Efficient and Sustainable Removal of Magnesium from Brines for Lithium/Magnesium Separation Using Binary Extractants

Abstract

Lithium is becoming increasingly important due to its essential role in lithium-ion batteries. Over 70% of the global lithium resources are found in salt lake brines, but lithium is always accompanied by magnesium. It is a challenge to efficiently separate lithium from magnesium in brines. The state-of-the-art processes for lithium/magnesium separation either consume large quantities of chemicals and generate large amounts of waste, or are energy-intensive. In this study, we develop a sustainable solvent extraction process based on binary extractants to efficiently separate lithium and magnesium. A binary extractant comprised of Aliquat® 336 and Versatic Acid® 10, [A336][V10], was prepared and investigated for removal of magnesium from both a (synthetic) concentrated brine (106 g L-1 Mg and 10 g L-1 Li) and an (synthetic) original brine (15 g L-1 Mg, 80 g L-1 Na and 0.2 g L-1 Li). Through batch counter-current experiments and mixer-settler experiments, it was found that [A336][V10] is able to quantitatively remove magnesium from the original brine in three continuous counter-current extraction stages with as little as about 10% co-extraction of lithium. The loaded organic phase can be stripped and regenerated by water. The whole process (extraction and stripping) does not consume any acid or base, but makes use of the differences in the chloride concentration during extraction and stripping. This process is an environmentally friendly alternative to the state-of-the-art processes and represents a step forward in the sustainable production of Li2CO3 from brines.

Full text

Efficient and Sustainable Removal of Magnesium from Brines for
Lithium/Magnesium Separation Using Binary Extractants
Zheng Li,*Jonas Mercken, Xiaohua Li, Sofía Riaño, and Koen Binnemans
Department of Chemistry, KU Leuven, Celestijnenlaan 200F, Bus 2404, B-3001 Heverlee, Flemish Brabant, Belgium
*
SSupporting Information
ABSTRACT: Lithium is becoming increasingly important
due to its essential role in lithium-ion batteries. Over 70% of
the global lithium resources are found in salt lake brines, but
lithium is always accompanied by magnesium. It is a challenge
to efficiently separate lithium from magnesium in brines. The
state-of-the-art processes for lithium/magnesium separation
either consume large quantities of chemicals and generate
large amounts of waste or are energy-intensive. In this study,
we develop a sustainable solvent extraction process based on
binary extractants to efficiently separate lithium and
magnesium. A binary extractant composed of Aliquat 336
and Versatic Acid 10, [A336][V10], was prepared and
investigated for removal of magnesium from both a
(synthetic) concentrated brine (106 g L−1Mg and 10 g L−1Li) and an (synthetic) original brine (15 g L−1Mg, 80 g L−1
Na and 0.2 g L−1Li). Through batch counter-current experiments and mixer−settler experiments, it was found that
[A336][V10] is able to quantitatively remove magnesium from the original brine in three continuous counter-current extraction
stages with as little as about 10% coextraction of lithium. The loaded organic phase can be stripped and regenerated by water.
The whole process (extraction and stripping) does not consume any acid or base but makes use of the differences in the
chloride concentration during extraction and stripping. This process is an environmentally friendly alternative to the state-of-
the-art processes and represents a step forward in the sustainable production of Li2CO3from brines.
KEYWORDS: binary extractant, solvent extraction, lithium, magnesium, brine
■INTRODUCTION
The transition from fossil fuel-based cars to electric (and
hybrid electric) cars is an important measure to reduce CO2
emissions and hence to counteract global warming. Lithium-
ion batteries (LIBs) are nowadays commonly used in electric
vehicles due to their high energy density and lightweight
compared to other types of batteries.
1,2
The improved
performance of the new generations of LIBs is an additional
incentive to electrification of transportation.
3,4
The large
amounts of lithium (as battery-grade Li2CO3) needed for the
production of LIBs lead to a sharply increasing demand of
lithium.
5−7
Pegmatites and brines are the two main resources of lithium.
Although seawater contains huge amounts of lithium, the
lithium concentration (about 0.17 ppm) is too low to be
exploitable in the short term.
