Articles

https://doi.org/10.1038/s41560-019-0368-4

Deep eutectic solvents for cathode recycling

of Li-ion batteries
Mai K. Tran   , Marco-Tulio F. Rodrigues   , Keiko Kato, Ganguli Babu   * and Pulickel M. Ajayan   *

As the consumption of lithium-ion batteries (LIBs) for the transportation and consumer electronic sectors continues to grow, so
does the pile of battery waste, with no successful recycling model, as exists for the lead–acid battery. Here, we exhibit a method
to recycle LIBs using deep eutectic solvents to extract valuable metals from various chemistries, including lithium cobalt (iii)
oxide and lithium nickel manganese cobalt oxide. For the metal extraction from lithium cobalt (iii) oxide, leaching efficiencies of
≥90% were obtained for both cobalt and lithium. It was also found that other battery components, such as aluminium foil and
polyvinylidene fluoride binder, can be recovered separately. Deep eutectic solvents could provide a green alternative to conven-
tional methods of LIB recycling and reclaiming strategically important metals, which remain crucial to meet the demand of the
exponentially increasing LIB production.

W
ith the ever-increasing demand for energy being fulfilled
by rechargeable lithium-ion batteries (LIBs), the acceler-
ated production of these devices will soon translate into
massive amounts of waste. Because LIB lifetimes barely surpass
3 years in smaller electronics and 5–10 years in electric vehicles, large
numbers of spent LIBs are expected to be generated, with 500,000
metric tonnes predicted from China alone by 20201,2. The develop-
ment of LIB-specific recycling methods can minimize the impact of
this waste and also alleviate constraints on the supply chain of bat-
tery manufacturers. Cobalt, for example, is a strategic material that
typically constitutes up to 15 wt% of LIB cathodes. However, it is
relatively expensive and is mainly sourced from potentially corrupt
conflict zones in what are not always reliable areas in the world3.
Cobalt in large accumulated quantities is also of health and environ-
mental concerns3,4. Although the efficient recovery of raw materi-
als from spent batteries will be economically and environmentally
advantageous, existing technologies still face practical problems.
Most of the recoverable value in spent batteries is in the cathode.
Common approaches to extracting the metal contained in the active
lithium oxides are either based on pyrometallurgy1, hydrometal-
lurgy1,5, biometallurgy1, mechanical methods6,7 or some combina-
tion of the four1,8–10. Among these, pyrometallurgy reigns supreme
in industry despite the high energetic cost from the extreme tem-
peratures (1,400 °C or higher), and rampant off-gassing of harm-
ful fumes, which require drastic safety precautions as well as the
scrubbing of infrastructure to reduce subsequent pollution1,8.
Furthermore, the resulting mixed slag makes it nearly impossible
to fully recover all of the metals, such as lithium, using this method
alone8. Hydrometallurgy is one of the most viable options, due to its
high metal leaching rate and purity of the recovered product1,2,10,11.
Unfortunately, even this technique usually involves caustic reagents
such as hydrochloric, nitric and sulfuric acids, which pose a danger
to workers and the environment2. ‘Green solvents’, such as organic
acids reported in the literature, show promise in addressing the
environmental safety concern by utilizing non-hazardous materi-
als, but face their own set of challenges. For example, malic acid
has been shown to leach out divalent metal ions, but requires an
additional reducing agent to accelerate the process for dissolving the
higher-valence ions typically found in LIBs12. In contrast, oxalic acid
has been shown to have some ability as both a leaching and reduc-
ing agent, but faces the problem of requiring a high-temperature
pyrolysis step because of its inability to fully dissolve all of the cath-
ode active materials in their original form13,14.
In the present work, we propose an alternative and highly effi-
cient concept to extract value from LIB waste using a deep eutectic
solvent (DES). While this particular DES is also a ‘green solvent’, it
can act as both an effective leaching and reducing agent, eliminating
the need for supporting chemicals and processes to achieve metal
extraction. We further demonstrate that this DES is capable of tack-
ling multiple LIB chemistries, including lithium cobalt (iii) oxide
(LCO; chemical formula LiCoO2) and the increasingly relevant
lithium nickel manganese cobalt oxide (NMC; chemical formula
LiNi1/3Mn1/3Co1/3O2)3,10.
The green solvent approach
DESs are a class of compounds that generally present an unusu-
ally high capability of dissolving metal oxides15–18. These solvents
are eutectic mixtures of hydrogen bond donors and acceptors that
are often cheaply obtained, easily prepared and made of relatively
nontoxic and biodegradable compounds19–22. Multiple reports have
demonstrated DESs’ ability to digest common metal oxides15,17,23–25,
with certain compositions even rivalling the performance of acids18.
A major advantage of using DESs is that there is no additional need
for an extra step involving a reducing agent and/or expensive sol-
vent extractants, which are commonly added to many conventional
hydrometallurgical procedures13. Here, we extend these virtues to
the universe of LIBs, and provide a proof of concept for the use
of these sustainable solvents as vehicles to enable the recycling of
spent LIBs.
The proposed recycling scheme is depicted in Fig. 1. The gen-
eral recycling process begins with the dismantling of the LIB and
insertion of the cathode into a DES. After heating and stirring, the
leachate can be filtered, and the foil, binder and residual conductive
carbon can be recovered separately. With precipitation or electrode-
position, the dissolved metal ions, such as lithium or cobalt, can be
recovered for use in other energy applications.
LCO was employed here as an initial model compound, given
the importance of repurposing the cobalt used in LIBs3. A DES

