61–70. Leaist, D.G., 1989. Tracer diffusion of cobalt(II), nickel(II), copper(II) and hydrogen ions in aqueous sodium sulphate solutions. Electrochim. Acta 34 (5), 651–655. Lupi, C., Pasquali, M., Dell'Era, A., 2005. Nickel and cobalt recycling from lithium-ion batteries by electrochemical processes. Waste Manag. 25, 215–220. Lupi, C., Pasquali, M., Dell'Era, A., 2006. Studies concerning nickel electrowinning from acidic and alkaline electrolytes. Miner. Eng. 19, 1246–1250. Lupi, C., Dell'Era, A., Pasquali, M., 2009. Nickel–cobalt electrodeposited alloys for hydrogen evolution in alkaline media. Int. J. Hydrogen Energy 34, 2101–2106. Lupi, C., Dell'Era, A., Pasquali, M., Imperatori, P., 2011. Composition, morphology, structural aspects and electrochemical properties of Ni–Co alloy coatings. Surf. Coat. Technol. 205, 5394–5399. Lupi, C., Dell'Era, A., Pasquali, M., 2013. In situ activation with Mo of Ni–Co alloys 378 for hydrogen evolution reaction. Int. J. Hydrogen Energy 39, 1932–1940. Pasquali, M., Lupi, C., 2003. Electrolytic nickel recovery from lithium-ion batteries. Miner. Eng. 16, 537–542. Pistoia, G., Wiaux, J.P., Wolsky, S.P., 2001. Used battery collection and recycling. Industrial Chemistry Library. Elsevier, Amsterdam. Pletcher, D., Whyte, I., Walsh, F.C., Millington, J.P., 1991. Reticulated vitreous carbon cathodes for metal ion removal from process streams. Part I: mass transport studies. J. Appl. Electrochem. 21, 659–666.

8

A. Dell'Era et al. / Hydrometallurgy 150 (2014) 1–8

Podlaha, E.J., Fenton, J.M., 1995. Characterization of a flow-by RVC electrode reactor for the removal of heavy metals from dilute solutions. J. Appl. Electrochem. 25, 299–306. Polcaro, A.M., Palmas, S., Mascia, M., Renoldi, F., Correa dos Santos, R.L., 1999. Reticulated vitreous carbon electrode for copper(II) removal from dilute acid solutions: effect of the dissolved oxygen. Ind. Eng. Chem. Res. 38 (4), 1400–1405. Ponce de Leon, C., Pletcher, D., 1996. The removal of Pb(II) from acqueous solution using a reticulated vitreous carbon cathode cell-the influence of the electrolyte medium. Electrochim Acta 41 (4), 533–541. Ramalan, N.H.M., Karoonian, F.S., Etesami, M., Wen-Min, S., Hasnat, M.A., Mohamed, N., 2012. Impulsive removal of Pb(II) at a 3-D reticulated vitreous carbon cathode. Chem. Eng. J. 203, 123–129.

Tangirala, R.C., Low, C.T.J., Ponce-de-Leon, C., Campbell, S.A., Walsh, F.C., 2010. Copper deposition at segmented, reticulated vitreous carbon cathode in Hull cell. Trans. Inst. Met. Finish. 88 (2), 84–92. Tramontina, J., Azambuja, D.S., Piatnicki, C.M.S., 2002. Removal of Cd 2+ ion from diluted aqueous solutions by electrodeposition on reticulated vitreous carbon electrodes. J. Braz. Chem. Soc. 13, 4. http://dx.doi.org/10.1590/S0103-50532002000400010. Wang, L.K., Vaccari, D.A., Li, Y., Shammas, N.K., 2005. Chemical precipitation. Handbook of Environmental Engineering. Springer, New Jersey, pp. 141–197. Widner, R.C., Sousa, M.F.B., Bertazzoli, R., 1998. Electrolytic removal of lead using a flowthrough cell with a reticulated vitreous carbon cathode. J. Appl. Electrochem. 28, 201–207.