8
The global exploitable lithium
resources are estimated to be 31−34 Mt, with salt lake brines
comprising over 70% of these resources.
9
Currently, the
majority of the global lithium production is obtained from
these brines.
6,10,11
Magnesium always accompanies lithium in
salt lake brines. It is challenging to separate lithium from
magnesium because these elements exhibit similar chemical
properties because of their diagonal relationship in the periodic
table.
The Mg/Li ratio is a key parameter determining whether a
brine can be exploited for lithium. The current commercial
process for lithium recovery from the brines with low Mg/Li
ratios (<8) consists of several steps: (1) concentration of
brines by evaporation in a solar pond for about 1 year, (2)
removal of magnesium by lime milk (Ca(OH)2), (3) removal
of calcium by addition of Na2CO3/Li2CO3, and finally (4)
precipitation of Li2CO3by addition of Na2CO3.
12,13
This
process creates a heavy environmental burden because of the
high consumption of precipitation reagents, generation of large
amounts of waste, and high consumption of water for washing
the precipitate. Furthermore, it suffers from a low recovery
efficiency due to the loss of lithium during precipitation.
Recovery of lithium from brines with high Mg/Li ratios (>8,
and the ratio can sometimes even be >1000) is more
challenging, and only a few processes are operated at an
industrial scale worldwide. The West Taijinar Salt Lake in
China is a representative brine with a high Mg/Li ratio (Mg/Li
≈100) and it has been mined by CITIC China since
2007.
14−16
The key operation is the conversion of MgCl2to
Received: September 12, 2019
Revised: October 28, 2019
Published: October 31, 2019
Research Article
pubs.acs.org/journal/ascecg
Cite This: ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
© 2019 American Chemical Society 19225 DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
Downloaded via KU LEUVEN on December 4, 2019 at 17:48:11 (UTC).
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

MgO by roasting powder mixtures of chloride salts that are
obtained by spray drying of concentrated brines after
extraction of potassium (for production of fertilizers). This
process is not only energy-intensive and water-consuming but
also environmentally detrimental because HCl gas is produced
during the roasting step. The process is economically viable
only when the lithium price is high and it is shut down when
the lithium price is too low. In conclusion, the state-of-the-art
processes for the separation of lithium and magnesium in
brines have severe limitations, independent of the Mg/Li
ratios. Considering the increasing demand for lithium, it is
obvious that the development of more environmentally
friendly processes for the recovery of lithium from brines is
indispensable for the sustainable utilization of lithium
resources.
Selective solvent extraction of lithium from magnesium-rich
brine would be the best solution. However, this is also one of
the most difficult separation challenges and it has been studied
for about half a century without a real breakthrough. The most
intensively studied solvent extraction system for lithium and
magnesium separation is the synergistic solvent extraction
system consisting of tributyl phosphate and FeCl3(or ionic
liquids).
17−19
Despite good lithium selectivity over magnesium,
this system suffers from difficult stripping and loss of FeCl3(or
cations of ionic liquids) into the aqueous phase. These
limitations prevent the application of this extraction system.
Following a different route, selective solvent extraction of
magnesium can be an alternative process for lithium and
magnesium separation because the extraction of magnesium is
easier than the extraction of lithium. Acidic extractants are
known to be able to efficiently extract magnesium. Zhang et al.
used saponified D2EHPA (di-(2-ethylhexyl)phosphoric acid)
to extract magnesium from a synthetic brine solution
containing 19.5 g L−1Mg and 0.02 g L−1Li.
20
Despite good
magnesium extraction, the coextraction of lithium was too
high. Following the same concept, Virolainen et al. studied the
removal of small amounts of calcium and magnesium
impurities from concentrated lithium solutions using
D2EHPA and Versatic Acid 10 (V10, a mixture of carboxylic
acids with the common structural formula of C10H20O2) and
found that V10 has less co-extraction of lithium.
21
Unfortu-
nately, the method of Virolainen et al. is not suitable for direct
removal of magnesium from brine, because magnesium in
brine is so abundant that the consumption of base during
extraction (to neutralize the protons) and acids during
stripping (to regenerate the acidic extractant) would be too
large.