Rice University, Houston, TX, USA. *e-mail: babu.ganguli@rice.edu; ajayan@rice.edu

Nature Energy | VOL 4 | APRIL 2019 | 339–345 | www.nature.com/natureenergy 339

Articles NAture Energy

DES
Used LIBs

Aluminium foil
Housing
4% Disassemble
17%
Separator
3%

Li-ion
41%
Electrolyte 10% Cathode

18%
Binder
7%
Anode
Copper foil Aluminium foil

Heat
Conductive carbon
Cobalt recovery

Extraction of
cobalt and lithium

3 µm Precipitation
or
electrodeposition

3 µm

Fig. 1 | Battery recycling schematic. Starting with disassembly of the LIB, cathode waste is inserted into a DES, which is then heated and stirred.
Extraction of cobalt and lithium ions occurs through dissolution, and at this step, aluminium foil, binder and conductive carbon can be recovered
separately when the leachate is filtered. Cobalt compounds can then be recovered either through precipitation or electrodeposition, allowing reutilization
of these valuable materials.

composed of choline chloride and ethylene glycol (ChCl:EG) was
used to extract metal ions from LCO, which were then precipitated
and converted into Co3O4—a common precursor for the synthesis
of LCO26,27. We emphasize that both precursors used to prepare the
DES are biodegradable, making the proposed process more envi-
ronmentally friendly than existing alternatives.
Metal extraction from battery cathodes
In a typical experiment, 100 mg of LCO powder was added to 5 g of
the ChCl:EG DES. A series of these LCO–DES mixtures were sub-
jected to a range of temperatures from 25–220 °C for constant time
(24 h). The dissolution of cobalt was found to be highly dependent
on the temperature in which the extraction took place; at increas-
ing temperatures, the clear eutectic transformed into darker hues
of blue (Fig. 2a). This colour variation is indicative of different
cobalt extraction efficiencies at different temperatures. Very slight
blue-green colouration was seen at temperatures as low as 80 °C,
coinciding with the first detectable amount of cobalt registered by
inductively coupled plasma optical emission spectrometry (ICP-
OES; Fig. 2b and Supplementary Table 1). Until 120 °C, only nomi-
nal increases in the cobalt concentration could be perceived, as seen
in the inset of Fig. 2b. However, at 135 °C, there was an observable
discontinuity in the shading of colour, going from a pale blue-green
to a vibrant emerald green that coincided with a jump in the total
cobalt concentration by nearly an order of magnitude. As can be
seen from the nonlinear right axis of Fig. 2b, cobalt leaching efficien-
cies of up to 94.1% were reached at the highest temperature tested
(Supplementary Table 1). At 180 °C, leaching efficiencies as high as
99.3% could also be obtained by reducing the initial amount of LCO
powder used in the experiments (from 100 to 13 mg), which rivals
the leaching efficiencies of typical hydrometallurgical reagents such
as phosphoric and concentrated hydrochloric acid, which report
efficiencies of 97.8% (ref. 28) and 100% (ref. 29), respectively.
Ultraviolet-visible spectroscopy analysis of the liquids (after
removing unreacted LCO particles; see Methods) showed the sig-
nature bands of the tetrachlorocobaltate (ii) anion ([CoCl4]2−),
with 3 distinct bands between 600 and 700 nm (Supplementary
Fig. 1)30–33. For extraction temperatures between 160 and 195 °C,
the triplet fell into the same approximate wavenumbers; a system-
atic blue shift observed after thermal treatment at 220 °C could
have been indicative of solvation effects due to the formation of
additional complexes30.
The tetrachlorocobaltate (ii) anion has a characteristic blue
colour, and is responsible for the tonality of the samples exposed
to different temperatures. The formation of this anionic complex is
not surprising: chlorometalates have previously been proposed as
the dissolved species of certain metal oxides in the ChCl:EG eutec-
tic18. What is curious is the reduction associated with the dissolu-
tion process. In fact, Co(ii) exists in [CoCl4]2−, while Co(iii) occurs
in the oxide. Ethylene glycol, which is present in the eutectic, is a
common reducing agent used in the synthesis of metallic nanopar-
ticles34,35. It may be possible that the dissolution of LCO in ChCl:EG
involves the simultaneous oxidation of ethylene glycol and reduc-
tion of cobalt. Although the exact mechanisms for dissolution of

340 Nature Energy | VOL 4 | APRIL 2019 | 339–345 | www.nature.com/natureenergy

NAture Energy Articles
a
a b

MO stretching

% Transmittance
OH
25 50 60 80 105 120 135 160 180 195 220 stretching

Increasing temperature (°C) After calcining

Before calcining 2–
3 μm
b c CO3
120
14 100 72 h 3,000
Co concentration (×103 ppm)

80 2,000 1,000
Co concentration (ppm)
Leaching efficiency (%)
80 48 h
24 h Wavenumber (cm–1)
40
80 60
0 c d
(311)
7 30 60 90 120 150
50 40 Co3O4 Co 2p3/2 779.5

20
20
781.1

Intensity (a.u.)

