Recycling of spent lithium-ion battery cathode materials by ammoniacal leaching

Journal of Hazardous Materials 313 (2016) 138–146

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Recycling of spent lithium-ion battery cathode materials by ammoniacal leaching Heesuk Ku a , Yeojin Jung a , Minsang Jo a , Sanghyuk Park a , Sookyung Kim b , Donghyo Yang b,∗ , Kangin Rhee b , Eung-Mo An b , Jeongsoo Sohn b , Kyungjung Kwon a,∗ a b

Department of Energy & Mineral Resources Engineering, Sejong University, Seoul 05006, Republic of Korea Urban Mine Department, Korea Institute of Geoscience and Mineral Resources, 124 Gwahang-no, Yuseong-gu, Daejeon, Republic of Korea

g r a p h i c a l

a b s t r a c t

h i g h l i g h t s • • • • •

Ammoniacal leaching is used to recover spent Li-ion battery cathode materials. Leaching agents consist of ammonia, ammonium sulfite and ammonium carbonate. Ammonium sulfite is a reductant and ammonium carbonate acts as pH buffer. Co and Cu can be fully leached while Mn and Al are not leached. Co recovery via ammoniacal leaching is economical compared to acid leaching.

a r t i c l e

i n f o

Article history: Received 16 November 2015 Received in revised form 3 February 2016 Accepted 23 March 2016 Available online 31 March 2016

a b s t r a c t As the production and consumption of lithium ion batteries (LIBs) increase, the recycling of spent LIBs appears inevitable from an environmental, economic and health viewpoint. The leaching behavior of Ni, Mn, Co, Al and Cu from treated cathode active materials, which are separated from a commercial LIB pack in hybrid electric vehicles, is investigated with ammoniacal leaching agents based on ammonia, ammonium carbonate and ammonium sulfite. Ammonium sulfite as a reductant is necessary to enhance leaching kinetics particularly in the ammoniacal leaching of Ni and Co. Ammonium carbonate can act as a pH buffer

∗ Corresponding authors. E-mail addresses: [email protected] (D. Yang), [email protected], [email protected] (K. Kwon). http://dx.doi.org/10.1016/j.jhazmat.2016.03.062 0304-3894/© 2016 Elsevier B.V. All rights reserved.

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146 Keywords: Lithium ion battery Cathode Ammoniacal leaching Recycling Cobalt

139

so that the pH of leaching solution changes little during leaching. Co and Cu can be fully leached out whereas Mn and Al are hardly leached and Ni shows a moderate leaching efficiency. It is confirmed that the cathode active materials are a composite of LiMn2 O4 , LiCox Mny Niz O2, Al2 O3 and C while the leach residue is composed of LiNix Mny Coz O2 , LiMn2 O4 , Al2 O3 , MnCO3 and Mn oxides. Co recovery via the ammoniacal leaching is believed to gain a competitive edge on convenitonal acid leaching both by reducing the sodium hydroxide expense for increasing the pH of leaching solution and by removing the separation steps of Mn and Al. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Lithium-ion batteries (LIBs) offer several advantages including high power density, high energy density, high potential, long storage life, low self-discharge rate, and a wide operating temperature range [1]. The LIBs are widely used in mobile phones, laptops, video cameras, and other modern life appliances [2]. They now become more important because of usage as electric vehicle power sources [3]. The world’s production of LIBs reached 2.05 billion in 2005 and 5.86 billion in 2012 respectively [4]. It is anticipated that the production and consumption of LIBs will increase steadily in the forthcoming years. The LIBs, of which main components are anode, cathode, electrolyte, and separator, consist of lithium transition metal oxides, organic chemicals, carbons and polymers [1]. While the anode is basically a composite of carbon powder and polymer binder, the cathode is a composite of conducting carbon, polymer binder, and lithium transition metal oxides such as LiCoO2 , LiMn2 O4 , LiNiO2 , and LiCox Mny Niz O2 . These anode and cathode active materials are respectively coated on Cu and Al foils, which act as current collectors [5,6]. If spent LIBs are simply disposed by dumping them in landfill, soil contamination resulting from the leakage of organic electrolytes as well as heavy metals in the batteries will become a serious environmental concern [7]. Aside from their potentially hazardous nature, valuable materials from spent LIB wastes need to be recycled with considering their limited resources. Therefore, the recovery of the major components in the spent LIBs appears beneficial to prevent environmental pollution and raw material consumption [8,9]. From an environmental and health viewpoint, hydrometallurgical treatment could be a favored technology for recycling of metals from LIBs compared with pyrometallurgical processes because it offers such advantages as low energy consumption, no air emissions, and complete recovery of valuable components with high purity [4,10]. It is widely known that valuable metals such as Ni, Mn, Co, Al and Cu from the spent LIBs can be leached out by using acids such as hydrochloric acid and sulfuric acid. In particular, Co is the most extensively studied metal for recycling because of its relatively high price. Furthermore, separation of materials containing Co is particularly important from an environmental and health viewpoint as Co is classified as carcinogenic, mutagenic and toxic to reproduction [6]. The majority of Co can be leached out with hydrochloric acid or sulfuric acid and its leaching efficiency increases further in the presence of reducing agent such as hydrogen peroxide [1,5,8,11,12]. However, this acid leaching incurs an extra cost for regulating the pH of leaching solution to separate Co from Ni and Mn as a precipitated form. For example, when sodium hydroxide was sequentially added to an acidic leach liquor containing Ni, Mn and Co, Wang et al. observed that the precipitation of Ni, Mn and Co began at a pH value of 2, 1 and 3 and was completed at a pH value of 8, 12 and 10, respectively [13]. Therefore, a large amount of sodium