To avoid the consumption of acids and bases for the
selective removal of magnesium from magnesium-rich brines,
this study investigates the use of binary extractants instead of
acidic extractants. Binary extractants consist of an acidic
extractant saponified with a basic extractant.
22−29
In other
words, it is an ionic liquid with the cation being (quaternary)
ammonium and the anion being a de-protonated acidic
extractant molecule. It is worth mentioning that quaternary
phosphonium compounds work in the same way as quaternary
ammonium compounds although the former is much more
expensive. Some phosphonium-based binary extractants (e.g.,
trihexyl(tetradecyl)phosphonium bis-2,4,4-(trimethylpentyl)
phosphinate, also known as Cyphos IL 104) have also been
studied for extraction of metal ions.
30,31
In the extraction of
magnesium using acidic extractants, bases are needed to
neutralize the released protons in order to drive the extraction
reaction; for the stripping, acids are needed to regenerate the
extractant. In other words, the extraction and stripping are
driven by acidity (or pH) of the aqueous solution. On the
contrary, extraction and stripping of magnesium by a binary
extractant is driven by the common ion effect, that is, the
concentration of chloride ions. When the chloride concen-
tration is high, magnesium is extracted; when the loaded
organic phase is contacted with water, magnesium is stripped
due to the low chloride concentration. Therefore, no acids or
bases are consumed during the extraction or stripping.
Unfortunately, binary extractants have not found any real
applications so far, despite many studies about magnesium
extraction from seawater mentioned above. The main obstacle
for the application of binary extractants is perhaps the lack of
driving force during extraction, that is, the concentration of
chloride in seawater (about 19 g L−1) is not sufficient to drive
efficient magnesium extraction. By contrast, (concentrated)
salt lake brine has much higher chloride concentrations (can be
>150 g L−1); hence, it is very suitable for magnesium
extraction using binary extractants.
In this study, we use Aliquat 336 (A336, a mixture of
methyltrioctylammonium chloride and methyltridecylammo-
nium chloride, with the former dominating) as the basic
extractant and Cyanex 272 (C272, bis(2,2,4-trimethylpentyl)-
phosphinic acid) and Versatic Acid 10 as the acidic extractant,
to prepare binary extractants. These extractants were chosen
because they are commercially available in large scale at
reasonable prices, and Versatic Acid 10 and Cyanex 272 are
known to selectively extract magnesium over lithium.
21,32
Two
binary extractants, [A336][V10] and [A336][C272] were
synthesized and tested for the removal of magnesium from two
synthetic brines: concentrated brine (106 g L−1Mg and 10 g
L−1Li) and original brine (15 g L−1Mg, 80 g L−1Na and 0.2 g
L−1Li). We show that [A336][V10] is able to remove
magnesium quantitatively from the original brine in a three-
stage counter-current extraction with about 10% co-extraction
of lithium, without consumption of any acids or bases.
■EXPERIMENTAL SECTION
Chemicals. Aliquat 336 (∼90%) was obtained from Sigma-Aldrich
(Diegem, Belgium); Cyanex 272 (85%) was obtained from Cytec
Industries B.V. (Vlaardingen, Netherlands); Versatic Acid 10 (>90%)
was obtained from Resolution Europe B.V. (Amsterdam, The
Netherlands); NaOH (analytical reagent), NaCl (≥99.5%), HCl
(37%), toluene (analytical reagent), and ethylenediaminetetraacetic
acid (EDTA, 0.10 M) were supplied by Fisher Scientific (Merelbeke,
Belgium); LiCl (≥99.5%) was obtained from Carl ROTH (Karlsruhe,
Germany); MgCl2(99%), p-cymene (>99%), n-heptane (99% for
HPLC), MIBK (99.5%), n-decanol (98%), and Eriochrome Black T
were obtained from Acros Organics (Geel, Belgium); KCl (analytical
reagent), NH4Cl (analytical reagent), Li, Na, and Mg standard
solutions (1000 ±10 mg L−1) were purchased from Chem-Lab
(Zedelgem, Belgium); NH3aqueous solution (25%) was purchased
from VWR Chemicals (Leuven, Belgium). Milli-Q water (18.2 MΩ
cm at 298.2 K) was used to prepare the aqueous solutions. All
chemicals were used as received, without any further purification.