Intensity (a.u.)
0 0
50 100 150 200 25 50 105 (440)

Temperature (°C) Temperature (°C) (511) 782.5

(220)
(400)

Fig. 2 | Extraction of cobalt using ChCl:EG DES. a, Photograph of the (422) (533)
colour change when LCO is dissolved to DES (which is clear in its
pure form) at different temperatures. Deeper blues with increasing
temperature correspond to the increasing concentration of cobalt. b, Cobalt 30 40 50 60 70 80 784 782 780 778 776
concentration versus temperature, showing the dependency on heat of the 2θ (°) Binding energy (eV)
dissolution of LCO in the DES for 24 h, with leaching efficiency shown on the
(nonlinear) right axis. Error bars of the averaged concentrations represent Fig. 3 | Characterization of precipitate from leaching cobalt ions from the
s.d. (at least three replicates). Inset, magnified view of the concentrations DES. a, FTIR spectra indicating the presence of carbonate groups before
between 25 and 135 °C. c, Cobalt concentration versus temperature for calcining (blue curve). The corresponding band vanishes on calcination
three different durations of dissolution show the time dependency of cobalt (red curve), which prompts the conversion of the powder into Co3O4.
leaching. b, Scanning electron microscopy image of the calcined powder, showing a
uniformly spherical microstructure. c, XRD spectra of the calcined powder,
exhibiting crystalline peaks matching those of the cubic cobalt oxide spinel
oxides by DESs have yet to be unravelled, Abbott et al.18 observed JCPDS 42-1467. d, XPS spectra of the calcined powder, showing fitted
that the presence of an oxygen acceptor in the eutectic appears to be peaks indicative of the binding energies of Co3O4.
essential for digesting oxides, as it facilitates the cleavage of metal–
oxide bonds. Ethylene glycol could be chemically acting as such
an acceptor, being oxidized (and reducing cobalt) in the process.
Note that typical routes for the dissolution of oxides by eutectics could be useful in the creation of new LIBs37–39). The caveat is that
implies partial alterations of the chemical makeup of the solvent. it is typically difficult to obtain crystalline cobalt carbonates by pre-
Yet, even when such transformations are not fully reversible during cipitation40; additionally, as Na2CO3 is basic and the blue solutions
recovery (by precipitation or electrodeposition, for example), the have a high cobalt content, precipitation of Co(OH)2 is also likely.
eutectic may still be functional: partially oxidized compounds may The precipitation experiments resulted in a light-brown powder,
still accept additional oxygen, while fully oxidized alcohols would containing essentially amorphous materials (Supplementary Fig. 2a)
provide protons, believed to be effective oxygen acceptors18. The —possibly a mixture of CoCO3, Co(OH)2 and Co3O4. The obser-
recycled DES can actually leach cobalt with efficiencies close to that vation of faint diffraction peaks of Co3O4 suggests the existence
of the fresh solvent, as is discussed later. of Co(iii) in the leachate. The Fourier transform infrared (FTIR)
Time was also investigated as a variable. The cobalt extrac- spectrum of the recovered powder (Fig. 3a, blue curve) showed
tion using DES was tested at 3 different temperatures (25, 50 and intense bands at 660, 1,380 and ~3,400 cm−1, which were respec-
105 °C) for 24, 48 and 72 h. The results shown in Fig. 2c indicate tively indicative of stretching modes from octahedrally and tetrahe-
the significant effect that the duration of heat exposure had on drally coordinated Co3+ and Co2+ metal oxide ions corresponding
the dissolution of cobalt. Dissolved metal concentrations more with spinel Co3O4, carbonate formation (that also had a small band
than doubled from 24–48 h, and at higher treatment temperatures, at 832 cm−1) and OH stretching, which arose from hydroxides and
extending the thermal treatment from 24–72 h increased extraction residual water in the sample41,42. X-ray photoelectron spectroscopy
levels by more than an order of magnitude. Hence, higher cobalt (XPS) provides additional compositional evidence for the precipi-
leaching efficiencies are also obtainable over greater exposure tated product. The XPS spectrum in Supplementary Fig. 2 suggests
times (Supplementary Table 2). that Co(OH)2 was present, due to peaks at 781.0, 782.7 and 786.4 eV,
as well as possible trace amounts of cobalt metal, which may have
Recovery strategies been the cause of the smallest peak at 778.8 eV (Supplementary
A full recycling scheme (Fig. 1) requires the recovery and repur- Fig. 2b)43–45. The inset therein is consistent with the 15 eV difference
posing of the metal ions. Given that H2O is typically a stronger in binding energy (spin–orbit coupling) between the 2p1/2 and 2p3/2
ligand than chloride36, Co(ii) exists in solution and CoCO3 has peaks of pure cobalt metal44,45. In the O 1s band (Supplementary
an extremely low solubility product in water, precipitation was Fig. 2c), a peak at 531.7 eV was observed, probably due to the
attempted by diluting the blue leachates with Na2CO3(aq.) (CoCO3 presence of carbonates46.