hydroxide would be necessary to increase the pH from a very low value to as high as 12. Moreover, the selective precipitation of each metal becomes difficult because the pH range for metal precipitation is overlapped between Ni, Mn and Co. Therefore, the spent LIBs leached with acidic medium undergo various steps of separation and recovery techniques such as solvent extraction, selective precipitation and electrochemical method [2,5,13–17]. For example, Chen et al. separated and recovered Mn, Ni, Cu and Co in a series of solvent extraction and precipitation steps. Mn and Ni were respectively precipitated with KMnO4 , NaOH and Na3 PO4 while Cu and Co were respectively extracted with Mextral® 5640H and Mextral® 272P [15]. By contrast, alkaline leaching employing ammoniacal solutions based on ammonia, ammonium carbonate, ammonium chloride or other alkaline reagents is known to show different leaching behavior for Ni, Mn, Co and Cu from acidic leaching. Bhuntumkomol et al. investigated the leaching behavior of nickel oxides in acid and ammoniacal solutions [18]. The leaching behavior of Mncontaining ores with ammoniacal medium was also studied by some research groups [19–23]. Senanayake et al. compared the leaching behavior of Zn-Mn-C batteries between ammoniacal and sulfuric acid solutions [24]. They attributed the higher and lower leaching efficiency of Zn and Mn in ammoniacal solutions to the formation of soluble Zn(NH3 )4 2+ and the inability of Mn to form a soluble complex ion, respectively. Rokukawa carried out extraction of Ni, Co, Cu, Mn, and Fe from cobalt crusts and ocean manganese nodules, and obtained the highest leaching efficiency of Ni, Co and Cu using combined ammonium carbonate and ammonium sulfite as leaching agent [20,25]. In this paper, the leaching behavior of Ni, Mn, Co, Al and Cu is investigated with ammoniacal medium for cathode active materials, which are separated from a commercial LIB pack in hybrid electric vehicles. We adopted ammoniacal leaching agents based on ammonia, ammonium carbonate and ammonium sulfite. The effects of leaching agent composition, leaching time and temperature are examined herein.

2. Experimental 2.1. Materials and methods LIB cathode active materials were obtained from a commercial, spent LIB pack in hybrid electric vehicles (Hyundai Motor Company). The flowchart of treatment process of spent LIBs is illustrated in Fig. 1. The physical treatment procedure involves discharging, dismantling, separating, drying, crushing, grinding, sieving and grain size separation in sequence. After a spent LIB pack was connected to a discharger, the pack was discharged to less than 0.1 V, and then, the pack proceeded with the dismantling process. Dismantled cells from the pack were forced to be short-circuited to release residual electrical charge. The discharging step is necessary before dismantling to avoid the potential danger of short-circuit or self-ignition of battery rolls when anode and cathode are put

140

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146 Table 1 The chemical composition of treated cathode materials. Elements

Ni

Mn

Co

Content (wt%)

15.3

14.3

6.0

Al 2.3

Cu

Etc.

0.3

61.8

Fig. 2. Effect of ammonia concentration from 1 to 3 M on leaching efficiency of the metals (80 ◦ C, 300 rpm and 1 h leaching).

(15.3 wt% Ni, 14.3 wt% Mn, 6.0 wt% Co, 2.3 wt% Al and 0.3 wt% Cu) as summarized in Table 1. The chemical treatment procedure used ammoniacal solutions to dissolve the valuable metals. Reagent grade ammonia solution (28 wt%), ammonium carbonate and ammonium sulfite were purchased from Junsei Chemical Co., Ltd. A leaching agent containing ammonia, ammonium sulfite and ammonium carbonate was prepared. Temperature of a flask reactor containing 500 ml leaching agent was fixed by using a heating mantle with a condenser and an agitator of which rotating speed was 300 rpm. When the leaching agent was heated to a setting temperature, 5 g of the LIB cathode powders was added into the flask reactor. The solid-liquid density was fixed at 10 g/l. After leaching experiments, the leaching solution was filtered and Ni, Mn, Co, Al and Cu content was analyzed by ICP. The pH of leaching solution was monitored during leaching experiments. 2.2. Material characterization

Fig. 1. Flowchart of the treatment process of spent LIBs.

in contact with each other in the dismantling course [26]. After the LIB cells were dismantled, anode and cathode active materials and separators were separated. The cathode active materials were first crushed with a shredder, which reduced the size of cathode to 40 mm × 50 mm. The shredded cathode was then ground with grinders such as pin mill leading to the resulting particle size of treated cathode materials being −65 mesh. The chemical composition of the majority of treated cathode materials is carbon as a conducting material and, in addition, some valuable metals exist