General Synthesis Procedure of the Binary Extractants. The
synthesis of the binary extractants was based on literature
procedures.
22,26,33
The following general reaction scheme is applicable
for the synthesis
V
[
′] + + [ ′] + +
+− +−
NR R Cl HA NaOH NR R A NaCl H
O
332
(1)
with R′= methyl, R = octyl or decyl and HA is the acidic extractant.
NaOH was used to deprotonate the acidic extractant and to drive the
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19226

equilibrium to the right. No toxic byproducts were produced during
the synthesis, just NaCl and water.
Calculated amounts of the basic (Aliquat 336) and the acidic
extractants (Cyanex 272 or Versatic Acid 10) were added together in
a 1:1 stoichiometric ratio. The amounts (in moles) added were
estimated taking the purity of the commercial products into account.
A freshly prepared NaOH solution (4.0 mol L−1) was added to the
mixture before it was cooled down completely. The still warm NaOH
solution (51 °C) helped to decrease the viscosity of the mixture and
increased the stirring efficiency. The mixture was agitated vigorously
for 4 h. Afterward, the mixture was left for phase separation. The
organic layer was collected and washed three times with Milli-Q water
to remove chloride ions. The complete removal of chloride ions was
checked with an acidified silver nitrate solution. Any entrained water
in the organic phase after washing was removed using a rotary
evaporator. The 1H(Figures S1 and S3) and 13C NMR spectra
(Figures S2 and S4) together with the elemental analysis of the
compounds can be found in the Supporting Information. The yield of
[A336][C272] was 94% and that of [A336][V10] was 96%. The
structures of [A336][C272] and [A336][V10] are shown in Figure 1.
Preparation of Synthetic Brine Solutions. Two synthetic feed
solutions were prepared based on the composition of the West
Taijinar Salt Lake in China, which is a representative salt lake with a
high Mg/Li ratio that has been mined for lithium. The first feed
solution (106 g L−1Mg, 10 g L−1Li) mimics the concentrated brine,
which would be spray-dried and roasted in the current process. The
second feed solution (15 g L−1Mg, 0.2 g L−1Li, 80 g L−1Na) mimics
the original brine of the West Taijinar Salt Lake. The feed solutions
were prepared by dissolving weighed amounts of chloride salts in
Milli-Q water.
Extraction Experiments. Each extraction experiment was carried
out in a 15 mL centrifuge tube with 5.0 mL of aqueous solution and
5.0 mL of organic solution, respectively. Mixtures of the two phases
were shaken for 30 min at 300 rpm using a Thermo Scientific 2000
shaker to attain equilibrium. Afterward, the samples were centrifuged
for 5 min at 4000 rpm. Scrubbing and stripping experiments were
carried out using the same method. Organic-to-aqueous phase ratios
(O/A) were varied by varying the organic and aqueous volumes
accordingly. Investigation of the effect of metal concentration was
carried out by varying the metal (salt) concentration of interest, while
keeping the concentrations of other salts constant. The aqueous
phases after equilibrium and resultant aqueous solutions after
stripping were analyzed for magnesium, lithium, and sodium
concentrations by ICP-OES and EDTA titration (only for
magnesium).
Thepercentageextraction%E, percentage stripping % S,
distribution ratio D, and separation factor αare defined as
=
·
·+·×E
cV
cV cV
%
100%
org org
org org aq aq (2)
=
−
×S
cc
c
%
100%
org org,S
org (3)
=
D
c
c
org
aq
(4)
α
=D
D
A
B(5)
where corg and caq and Vorg and Vaq are concentrations and volumes of
the organic and the aqueous phase at extraction equilibrium,
respectively; corg,S is the concentration of the organic phase after
stripping; DAand DBare the distribution ratios of metals A and B,
respectively.
Extraction Rate. A series of reaction vials (4.0 mL) with 2.0 mL
of aqueous phase and 2.0 mL of organic phase were shaken on the
Thermo Scientific 2000 shaker at 300 rpm, and samples were taken
out and analyzed at a certain time interval. The loading profile of
metals as a function of time was established.