Nature Energy | VOL 4 | APRIL 2019 | 339–345 | www.nature.com/natureenergy 341

Articles
LCO
LCO dissolved in
Pure DES
DES
Fig. 4 | Recyclability of the DES. LCO was first added to the pure DES. Then, after stirring under 135 °C heat, cobalt was leached into the solution, causing
the colour change from clear to green (first to second photograph). The Co2+ ions were electrodeposited onto a substrate, allowing the remaining DES to
be recovered (third photograph). LCO was added to the recovered DES for another cycle of electrodeposition, and resulted in a similar colour change and
leaching efficiency to the first cycle (last photograph).
To create a more useful compound for synthetic purposes, cal-
cination of the recovered powder was employed at 500 °C for 6 h,
which seemed to convert most (if not all) components into uni-
formly spherical particles (Fig. 3b). The X-ray diffraction (XRD)
pattern of the calcined powder (Fig. 3c) aligns well with that of
Co3O4 (JCPDS 42-1467, cubic). Figure 3d shows the XPS curve fit-
ting of the Co 2p3/2 band with peaks at 779.5, 781.1 and 782.5, which
also correspond to the binding energies of Co3O443. This change
from ‘hydrated cobaltous carbonate’ to Co3O4 powder was also clear
from the disappearance of the CO32− FTIR band (Fig. 3a), and could
be seen qualitatively as the precipitate darkened from light-brown
to black after oxidizing (Supplementary Fig. 3). Conservative mass
calculations of the predicted species before and after calcining indi-
cated that about 74% of the cobalt from the DES leachate could be
recovered in the form of Co3O4.
We note that alternative methods, such as electrodeposition,
can also be used to recover metal ions from the leachate. Although
they are energy intensive, such procedures could allow the DES
itself to be recycled, making it available for additional cycles of
metal extraction. A proof of concept of this approach is shown in
Fig. 4. Experiments using the LCO–DES mixture treated at 135 °C
(details provided in Supplementary Discussion subsection ‘Deep
eutectic solvent recyclability’) resulted in the electrodeposition of
Co(OH)2 onto a stainless steel mesh working electrode. The liquid
phase after this treatment presented as a very pale colour, almost
as clear as in the pristine DES. After a new addition of LCO and
exposure of the mixture to 135 °C, vibrant green hues were once
again visible, which is a visual indication of cobalt leaching. The
resulting leaching efficiency approached that of the pristine DES,
indicating that the eutectic can potentially maintain its perfor-
mance over multiple cycles.
To verify the feasibility of using DES extraction in an actual bat-
tery electrode, the study was extended to observe the behaviour
towards other parts of a LIB. Different battery components (carbon
black, aluminium foil, the polyvinylidene fluoride (PVDF) binder
and copper foil; Supplementary Fig. 4) were exposed to the DES at
various temperatures, and it was generally observed that no reac-
tion occurred when mixing between room temperature and 105 °C,
with the exception of copper foil. LCO/Li half-cells were assembled
and tested for 300 charge/discharge cycles (see Methods for details).
The batteries were then dismantled and the harvested cathodes were
immersed in the ChCl:EG DES at 180 °C without further process-
ing. The electrode coating was observed to easily detach from the
current collector. DESs have been investigated for their plasticizing
abilities in polymers25, and it is thought that the adhesion between
the cathode material and aluminium foil is weakened as the inter-
chain interactions diminish, aided by heat and mechanical agita-
tion during stirring47. Hence, the cathode powder was separated
from the aluminium foil easily, and bluish hues were immediately
342
NAture Energy
LCO
Recovered DES after LCO dissolved
electrode position a second time
observed in the liquid phase, indicating prompt cobalt extraction
(Supplementary Fig. 5). The foil was pulled out, and it was then pos-
sible to filter out the PVDF binder and conductive carbon (along
with eventual unreacted LCO particles) from the solution. The
cobalt leaching efficiency for this process was found to be 29.6%
(the LCO:DES weight ratio was the same as in the experiments
above).
This initial study can be further generalized by inspecting the
behaviour of the ChCl:EG DES in contact with the increasingly
popular NMC battery chemistry. Similar to observations with LCO,
immersion of NMC powder to the fixed amount of eutectic at 180 °C
caused an immediate colour change to dark green (Supplementary
Fig. 6a). The dissolution of all metal ions in the DES was observed
and quantified using ICP-OES, but as lithium and cobalt were of
primary interest due to the growing demand for these resources2,3,48,
these are the ones reported here. The remaining metals are reported
in Supplementary Table 3. Leaching efficiencies as high as 71%
for lithium and 32% for cobalt were obtained by varying the ini-
tial cathode powder added to 5 g of DES (Fig. 5). Interestingly, the
leachate underwent additional colour change on cooling to room
temperature (Supplementary Fig. 6b), indicating the existence of