After the treated cathode materials were dissolved in aqua regia, the chemical composition of treated cathode materials was analyzed by an inductively coupled plasma (ICP, ICAP 6500). Crystal structure, morphology and chemical composition of leach residues were characterized by X-ray diffraction (XRD, X’pert MPD), scanning electron microscope (SEM, 6380, Jeol Ltd, Japan) and energy dispersive spectroscopy (EDS, JEM 2300, Jeol Ltd, Japan), respectively. 3. Results and discussion 3.1. Optimization of leaching agent composition Fig. 2 shows the effect of ammonia concentration on the leaching efficiency of Ni, Mn, Co, Al, and Cu when a leaching agent was an ammonia solution. First of all, Ni, Mn and Co, which are the three major metals and of commercial interest, were hardly leached out

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146

Fig. 3. Effect of (a) ammonium sulfite and (b) ammonium carbonate concentration from 0 to 4 M on the leaching efficiency of the metals (1 M ammonia solution, 80 ◦ C, and 1 h leaching).

in the leaching agent based on single solute component ammonia up to its concentration of 3 M. By contrast, Al was more or less leached out in the single component leaching agent with Cu, of which leaching behavior can be predicted in a pourvaix diagram of Cu in the presence of NH3 [27], totally leached out in 3 M ammonia solution. However, it is certain that another component is necessary in the leaching agent to leach more economically important metals such as Co and Ni. The ammonia concentration, for convenience sake, was fixed at 1 M for the further optimization of leaching agent composition. The effect of addition of ammonium sulfite or ammonium carbonate to ammonia solution was investigated. First, the effect of ammonium sulfite concentration in ammonia solution on the leaching efficiencies of Ni, Mn, Co, Al and Cu is shown in Fig. 3a. When we compare the leaching efficiency of each metal in the absence and presence (0.5 M) of ammonium sulfite, a change in leaching efficiency is most notable for Ni and Co. The leaching effi-

141

ciency of Ni and Co increased from 0 to 26% and 78% respectively on the addition of ammonium sulfite. Whereas the leaching efficiency of Cu was further enhanced from 73 to 100%, the leaching efficiency of Mn and Al were invariably very low in the presence of ammonium sulfite. Ammonium sulfite seems indispensable particularly in the ammoniacal leaching of Ni and Co. Reductants such as ammonium sulfite, ammonium thiosulfate and sodium sulfite were reported to be necessary, and particularly sulfite-containing reductants are known to be effective in the ammoniacal leaching [19–23,25,28]. These reductants reduce the oxidation number of Ni and Co in their oxide form, leading to the enhancement of dissolution rate. Although the dissolution of higher valence state Ni and Co in their oxides is thermodynamically favorable in ammoniacal solution, the reaction kinetics is very slow [27]. The optimum amount of ammonium sulfite is around 0.5 M for Ni and Co on the basis of 1 M ammonia. The decrease in leaching efficiency of Ni and Co beyond 1 M ammonium sulfite might be related to the deviation from pH range for forming complex ions with NH3 . Das et al. argued that by increasing the concentration of reductant, precipitation of the hydroxide is facilitated, resulting in the formation of complex and subsequent co-precipitation of Ni and Co, thereby decreasing the overall dissolution of metals as soluble amines [19]. Therefore, it was concluded that the pH of leaching solutions need to be maintained in an optimum range for stable ammonia complex ions. Second, we varied ammonium carbonate concentration in a leaching agent where the concentrations of ammonia was fixed to 1 M. Because ammonium carbonate has the same kind and number of cation (two NH4 + ) and a different anion (CO3 2− ) that cannot act as a reducing agent from ammonium sulfite, a comparison of leaching behavior in ammonia solutions containing the same concentration of ammonium carbonate or ammonium sulfite would make the role of ammonium sulfite as a reductant more clear with an identical pH change. The leaching efficiency of Ni, Mn, Co, Al and Cu depending on ammonium carbonate concentration is illustrated in Fig. 3b. When we compare the leaching efficiency of each metal in the presence of ammonium carbonate or ammonium sulfite at low concentration, a difference in leaching efficiency of Ni and Co is the most remarkable. For example, the leaching efficiency of Ni and Co was 8 and 14% with 0.5 M ammonium carbonate while that of Ni and Co was 26 and 78% with 0.5 M ammonium sulfite respectively. By contrast, the leaching efficiency of Ni and Co started to converge to ∼20 and ∼45% respectively as the concentration of ammonium carbonate or ammonium sulfite increased. On the other hand, the other metals exhibited the same leaching behavior regardless of the presence of ammonium carbonate or ammonium sulfite. Therefore, the necessity of reductant such as ammonium sulfite for the enhancement of leaching efficiency of Ni and Co in ammoniacal leaching can be supported by the comparison of leaching behavior in Fig. 3a and b. To find the optimum composition of leaching agents, both ammonium sulfite and ammonium carbonate were added to an 1 M ammonia solution because ammonium sulfite and ammonium carbonate showed different optimum concentration for the maximum leaching efficiency of Ni and Co when they were used separately. That is, ammonium sulfite had the optimum concentration of 0.5 M whereas the leaching efficiency of Ni and Co kept increasing as the concentration of ammonium carbonate increased. Thus, we varied ammonium carbonate concentration in a leaching agent where the concentrations of ammonia and ammonium sulfite were fixed to 1 and 0.5 M respectively. The leaching efficiency of Ni, Mn, Co, Al and Cu depending on ammonium carbonate concentration is illustrated in Fig. 4a. Although there are some fluctuation for Ni and Co, the leaching efficiency of all the metals maintains the maximum leaching efficiency obtained with the binary (ammonia + ammonium sulfite) system regardless of the amount of ammonium carbon-