Batch Counter-Current Extraction Test. A batch counter-
current extraction (BCE) test was conducted to simulate a counter-
current multistage solvent extraction. The flowsheet can be found in
Figure 10. Up to 12 series of extraction (10 mL of aqueous phase and
17.5 mL (or 15 mL) of organic phase in each extraction) were carried
out to attain a stable state in the system. Each batch was shaken for 30
min, and the phase disengagement was accelerated by centrifugation
for 5 min at 4000 rpm. The last batches (nos. 11 and 12) were used to
determine the concentration of magnesium, lithium, and sodium in
the three extraction stages of the counter-current extraction test.
Mixer−Settler Test. Based on the results of the BCE test, the
feasibility to run the extraction in a continuous mode was tested using
a small mixer−settler battery (Rousselet Robatel, model UX 1.1).
Each mixer−settler stage is made of polytetrafluoroethylene (PTFE)
and has an effective volume of 35 mL for the mixer and 143 mL for
the settler. The settler is provided with a removable baffle and two
coalescence plates made also of PTFE to accelerate the phase
disengagement. A glass window at the end of the settling chamber
allows checking the O/A ratio in the settler. The position of the O/A
interphase in the settler is regulated by adjusting the height of a weir.
Two Masterflex L/S Cole-Parmer peristaltic pumps were used to
pump the aqueous and organic phases. For the extraction, the flow
rate was 1.7 and 1.0 mL min−1for the organic and aqueous phases,
respectively. On the basis of the flow rates and volumes of the mixer
and settler, the residence time in the mixer and settler is 13 and 53
min, respectively. The total residence time is sufficient for the
extraction to reach equilibrium. Samples of both the aqueous and the
organic phases were taken every hour during the operation of the
mixer−settler. The aqueous phases and the stripping solutions of the
organic phases were analyzed by ICP-OES.
EDTA Titration. EDTA titration was used for the determination of
the magnesium concentrations, in addition to measurement by ICP-
OES. The samples were buffered at pH = 10 using an NH3/NH4Cl
buffer. Eriochrome black T was used as an indicator. No interference
of lithium and sodium was expected during analysis based on the work
of Gao et al.
34
Samples were analyzed in triplicate.
Analytical Instruments. Magnesium, lithium, and sodium were
analyzed using a PerkinElmer Optima 8300 ICP-OES equipped with a
Scott crossflow nebulizer. A Bruker AVANCE NEO 400 nuclear
magnetic resonance spectroscopy device was used to record the 1H
NMR (400 MHz) and 13C NMR (100 MHz, decoupled) spectra. A
Thermo Scientific FLASH 2000 Elemental Analyzer was used for the
elemental analysis of the synthesized compounds.
Figure 1. Structures of (a) [A336][C272] and (b) [A336][V10]. The
main component of [A336] and [C272] is shown. For [V10], one
isomer of the mixture is shown.
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19227

■RESULTS AND DISCUSSIONS
Both the synthetic concentrated brine (106 g L−1Mg, 10 g L−1
Li) and synthetic original brine (15 g L−1Mg, 0.2 g L−1Li and
80 g L−1Na) were investigated for the removal of magnesium
by solvent extraction using binary extractants [A336][C272]
and [A336][V10].
Extraction with the [A336][C272] Extractant. The
extraction of magnesium from the concentrated brine solution
was tested with different concentrations of [A336][C272]
(0.1−0.8 mol L−1) dissolved in p-cymene. p-Cymene is a
bioderived solvent that is suitable as a diluent for solvent
extraction applications, especially suitable for diluting ionic
liquids because p-cymene is slightly polar.
35
Third phase
formation occurred during extraction when [A336][C272] was
>0.4 mol L−1(31 vol %, O/A = 1/1). Under this condition, the
loading of magnesium was only 3.8 g L−1, which is too low to
efficiently remove magnesium because too many extraction
stages would be needed. Therefore, [A336][C272] was not
tested further. A possible reason for the third phase formation
might be the formation of large complexes that are poorly
soluble in p-cymene, due to the synergistic effect between
[A336]+and [C272]−.