multiple dissolved complexes. The existence of such complexes may
explain the non-monotonic dependence between leaching and ini-
tial NMC content, as different species may be favoured under vary-
ing conditions. It may also justify the lower efficiencies for cobalt
extraction in NMC (versus LCO), as at a given NMC:DES ratio, the
different metal ions compete for a limited number of binding sites.
Dissolution of both lithium and cobalt favoured a ratio of 20 mg of
NMC powder to 5 g of DES, while the other metals behaved differ-
ently (Supplementary Table 3).
As the metal oxides become more complex (by having mul-
tiple transition metal ions), the separation processes required to
recover individual species also become more intricate. For similar
problems, industry relies on a combination of commercial solvent
extractants (liquid–liquid extraction) that can bond specifically to
certain metals of interest, as well as chemical precipitation for other
steps1,47. Other recovery possibilities include using electrowinning
techniques that can selectively deposit the metals, one at a time13. It
is then reasonable to expect that all of these concepts can be simi-
larly applied to DESs, enabling the reutilization of the many ele-
ments composing the cathode materials.
Conclusions
Although DESs are known to be capable of dissolving metal oxides,
this activity has never been fully explored with materials of inter-
est to the LIB industry. Here, we demonstrate that a DES made of
ChCl:EG can extract metals from LIB cathode materials, attaining
leaching efficiencies as high as 99.3%. Time and temperature
played an important role in dictating the effectivity of the metal
Nature Energy | VOL 4 | APRIL 2019 | 339–345 | www.nature.com/natureenergy
NAture Energy
Cobalt
80 Lithium
Leaching efficiency (%)
60
40
20
0
0.02 0.04 0.10
NMC concentration (M)
Fig. 5 | Leaching efficiency of NMC powder dissolved in the DES at 180 °C.
Lithium and cobalt can be extracted with different efficiencies depending
on the initial concentration of NMC powder added to the DES.
extraction, with oxide dissolution being favoured by longer and
warmer experiments.
In addition to reporting the dissolution of LCO and NMC in
a DES, we have demonstrated that the metals can be recovered as
synthetically useful products, by precipitating dissolved ions and
calcining the obtained compounds. Furthermore, we have shown
that when dilution of the DES is avoided (by recovering cobalt by
electrodeposition), the residual eutectic can once more be employed
to leach metals from LCO, with no apparent loss in efficiency.
Environmentally friendly DESs not only present a possibility to
change the way we bring new life back into waste products, as is
proposed in this work, but also inspire the search of new paradigms
for sustainably recycling LIBs.
Methods
Metal extraction using a DES. The DES was formed at room temperature by
combining choline chloride (HOC2H4N[CH3]3Cl; ≥98%; Sigma–Aldrich) and
ethylene glycol (HOCH2CH2OH, anhydrous; 99.8%; Sigma–Aldrich) in a 1:2 molar
ratio. The components were constantly stirred until a clear, homogenous solution
was obtained. Additional details regarding the preparation and characterization
of this DES can be found elsewhere19,24,25,49. For the metal extraction experiments,
0.1 g of LCO powder (99.8%; Sigma–Aldrich) was added to 5 g of DES and mixed
in a closed glass vial. The temperature dependency of the leaching efficiency
was evaluated by placing individual vials into an oil bath for 24 h. Average bath
temperatures ranged from 25–220 °C (±10 °C, limited by the heating element
in the hot plate). In separate experiments, the effect of time on metal extraction
was assessed by maintaining the mixtures for 48 and 72 h at different average
temperatures (25, 50 and 105 °C). After each thermal treatment, the LCO + DES
leachate was poured into a funnel lined with a hydrophilic polyvinylidene
Durapore membrane filter (90 mm diameter; 0.45 µm pore size; MilliporeSigma),
and the resulting filtrate was used for characterization and subsequent recovery.
Leaching efficiency. The leaching efficiency (η) was defined as:
η = (CV ∕mx ) × 100% (1)
where C is the final concentration of the metal (in mg l−1), V is the volume of initial
leaching solution (in l) and mx is the mass of the initial amount of x (Li+ or Co3+, for
example) in the active material (in mg)50–53.
ICP-OES. The metal content in filtrates obtained under different conditions
was quantified using a PerkinElmer Optima 8300 ICP-OES system. The samples
were diluted with a 2% aqueous solution of nitric acid, and calibration curves
were generated using at least 5 ICP standard solutions, with the results used only
from correlation coefficients that were greater than 0.999. The gas nebulizer
flow rate range was set between 0.45 and 0.75 l min−1, and 2 wavelengths per
element were used in the axial mode unless otherwise stated: cobalt (228.616 and
230.786 nm), lithium (670.784 nm (radial mode) and 610.362 nm), nickel (231.604
and 341.476 nm) and manganese (257.610 and 259.372 nm). The concentrations
Nature Energy | VOL 4 | APRIL 2019 | 339–345 | www.nature.com/natureenergy