142

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146

ate. Therefore, the optimum composition of leaching agent would be either binary (ammonia (1 M) + ammonium sulfite (0.5 M)) or ternary (ammonia (1 M) + ammonium sulfite (0.5 M) + ammonium carbonate (0 ∼ 2 M)). Despite no further enhancement of leaching efficiency on the addition of ammonium carbonate, the ternary system could be advantageous considering the insensitiveness to a sudden pH change that could lead to an unfavorable condition for soluble complex ion formation. The relation between the base NH3 and its conjugate acid NH4 + from a pH buffer can be expressed as follows: 4NH3 + H+ = NH4 +

(1)

Therefore, the pH of an ammonia-ammonium system at room temperature is given by [19] pH = 9.26 + log([NH3 ]/[NH4 + ])

(2)

Because the pH of ammoniacal solution depends on the ratio of NH3 to NH4 + , the initial addition of ammonium sulfite alters substantially the pH of binary (ammonia + ammonium sulfite) system. However, once there is the third component containing NH4 + in an ammoniacal solution, the pH of ternary (ammonia + ammonium sulfite + ammonium carbonate) system does not change significantly. Namely, ammonium carbonate can act as a pH buffer so that the pH of leaching solution changes little during leaching for the formation of stable Ni, Co and Cu ammonia complexes [19,27–31,33]. To confirm the pH buffering effect of ammonium carbonate, the pH change of the binary or ternary system was monitored as in Fig. 4b. Measured pH values at room temperature were well matched with theoretical pH values at 25 ◦ C according to Eq. (2) (a pH value of the binary system with 0 M ammonium sulfite cannot be calculated). The pH of solution is usually shifted to a lower value as temperature increases, and indeed the pH of the binary or ternary system was lowered by approximately 1 when temperature changed from room temperature to 80 ◦ C. When the same concentration (0.25 M) of ammonium sulfite or ammonium carbonate was initially added to the binary or ternary system respectively, a pH change in the binary system was 2.0 whereas that in the ternary system was 0.2. Because the stable pH range for Ni(NH3 )6 2+ , Co(NH3 )6 2+ and Cu(NH3 )4 2+ ions is in between 8 and 10 for leaching of Ni-Co-Cu system referring to potential-pH diagrams for Co/Ni/Cu-NH3 -H2 O system at 25 ◦ C [27,32], the pH-buffered ternary system would have an advantage over the binary system for maintaining the optimum pH range. The leaching behavior of each metal can be explained as follows when the pH of leaching solution is in the favorable range for Ni, Co and Cu complex ions. Once Ni, Co and Cu form low valence ions such as a divalent cation, they can exist as complex ions with NH3 in ammoniacal solutions. If an excess of ammonia is added to a solution containing Ni2+ , Co2+ and Cu2+ , complex ions such as Ni(NH3 )6 2+ , Co(NH3 )6 2+ and Cu(NH3 )4 2+ are formed as given by the following reactions [27,34]. Ni2+ + 6NH3 ↔ Ni(NH3 )6 2+ Co2+ + 6NH3 ↔ Co(NH3 )6 2+ Cu2+ + 4NH3 ↔ Cu(NH3 )4 2+

Fig. 4. (a) Effect of ammonium carbonate concentration from 0 to 2 M on the leaching efficiency of the metals (1 M ammonia solution, 0.5 M ammonium sulfite, 80 ◦ C, and 1 h leaching), (b) pH change of the binary (1 M ammonia solution + ammonium sulfite) or ternary (1 M ammonia solution + 0.5 M ammonium sulfite + ammonium carbonate) system, (C) Effect of total amount of leaching agents on the leaching efficiency of the metals (ammonia:ammonium sulfite:ammonium carbonate = 1:0.5:1, 80 ◦ C, and 1 h leaching).