36−38
Extraction of Magnesium by [A336][V10] from the
Concentrated Brine. Extractant Concentration. The
influence of the [A336][V10] concentration on the extraction
efficiency of magnesium and lithium from the concentrated
brine was investigated (Figure 2). A concentration of
[A336][V10] up to 1.25 mol L−1(80 vol %) was used for
magnesium removal without the occurrence of a third phase,
although the solution was very viscous at the highest extractant
concentration.
The extraction of magnesium increased with increasing
[A336][V10] concentration. The molar ratio of loaded
magnesium to [A336][V10] gave an average value of (0.539
±0.005) over the studied concentration range. This ratio
means that the maximum loading of the extractant is achieved
and that two [A336][V10] molecules are used to extract one
Mg2+ cation, following eq 6.
V
[][]+
[] +[ ]
m
m
A336 V10 MCl
M V10 A336 Cl
m
m
org (aq)
(org) (org) (6)
where MClmrepresents the salts that are extracted and m=2
for magnesium extraction. This observation is consistent with
literature.
23,24,26
[A336][V10] (1.0 mol L−1, 64 vol %) in p-
cymene was chosen for further tests because this concentration
combines a high magnesium loading and operational viscosity.
Under this condition, about 13.5 g L−1magnesium was
extracted with coextraction of 0.5 g L−1lithium, corresponding
to 13% magnesium extraction and 5% lithium coextraction.
The separation factor of Mg/Li is only 2.8. The coextraction of
lithium also increased with increasing extractant concentration.
Although the extractant was fully loaded with magnesium, the
percentage extraction of magnesium was not high enough for
efficient removal of magnesium because of the feed solution
containing a too high magnesium concentration. Too many
stages would be needed to achieve quantitative extraction of
magnesium. Furthermore, the coextraction of lithium was too
high, resulting in a low lithium recovery at the end of the
process and an unsatisfactory separation of magnesium and
lithium. Several parameters were varied, including changing
phase ratio (O/A), changing diluents and scrubbing with
various solutions, to improve the separation of magnesium and
lithium, but the improvements were marginal (Figures S5−
S9). It is worth mentioning that NaCl was found to be able to
efficiently scrub lithium from the loaded organic phase, but
magnesium was also partly scrubbed (Figure S9).
Stripping the Loaded Organic Phase. The loaded organic
phase after extraction of magnesium from the concentrated
brine was stripped by both water and varying concentrations of
HCl to investigate the stripping performance (Figure 3).
Stripping by water is the reverse of extraction, that is, the
equilibrium in eq 6 is shifted from the right side to left side.
Water was able to strip 100% of lithium and 80% of
magnesium. Acids were more efficient for magnesium
stripping, and 100% magnesium was stripped by >0.5 mol
L−1HCl. It should be noted that to reuse the binary extractant,
Figure 2. Influence of the [A336][V10] concentration on the loading
and the percentage extraction of magnesium and lithium. The phase
ratio was 1/1.
Figure 3. Effect of the acid concentration on the stripping of
magnesium and lithium. The loaded organic phase composed of 13.5
gL
−1Mg and 0.5 g L−1Li. The phase ratio was 1/1.
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19228

no acid should be used for stripping because acids decompose
the binary extractant into the acidic extractant and the basic
extractant according to eq 1.
Extraction of Magnesium by [A336][V10] from the
Original Brine. Extractant Concentration. Because the
extraction characteristics of the concentrated brine led to a
low percentage extraction of magnesium (despite a full loading
of magnesium) and difficult scrubbing of the co-extracted
lithium, the synthetic original brine solution (15 g L−1Mg, 0.2
gL
−1Li and 80 g L−1Na) was investigated. Processing the
original brine directly would shorten the duration of the
process because the time-consuming solar evaporation can be
omitted, despite that larger volume of brine has to be
processed. The results for the percentage extraction of
magnesium, lithium, and sodium as a function of the
[A336][V10] concentration in p-cymene are shown in Figure
4. Extraction of magnesium, lithium, and sodium all increased
with the increasing concentration of [A336][V10] from 8.3,
0.7, and 0.1% at 0.15 mol L−1[A336][V10] to 60, 5.4, and
1.6% at 1.25 mol L−1[A336][V10], respectively. Under the
condition of 1.0 mol L−1[A336][V10], 54% magnesium was
loaded in a single contact with coextraction of 4.8% lithium
and 1.2% sodium (loading of magnesium, lithium, and sodium
can be found in Figure S10). The coextraction of sodium is not
a problem because lithium can be easily separated from sodium
by precipitation as Li2CO3. The loading of lithium is lower
than sodium, but the percentage extraction is higher. The
separation factor of Mg/Li increased from 2.8 for the
concentrated brine to 21.8 for the original brine using 1.0
mol L−1[A336][V10]. This is a significant improvement in the
separation of magnesium and lithium.