Articles
reported here are the average of at least three replicates, while the error bars
illustrate the observed standard deviation between the replicates.
Ultraviolet-visible spectrometry. Inside an Agilent Cary 60 UV-Vis
spectrophotometer, ChCl:EG DES liquid samples were directly inserted into
a 1-mm-pathlength, 0.35-ml-capacity plastic cuvette using a syringe. Baseline
measurements in air were taken and compared with measurements with samples
in the range 1,600 to 250 nm. Pure DES, as well DES treated with LCO at average
temperatures between 25 and 220 °C were measured.
XRD. Using 40 kV and 40 mA Cu Kα radiation, a Rigaku D/Max Ultima II Power
XRD was used to characterize the structure of the precipitated powder before
and after calcining. The powder was ground finely in a mortar and pestle, then
placed on a zero-background sample holder. The instrument scan speed was set to
0.5° min−1, with an automatic slit control (Rec. Slit) of 0.6 mm for high intensity,
and scanned from 10–90° (2θ).
FTIR spectrometry. A Thermo Fisher Scientific Nicolet iS10 FTIR equipped
with a diamond crystal accessory and ZnSe beam splitter was used to produce
high-resolution infrared information about the vibrational modes of the
precipitated powder. The finely ground powder sample was loaded directly onto
the cleaned collection surface and scanned from the spectral range of 4,000 to
650 cm−1. All absorbance data collected were then processed using an automatic
baseline correct, normalized scale, and automatic smoothing, then converted into
transmittance readings.
XPS. A PHI Quantera XPS Microprobe was used to analyse the bonding state of
the precipitated powder. An initial surface scan sweep was set at a pass energy of
140 eV and a time step of 20 ms, then multiple elemental sweeps for carbon, cobalt
and oxygen were set at a pass energy of 26 eV and run. The resulting peaks were
calibrated to carbon’s shift at 284.8 eV and fitted using the MultiPak software.
Scanning electron microscopy. On a FEI Quanta 400, the precipitated powder
samples were mounted on carbon tape and analysed using 25 kV with a working
distance of 10.2 mm. The dwell time was set to 2 μs, and the images used in this
work were magnified at 20,000×. No further image processing was performed.
Metal recovery. The dissolved cobalt contained in the filtrate was recovered by
precipitation in a carbonate-rich solution; 4 ml of the filtrate was diluted in 10 ml
of a 20 wt% aqueous solution of Na2CO3. On mixing, the blue filtrate immediately
turned brown and was centrifuged at 12,000 r.p.m. for 30 min. The precipitate was
transferred to a Durapore membrane filter and rinsed with deionized water until
the collected liquid would not cloud on addition of an AgNO3 solution (indicating
the removal of ChCl’s chloride from the solid). The cleaned precipitate powder
was dried in an oven at 90 °C, then calcined at 500 °C for 6 h. A description of
the experiments using electrodeposition to recover cobalt is provided in the
Supplementary Discussion subsection ‘Deep eutectic solvent recyclability’.
Half-cell battery. LCO/Li half cells were prepared using cathode slurries composed
of 80% LCO (99.8% purity; Sigma–Aldrich), 10% carbon black super P conductive
(≥99%; Alfa Aesar) and 10% PVDF binder dissolved in N-methyl-pyrrolidinone
(anhydrous; 99.5%; Sigma–Aldrich). The slurry was homogenized in a Thinky
planetary mixer, employing three successive mixing/resting loops (4 min at
1,500 r.p.m., then 3 min at rest). The slurry was then transferred to a cleaned
aluminium foil and rolled into a 120 µm film (wet thickness) using a TQC applicator.
The electrode was laid flat to dry and subsequently left overnight in a vacuum oven
at 90 °C. The dried sheet was pressed at 90 °C in an MTI HR01 Hot Rolling Machine
down to an average thickness of 12.2 µm (25.7 µm including the aluminium foil).
Punched discs (16 mm diameter) were used for cell assembly. The LCO/Li coin cells
were constructed in an argon-filled glovebox (MBraun UNIlab) using lithium chips
from Alfa Aesar (0.7 mm thick). Cells were potentiostatically charged and discharged
for 300 cycles at a 1 C rate, using 3.4 and 4.2 V as cut-offs on a Lanhe cycler. After the
final discharge, the spent batteries were dismantled, and the entire harvested LCO
electrode was immersed in the DES at 180 °C for 24 h without any further processing.
Data availability
The data that support the plots and tables within this paper and its Supplementary
Information files, as well as the other findings of this study, are available from the
corresponding author upon reasonable request.
Received: 12 October 2018; Accepted: 1 March 2019;
Published online: 1 April 2019
References
1. Zheng, X. et al. A mini-review on metal recycling from spent lithium ion
batteries. Engineering 4, 361–370 (2018).
2. Zeng, X., Li, J. & Singh, N. Recycling of spent lithium-ion battery: a critical
review. Crit. Rev. Env. Sci. Tech. 44, 1129–1165 (2014).
343
Articles
3. Frankel, T. C. The cobalt pipeline: tracing the path from deadly hand-dug
mines in Congo to consumers' phones and laptops. The Washington Post
(30 September 2016); https://www.washingtonpost.com/graphics/business/