Formation constants for the complexes of Ni(NH3 )6 2+ , Co(NH3 )6 2+ and Cu(NH3 )4 2+ are given as 5.5 × 108 , 1.3 × 105 and 1.1 × 1013 at 25 ◦ C, respectively [34]. As mentioned in Section 2, the spent LIB pack was fully discharged before dismantling and the oxidation state of Ni and Co should be higher than 2+. Therefore, it is difficult to predict the leaching efficiency of Ni and Co from the values of formation constants for Ni(NH3 )6 2+ and Co(NH3 )6 2+ . However, we can expect from the high values of the

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146

(a)

143

(a) 0.12

80

Ni Mn Co Al Cu

60 40

0.06 0.04

0.00

0

20

40

60

80

100

120

Time (min)

0

5

100

0.40

60

20

25

30

20

25

30

o

0.30

1-(1-X)1/3

80

15

20 C o 40 C o 60 C o 80 C

0.35

Ni Mn Co Al Cu

10

Time (min)

(b)

(b)

Leaching efficiency (%)

0.08

0.02

20 0

o

20 C o 40 C o 60 C o 80 C

0.10

1-(1-X)1/3

Leaching efficiency (%)

100

0.25 0.20 0.15 0.10

40

0.05

20 0 20

0.00

30

40

50

60

70

80

5

10

15

Time (min)

(c)

Leaching temperature (oC)

0

Co Ni

-1

Fig. 5. (a) Effect of leaching time from 0 to 120 min and (b) leaching temperature from 20 to 80 ◦ C on the leaching efficiency of the metals (1 M ammonia solution, 0.5 M ammonium sulfite and 1 M ammonium carbonate).

-2 lnK

formation constants that a considerable portion of Ni and Co in LIB cathode can be leached out in the presence of the reducing agent in ammoniacal solutions. On the other hand, thermodynamics favors Mn to form corresponding hydroxides or oxides in an ammonia solution [27,35]. Although Mn can form unstable ammonia complexes, manganese oxide is converted to manganese carbonate through an intermediate manganese ammine complex in the presence of ammonium carbonate [20,22,25,32]. The effect of total amount of leaching agent on the leaching efficiency of the metals is presented in Fig. 4c where the ratio between ammonia, ammonium sulfite and ammonium carbonate is fixed at 1:0.5:1. The leaching efficiency of the metals at 1 M ammonia in Fig. 4c corresponds to that at 1 M ammonium carbonate in Fig. 4a. As the amount of leaching agent was increased to an occasion of 3 M ammonia, 1.5 M ammonia sulfite and 3 M ammonium carbonate, the leaching efficiency of Ni and Co reached 37 and 94%, respectively. Although it is expected that Co can be fully leached out by using a higher concentration of leaching agent with the optimum composition, Ni is unlikely to be fully dissolved with the current leaching agent. On the other hand, Mn and Al show a negligible leaching efficiency regardless of the concentration of leaching agent. This selective leaching behavior of Co in contrast to Mn and Al could enable more economical recovery of Co from spent LIB cathode materials by removing separation steps of Mn and Al.

0

-3 -4 -5 -6

2.8

2.9

3.0

3.1

3.2

3.3

3.4

3.5

1/(T+273.15)*1000(K-1) Fig. 6. Conversion-time relation based on the reaction controlled shrinking core model for (a) Ni and (b) Co leaching and (c) Arrhenius plot.

3.2. Effect of leaching time and temperature The time dependency of leaching efficiency of the metals is given in Fig. 5a. Ni and Co showed relatively slow leaching kinetics and their leaching efficiency reached a plateau after 40 min. On the other hand, the leaching of Cu proceeded rapidly, and was completed in less than 10 min. The leaching efficiency of Mn and Al was almost negligible from the start. Although there was slight room for the improvement of Co leaching efficiency after 2 h, 1 h seemed sufficient for the leaching of the treated cathode active materials considering that the leaching of Cu and Ni can be stabilized in 1 h. Fig. 5b shows the effect of leaching temperature on the leaching efficiency of the metals from 20 to 80 ◦ C. Ni and Co had an increasing tendency of leaching efficiency with temperature whereas the

144

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146

Fig. 7. SEM–EDS analysis: (a) low magnification SEM and elemental mapping, (b) high magnification SEM of treated cathode material powders, and (c) low magnification SEM and elemental mapping, (d) high magnification SEM of leach residue (leaching condition: 1 M ammonia, 0.5 M ammonium sulfite, 1 M ammonium carbonate at 80 ◦ C and 1 h leaching).

leaching efficiency of Mn, Al and Cu did not correlate with temperature. Because the leaching efficiency of Ni and Co increased with temperature, one might expect that a higher leaching temperature than 80 ◦ C can achieve a higher leaching efficiency. However, it was reported that the Co ammine complexes become unstable and the leaching efficiency rather decreases when the leaching temperature is above 100 ◦ C [19].

3.3. Dissolution kinetics for Ni and Co The dissolution kinetics behavior of Ni and Co were studied at various temperatures in the ternary (1 M ammonia + 0.5 M ammonium sulfite + 1 M ammonium carbonate) system. To determine the kinetics behavior of Ni and Co, we fitted leaching data in reaction rate models based on either ash diffusion control or reaction control. A reaction controlled shrinking core model has been usually adopted for the study of particle leaching [23,36], and our fitting results were more supportive of the reaction controlled shrinking core model than the ash diffusion controlled one. The conversiontime relation for the reaction controlled shrinking core model is shown in the following Eq. (3) [36]. 5 − (1 − X)1/3 = kt

(3)

where k is the reaction rate constant, t is the time, and X is the fraction of metal leached. Fig. 6a and b are the plots of 1 − (1 − X)1/3 vs. time for Ni and Co, respectively. Arrhenius plots were constructed in Fig. 6c using obtained reaction rate constants to compute activation energy of Ni and Co dissolution. The activation energy of Ni and Co was respectively estimated to be 57.4 and 60.4 kJ/mol, which are

similar to activation energy values that were reported previously [18,32].

3.4. Characterization of treated cathode material powders and leach residue Fig. 7a displays a SEM image of treated cathode material powders at low magnification and its corresponding EDS elemental mapping with a color bar where white and black indicate the highest and lowest content respectively. The elemental distribution in the SEM image has the following conspicuous features. First, Ni, Mn and Co are located in similar places, and particularly Ni and Co are positioned on the exactly same spots. Second, Al is located in different places from where Ni, Mn and Co exist. Lastly, Cu appears to exist in small quantity and to be distributed relatively uniformly over the specimen. Judging from these features and as confirmed by LG Chem who is the manufacturer of the LIB pack for this study, the cathode active materials are a composite of LiMn2 O4 and LiCox Mny Niz O2 . We chose three points (1, 2 and 3) in the SEM image for Al-rich, Mn-rich and Ni/Mn/Co-rich positions, and their morphology was examined at a higher magnification in Fig. 7b. Taking Table 2 where elemental information at each point is displayed into consideration, we could figure out main components occupying each point. An Al-rich phase with a plate shape at the point 1 should be Al current collector debris, and an irregular-shaped Mnrich phase at the point 2 would be LiMn2 O4 active materials. An Ni/Mn/Co-rich sphere with a diameter of ∼10 ␮m at the point 3 has a typical LiCox Mny Niz O2 active material morphology and similar Ni/Mn/Co content.

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146

145

originate from unleached components and Li-leached LiMn2 O4 while MnCO3 could form in the ternary system adopting ammonium carbonate as mentioned in the preceding section 3.1. 4. Conclusions

Fig. 8. X-ray diffraction patterns of (a) treated cathode materials, and (b) leach residue (1 M ammonia, 0.5 M ammonium sulfite, 1 M ammonium carbonate at 80 ◦ C and 1 h leaching).

Table 2 EDS elemental analysis of treated cathode material powders and leach residue. Element (wt%)

O Al Mn Co Ni Cu

Treated cathode material powders Leach residue Point 1

Point 2

Point 3

Point 1

Point 2

Point 3

45.29 41.66 2.19 – – 10.86

23.2 1.7 69.33 2.48 – 3.27

21.8 – 27.82 27.12 23.26 –

45.89 50.36 3.75 – – –

23.16 1.6 70.7 – – 4.54

22.1 0.56 29.84 25.14 22.36 –

Fig. 7c displays a SEM image of leach residue at low magnification and its corresponding EDS elemental mapping. The leaching efficiency of Ni, Mn, Co, Al and Cu was 25, 1, 80, 2, and 100% respectively under these leaching conditions. As can be confirmed from the leaching efficiency values, Ni and Co content in the leach residue were reduced while Mn appeared dominant over the specimen. Unexpectedly, Cu still existed scattered over the leach residue surface although the leaching efficiency was 100%. The existence of Cu could originate from adsorbed Cu on Mn oxide during filtering of leaching residue [37,38]. Three points (1, 2 and 3) in the SEM image was again chosen for Al-rich, Mn-rich and Ni/Mn/Co-rich positions in Fig. 7d. There was no substantial difference in morphology and composition between Fig. 7b and d except that the point 3 in Fig. 7d shows a more porous surface of spherical LiCox Mny Niz O2 particle due to the dissolution of Ni and Co on the surface. Incidentally, the presence of carbon in treated cathode material powders and leach residue was confirmed in SEM/EDS analyses (Figs. S1 and S2). It appears that the majority of carbon remains unleached during the leaching process, considering a little change in carbon peak intensity in the EDS spectrum. Fig. 8 presents XRD patterns of treated cathode material powders and leach residue obtained from leaching with the ternary (1 M ammonia + 0.5 M ammonium sulfite + 1 M ammonium carbonate) system. Overall, the XRD patterns support the characterization results of SEM-EDS analysis related to Fig. 7. The treated cathode material powders are composed of LiNix Mny Coz O2, LiMn2 O4 , Al2 O3 and C. Al2 O3 should be the Al-rich phase observed in the SEM-EDS analysis, and C might be a conducting agent for LIB cathode. On the other hand, the leach residue is composed of LiNix Mny Coz O2 , LiMn2 O4 , Al2 O3 , MnCO3 and Mn oxides. The existence of LiNix Mny Coz O2 , LiMn2 O4 , Al2 O3 and Mn oxides should