batteries/congo-cobalt-mining-for-lithium-ion-battery/?noredirect=on
4. Peters, J. F., Baumann, M., Zimmermann, B., Braun, J. & Weil, M. The
environmental impact of Li-ion batteries and the role of key parameters—a
review. Renew. Sustain. Energy Rev. 67, 491–506 (2017).
5. Zhang, P., Yokoyama, T., Itabashi, O., Suzuki, T. M. & Inoue, K.
Hydrometallurgical process for recovery of metal values
from spent lithium-ion secondary batteries. Hydrometallurgy 47,
259–271 (1998).
6. Li, J., Wang, G. & Xu, Z. Environmentally-friendly oxygen-free roasting/wet
magnetic separation technology for in situ recycling cobalt, lithium carbonate
and graphite from spent LiCoO2/graphite lithium batteries. J. Hazard. Mater.
302, 97–104 (2016).
7. Guo, Y. et al. Improved extraction of cobalt and lithium by reductive acid
from spent lithium-ion batteries via mechanical activation process. J. Mater.
Sci. 53, 13790–13800 (2018).
8. Chagnes, A. & Swiatowska, J. Lithium Process Chemistry: Resources,
Extraction, Batteries, and Recycling (Elsevier, 2015).
9. Huang, B., Pan, Z., Su, X. & An, L. Recycling of lithium-ion batteries: recent
advances and perspectives. J. Power Sources 399, 274–286 (2018).
10. Chen, W.-S. & Ho, H.-J. Recovery of valuable metals from lithium-ion
batteries NMC cathode waste materials by hydrometallurgical methods.
Metals 8, 321 (2018).
11. Sun, C., Xu, L., Chen, X., Qiu, T. & Zhou, T. Sustainable recovery of valuable
metals from spent lithium-ion batteries using DL-malic acid: leaching and
kinetics aspect. Waste Manage. Res. 36, 113–120 (2018).
12. Yao, L., Yao, H., Xi, G. & Feng, Y. Recycling and synthesis of
LiNi1/3Co1/3Mn1/3O2 from waste lithium ion batteries using d,l-malic acid.
RSC Adv. 6, 17947–17954 (2016).
13. Yao, Y. et al. Hydrometallurgical processes for recycling spent
lithium-ion batteries: a critical review. ACS Sustain. Chem. Eng. 6,
13611–13627 (2018).
14. Sun, L. & Qiu, K. Organic oxalate as leachant and precipitant for the recovery
of valuable metals from spent lithium-ion batteries. Waste Manage. 32,
1575–1582 (2012).
15. Albler, F. J., Bica, K., Foreman, M. R. S. J., Holgersson, S. & Tyumentsev, M.
S. A comparison of two methods of recovering cobalt from a deep
eutectic solvent: implications for battery recycling. J. Clean. Prod. 167,
806–814 (2017).
16. Di Marino, D., Shalaby, M., Kriescher, S. & Wessling, M. Corrosion
of metal electrodes in deep eutectic solvents. Electrochem. Commun. 90,
101–105 (2018).
17. Foreman, M. R. S. Progress towards a process for the recycling of nickel
metal hydride electric cells using a deep eutectic solvent. Cogent Chem. 2,
1139289 (2016).
18. Abbott, A. P., Capper, G., Davies, D. L., McKenzie, K. J. & Obi, S. U.
Solubility of metal oxides in deep eutectic solvents based on choline chloride.
J. Chem. Eng. Data 51, 1280–1282 (2006).
19. Zhang, Q., De Oliveira Vigier, K., Royer, S. & Jérôme, F. Deep eutectic
solvents: syntheses, properties and applications. Chem. Soc. Rev. 41,
7108–7146 (2012).
20. Zhao, B.-Y. et al. Biocompatible deep eutectic solvents based on choline
chloride: characterization and application to the extraction of rutin from
Sophora japonica. ACS Sustain. Chem. Eng. 3, 2746–2755 (2015).
21. García, G., Aparicio, S., Ullah, R. & Atilhan, M. Deep eutectic solvents:
physicochemical properties and gas separation applications. Energy Fuels 29,
2616–2644 (2015).
22. Radošević, K. et al. Evaluation of toxicity and biodegradability of choline
chloride based deep eutectic solvents. Ecotoxicol. Environ. Saf. 112,
46–53 (2015).
23. Millia, L. et al. Bio-inspired choline chloride-based deep eutectic
solvents as electrolytes for lithium-ion batteries. Solid State Ion. 323,
44–48 (2018).
24. Tang, B., Zhang, H. & Row, K. H. Application of deep eutectic solvents
in the extraction and separation of target compounds from various samples:
other techniques. J. Sep. Sci. 38, 1053–1064 (2015).
25. Smith, E. L., Abbott, A. P. & Ryder, K. S. Deep eutectic solvents (DESs) and
their applications. Chem. Rev. 114, 11060–11082 (2014).
26. Contestabile, M., Panero, S. & Scrosati, B. A laboratory-scale lithium-ion
battery recycling process. J. Power Sources 92, 65–69 (2001).
27. Pegoretti, V. C. B., Dixini, P. V. M., Smecellato, P. C., Biaggio, S. R. & Freitas,
M. B. J. G. Thermal synthesis, characterization and electrochemical study of
high-temperature (HT) LiCoO2 obtained from Co(OH)2 recycled of spent
lithium ion batteries. Mater. Res. Bull. 86, 5–9 (2017).
28. Chen, X. et al. Separation and recovery of metal values from leaching
liquor of mixed-type of spent lithium-ion batteries. Sep. Purif. Technol. 144,
197–205 (2015).
344
NAture Energy
29. Joulié, M., Laucournet, R. & Billy, E. Hydrometallurgical process
for the recovery of high value metals from spent lithium nickel cobalt
aluminum oxide based lithium-ion batteries. J. Power Sources 247,
551–555 (2014).
30. Coleman, J. S. Chloride complexes of cobalt(II) in anion and cation
exchangers. J. Inorg. Nucl. Chem. 28, 2371–2378 (1966).
31. Hsieh, Y.-T., Lai, M.-C., Huang, H.-L. & Sun, I.-W. Speciation of cobalt-
chloride-based ionic liquids and electrodeposition of Co wires. Electrochim.