The leaching behavior of Ni, Mn, Co, Al and Cu from treated cathode active materials, which were separated from a commercial LIB pack in hybrid electric vehicles, was investigated with ammoniacal leaching agents based on ammonia, ammonium carbonate and ammonium sulfite. Ammonium sulfite as a reductant is necessary to enhance leaching kinetics particularly in the ammoniacal leaching of Ni and Co. Ammonium carbonate can act as a pH buffer so that the pH of leaching solution changes little during leaching. The ternary system could be advantageous considering the insensitiveness to a sudden pH change that could lead to an unfavorable condition for soluble complex ion formation. Co and Cu can be fully leached out by using a leaching agent with the optimum composition whereas Mn and Al were hardly leached and Ni showed a moderate leaching efficiency. The leaching of Ni and Co was completed in 40 min, and the leaching of Cu proceeded rapidly reaching 100% leaching efficiency in less than 10 min. The leaching behavior of Ni and Co appeared to follow the reaction controlled shrinking core model with the activation energy of Ni and Co being 57.4 and 60.4 kJ/mol respectively. The SEM-EDS and XRD analyses of treated cathode material powders indicated that the cathode active materials are a composite of LiMn2 O4 , LiCox Mny Niz O2, Al2 O3 and C while the leach residue is composed of LiNix Mny Coz O2 , LiMn2 O4 , Al2 O3 , MnCO3 and Mn oxides. Co recovery via the ammoniacal leaching is believed to gain a competitive edge on the acid leaching both by reducing the sodium hydroxide expense for increasing the pH of leaching solution and by removing the separation steps of Mn and Al. Acknowledgments This study was supported by the R&D Center for Valuable Recycling(Global-Top Environment Technolgy Development Program) funded by the Ministry of Environment (Project No.:GT14-C-01-036-0). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jhazmat.2016.03. 062. References [1] X. Zhang, Y. Xie, X. Lin, H. Li, H. Cao, An overview on the processes and technologies for recycling cathode active materials from spent lithium-ion batteries, J. Mater. Cycles Waste Manage. 15 (2013) 420–430. [2] A.S.W. Md. Al Hossaini Shuva, Kurny, Hydrometallurgical recovery of value metals from spent lithium ion batteries, Am. J. Mater. Eng. Technol. 1 (2013) 8–12. [3] V. Etacheri, R. Marom, R. Elazari, G. Salitra, D. Aurbach, Challenges in the development of advanced Li-ion batteries: a review, Energy Environ. Sci. 4 (2011) 3243–3262. [4] L. Li, W. Qu, X. Zhang, J. Lu, R. Chen, F. Wu, K. Amine, Succinic acid-based leaching system: a sustainable process for recovery of valuable metals from spent Li-ion batteries, J. Power Sources 282 (2015) 544–551. [5] D.A. Ferreira, L.M.Z. Prados, D. Majuste, M.B. Mansur, Hydrometallurgical separation of aluminium cobalt, copper and lithium from spent Li-ion batteries, J. Power Sources 187 (2009) 238–246. [6] A. Chagnes, B. Pospiech, A brief review on hydrometallurgical technologies for recycling spent lithium-ion batteries, J. Chem. Technol. Biotechnol. 88 (2013) 1191–1199. [7] L. Li, J. Lu, Y. Ren, X.X. Zhang, R.J. Chen, F. Wu, K. Amine, Ascorbic-acid-assisted recovery of cobalt and lithium from spent Li-ion batteries, J. Power Sources 218 (2012) 21–27.

146

H. Ku et al. / Journal of Hazardous Materials 313 (2016) 138–146

[8] X. Zeng, J. Li, N. Singh, Recycling of spent lithium-ion battery: a critical review, Crit. Rev. Env. Sci. Technol. 44 (2014) 1129–1165. [9] X. Zeng, J. Li, Innovative application of ionic liquid to separate Al and cathode materials from spent high-power lithium-ion batteries, J. Hazard. Mater. 271 (2014) 50–56. [10] J. Li, G. Wang, Z. Xu, Environmentally-friendly oxygen-free roasting/wet magnetic separation technology for in situ recycling cobalt, lithium carbonate and graphite from LiCoO2/graphite lithium batteries, J. Hazard. Mater. 302 (2016) 97–104. [11] S.M. Shin, N.H. Kim, J.S. Sohn, D.H. Yang, Y.H. Kim, Development of a metal recovery process from Li-ion battery wastes, Hydrometallurgy 79 (2005) 172–181. [12] C.K. Lee, K.I. Rhee, Preparation of LiCoO2 from spent lithium-ion batteries, J. Power Sources 109 (2002) 17–21. [13] R.C. Wang, Y.C. Lin, S.H. Wu, A novel recovery process of metal values from the cathode active materials of the lithium-ion secondary batteries, Hydrometallurgy 99 (2009) 194–201. [14] M. Joulie, R. Laucournet, E. Billy, Hydrometallurgical process for the recovery of high value metals from spent lithium nickel cobalt aluminum oxide based lithium-ion batteries, J. Power Sources 247 (2014) 551–555. [15] X. Chen, B. Xu, T. Zhou, D. Liu, H. Hu, S. Fan, Separation and recovery of metal values from leaching liquor of mixed-type of spent lithium-ion batteries, Sep. Purif. Technol. 144 (2015) 197–205. [16] M. Contestabile, S. Panero, B. Scrosati, A laboratory-scale lithium-ion battery recycling process, J. Power Sources 92 (2001) 65–69. [17] P. Zhang, T. Yokoyama, O. Itabashi, Y. Wakui, T.M. Suzuki, K. Inoue, Hydrometallurgical process for recovery of metal values from spent nickel-metal hydride secondary batteries, Hydrometallurgy 50 (1998) 61–75. [18] K. Bhuntumkomol, K.N. Han, F. Lawson, The leaching behavior of nickel oxides in acid and in ammoniacal solutions, Hydrometallurgy 8 (1982) 147–160. [19] R.P. Das, S. Anand, S.C. Das, P.K. Jena, Leaching of manganese nodules in ammoniacal medium using glucose as reductant, Hydrometallurgy 16 (1986) 335–344. [20] N. Rokukawa, Extraction of nickel, cobalt and copper from ocean manganese nodules with mixed solution of ammonium carbonate and ammonium sulfite, Shigen-to-Sozai 106 (1990) 2005–2209. [21] M. Niinae, N. Komatsu, Y. Nakahiro, T. Wakamatsu, J. Shibata, Preferential leaching of cobalt, nickel and copper from cobalt-rich ferromanganese crusts with ammoniacal solutions using ammonium thiosulfate and ammonium sulfite as reducing agents, Hydrometallurgy 40 (1996) 111–121. [22] R.K. Jana, D.S.R. Murthy, A.K. Nayak, M.S. Mahanty, S.K. Tiwary, D.D. Akerkar, Leaching of roast-reduced polymetallic sea nodules to optimize the recoveries of copper, nickel and cobalt, Int. J. Miner. Process 30 (1990) 127–141.

[23] K.S. Choi, O.H. Han, B.S., Shin Reduction leaching of manganese nodules with sodium sulfite in ammonium chloride solution, Shigen-to Sozai 112 (1996) 117–121. [24] G. Senanayake, S.M. Shin, A. Senaputra, A. Winn, D. Pugaev, J. Avraamides, J.S. Sohn, D.J. Kim, Comparative leaching of spent zinc-manganese-carbon batteries using sulfur dioxide in ammoniacal and sulfuric acid solutions, Hydrometallurgy 105 (2010) 36–41. [25] N. RoKukawa, Extraction of nickel, cobalt and copper from ocean cobalt crusts with ammoniacal alkaline solution, Shigen-to-Sozai 108 (1992) 187–191. [26] J. Nan, D. Han, X. Zuo, Recovery of metal values from spent lithium-ion batteries with chemical deposition and solvent extraction, J. Power Sources 152 (2005) 278–284. [27] K.N. Han, M. Hoover, D.W. Fuerstenau, Ammonia-ammonium leaching of deep-sea manganese nodules, Int. J. Miner. Process 1 (1974) 215–230. [28] S. Acharya, Reductive ammonia leaching of manganese nodules by thiosulfate, Metall. Trans. B 22B (1991) 259–261. [29] M. Miyake, M. Maeda, Dissolutino of nickel hydroxide in ammoniacal aqueous solutions, Metall. Mater. Trans. B 37B (2006) 181–188. [30] D. Mishra, R.R. Srivastava, K.K. Sahu, T.B. Singh, R.K. Jana, Leaching of roast-reduced manganese nodules in NH3 -(NH4 )2 CO3 medium, Hydrometallurgy 109 (2011) 215–220. [31] D. Bingol, M. Canbazoglu, S. Aydogan, Dissolution kinetics of malachite in ammonia/ammonium carbonate leaching, Hydrometallurgy 76 (2005) 55–62. [32] C. Vu, K.N. Han, F. Lawson, Leaching behavior of cobaltous and cobalto-cobaltic oxides in ammonia and in acid solutions, Hydrometallurgy 6 (1980) 75–87. [33] K. Osseo-Asare, J.W. Lee, H.S. Kim, H.W. Pickering, Cobalt extraction in ammoniacal solution: electrochemical effect of metallic iron, Mater. Trans. B 14B (1983) 571–576. [34] R.A. Kumbasar, S. Kasap, Selective separation of nickel from cobalt in ammoniacal solutions by emulsion type liquid membranes using 8-hydroxyquinoline (8-HQ) as mobile carrier, Hydrometallurgy 95 (2009) 121–126. [35] A. Katsiapi, P.E. Tasakiridis, P. Oustadakis, S. Agatzini-Leonardou, Cobalt recovery from mixed Co-Mn hydroxide precipitates by ammonia-ammonium carbonate leaching, Mier. Eng. 23 (2010) 643–651. [36] M.K. Jha, A. Kumari, A.K. Jha, V. Kumar, J. Hait, B.D. Pandey, Recovery of lithium and cobalt from waste lithium ion batteries of mobile phone, Waste Manage. 33 (2013) 1890–1897. [37] R.N. Sahoo, S.C. Das, B.R. Reddy, P.C. Rath, R.P. Das, Adsorption of copper on manganese residue obtained from NH3 -SO2 leaching, Hydrometallurgy 62 (2001) 185–192. [38] N. Das, R.K. Jana, Adsorption of some bivalent heavy metal ions from aqueous solutions by manganese nodule leached residues, J.Colloid Interface Sci. 293 (2006) 253–262.