Acta 117, 217–223 (2014).
32. Wellens, S. et al. Dissolution of metal oxides in an acid-saturated ionic liquid
solution and investigation of the back-extraction behaviour to the aqueous
phase. Hydrometallurgy 144–145, 27–33 (2014).
33. Wellens, S., Thijs, B. & Binnemans, K. An environmentally friendlier
approach to hydrometallurgy: highly selective separation of cobalt from
nickel by solvent extraction with undiluted phosphonium ionic liquids.
Green Chem. 14, 1657–1665 (2012).
34. Toshima, N. & Yonezawa, T. Bimetallic nanoparticles—novel
materials for chemical and physical applications. New J. Chem. 22,
1179–1201 (1998).
35. Luo, C., Zhang, Y., Zeng, X., Zeng, Y. & Wang, Y. The role of poly(ethylene
glycol) in the formation of silver nanoparticles. J. Colloid Interface Sci. 288,
444–448 (2005).
36. Huheey, J. E., Keiter, E. A., Keiter, R. L. & Medhi, O. K. Inorganic Chemistry:
Principles of Structure and Reactivity (Pearson Education, 2006).
37. Du, H. et al. Morphology control of CoCO3 crystals and their conversion to
mesoporous Co3O4 for alkaline rechargeable batteries application.
CrystEngComm 15, 6101–6109 (2013).
38. González-López, J., Fernández-González, Á. & Jiménez, A. Precipitation
behaviour in the system Ca2+-Co2+-CO32−-H2O at ambient conditions—
amorphous phases and CaCO3 polymorphs. Chem. Geol. 482,
91–100 (2018).
39. Barber, D. M., Malone, P. G. & Larson, R. J. The effect of cobalt ion
on nucleation of calcium-carbonate polymorphs. Chem. Geol. 16, 2
39–241 (1975).
40. Katsikopoulos, D., Fernández-González, Á., Prieto, A. C. & Prieto, M.
Co-crystallization of Co(ii) with calcite: implications for the mobility of
cobalt in aqueous environments. Chem. Geol. 254, 87–100 (2008).
41. Xia, X. et al. Freestanding Co3O4 nanowire array for high performance
supercapacitors. RSC Adv. 2, 1835–1841 (2012).
42. Cheng, J. P., Chen, X., Ma, R., Liu, F. & Zhang, X. B. A facile method to
fabricate porous Co3O4 hierarchical microspheres. Mater. Charact. 62,
775–780 (2011).
43. Biesinger, M. C. et al. Resolving surface chemical states in XPS analysis of
first row transition metals, oxides and hydroxides: Cr, Mn, Fe, Co and Ni.
Appl. Surf. Sci. 257, 2717–2730 (2011).
44. Khassin, A. A. et al. Metal–support interactions in cobalt–aluminum
co-precipitated catalysts: XPS and CO adsorption studies. J. Mol. Catal. Chem.
175, 189–204 (2001).
45. Moulder, J. F., Stickle, W. F., Sobol, P. E. & Moben, K. D. Handbook of X-Ray
Photoelectron Spectroscopy (Perkin-Elmer Corporation, 1992).
46. Stoch, J. & Gablankowska-Kukucz, J. The effect of carbonate contaminations
on the XPS O 1s band structure in metal oxides. Surf. Interface Anal. 17,
165–167 (1991).
47. Zeng, X., Li, J. & Singh, N. Recycling of spent lithium-ion battery: a critical
review. Crit. Rev. Environ. Sci. Technol. 44, 1129–1165 (2014).
48. Wanger, T. C. The lithium future—resources, recycling, and the environment:
the lithium future. Conserv. Lett. 4, 202–206 (2011).
49. Harifi-Mood, A. R. & Buchner, R. Density, viscosity, and conductivity of
choline chloride + ethylene glycol as a deep eutectic solvent and its binary
mixtures with dimethyl sulfoxide. J. Mol. Liq. 225, 689–695 (2017).
50. Wang, M.-M., Zhang, C.-C. & Zhang, F.-S. An environmental benign process
for cobalt and lithium recovery from spent lithium-ion batteries by
mechanochemical approach. Waste Manage. 51, 239–244 (2016).
51. Fan, B., Chen, X., Zhou, T., Zhang, J. & Xu, B. A sustainable process for the
recovery of valuable metals from spent lithium-ion batteries. Waste Manage.
Res. 34, 474–481 (2016).
52. Bhargava, S., Pownceby, M. & Ram, R. Hydrometallurgy (MDPI, 2017).
53. Jo, H., Jo, H., Rha, S. & Lee, P.-K. Direct aqueous mineral carbonation of
waste slate using ammonium salt solutions. Metals 5, 2413–2427 (2015).
Acknowledgements
The authors thank A. Kabbani and A. Puthirath for useful discussion, and L. Alexander
for assistance with the NMC dissolution experiments. M.K.T. acknowledges the National
Science Foundation for continued support and funding. This study is based on work
supported by the National Science Foundation Graduate Research Fellowship Program
under grant number 1450681. Any opinions, findings, conclusions or recommendations
expressed are those of the authors and do not necessarily reflect the views of the National
Science Foundation.
Nature Energy | VOL 4 | APRIL 2019 | 339–345 | www.nature.com/natureenergy
NAture Energy
Author contributions
M.-T.F.R. conceived of the experimental design. M.K.T. performed the experiments and,
alongside M.-T.F.R., co-wrote the paper and analysed the data. K.K. assisted in figure
creation, as well as XPS and electrochemical experimentation and analysis. P.M.A. and
G.B. conceived of and contributed to the overall project planning.
Competing interests
The authors declare no competing interests.
Nature Energy | VOL 4 | APRIL 2019 | 339–345 | www.nature.com/natureenergy
Articles
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41560-019-0368-4.
Reprints and permissions information is available at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to G.B. or P.M.A.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2019
345