Sodium Concentration. The effect of NaCl concentration
on the extraction of metals can be found in Figure 5. With
increasing concentration of NaCl, extraction of magnesium
increased considerably from 28% without NaCl to 56% with 80
gL
−1Na, while extraction of lithium and sodium only
increased slightly (loading of metals is presented in Figure
S11). The enhanced extraction can be explained by the
common-ion effect (higher concentration of Cl−due to the
addition of NaCl). It is obvious that, according to eq 6,a
higher chloride concentration drives the reaction to the right.
Interestingly, the increase in lithium coextraction is the
smallest among the three metal ions when increasing the
NaCl concentration. Based on these results, a feed solution
containing a high NaCl concentration, and thus a high chloride
concentration, helps to increase the magnesium extraction
while maintaining similar coextraction of lithium.
Magnesium Concentration. The effect of the magnesium
concentration on the extraction of the metal ions was
investigated, and the results are presented in Figure 6. The
loading of magnesium in the organic phase increased
substantially with increasing MgCl2, from 0.91 g L−1at 1.0 g
L−1initial Mg in the feed to a maximum of 8.0 g L−1at 15 g
L−1initial Mg in the feed. On the contrary, extraction of both
lithium and sodium decreased with increasing magnesium
loading, from 0.06 and 7.6 to 0.01 and 1.1 g L−1, respectively,
which is because of the competition between magnesium and
lithium and sodium for the extractant. In other words, the
Figure 4. Percentage extraction of magnesium, lithium, and sodium
from the original brine (15 g L−1Mg, 0.2 g L−1Li and 80 g L−1Na)
using [A336][V10] in p-cymene. The phase ratio was 1/1.
Figure 5. Percentage extraction of Mg, Li, and Na, from feed solutions
containing 15 g L−1Mg and 0.2 g L−1Li and varying NaCl. The phase
ratio was 1/1.
Figure 6. Loading (a) and percentage extraction (b) of magnesium,
lithium, and sodium with the increasing MgCl2concentration in the
feed. Lithium and sodium concentration in the feed was kept at 0.2
and 80 g L−1. The phase ratio was 1/1.
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19229

loading of magnesium scrubs lithium and sodium, which is an
important observation.
In terms of percentage extraction, the values decreased for
all of the three metal ions. The decrease of percentage
extraction of magnesium from 97 to 54% is because
magnesium increased too much in the feed solution, although
the loading of magnesium also increased. However, the
decrease of lithium and sodium percentage extraction from
25 and 9.6 to 5.5 and 1.5%, respectively, is due to the decrease
in the absolute loading caused by the scrubbing effect of
magnesium.
Effect of the Phase Ratio. The phase ratio variation method
was used to construct a McCabe−Thiele diagram for the
extraction of metals from the original brine using
[A336][V10]. The results can be found in Figure 7 in terms
of loading and percentage extraction. [A336][V10] was never
fully loaded with magnesium even using a low O/A phase ratio.
An upper limit of 9.0 g L−1Mg loading was achieved at an O/A
phase ratio of 1/10 and 1/5, which is equivalent to about 75%
of the extractant loaded based on 1.0 mol L−1[A336][V10].
The incomplete loading of the extractant is because the driving
force (the Cl−concentration) is not high enough. The
percentage extraction of magnesium was poor when the O/A
phase ratio was less than 1/1. Increasing the phase ratio led to
a steep decrease in loading, but the percentage extraction
increased to 100% for an O/A phase ratio of 10/1.
Nevertheless, the coextraction of lithium and sodium also
increased steeply with increasing phase ratio, reaching a
maximum of 58 and 33%, respectively, at an O/A phase ratio
of 10/1. Based on these results, a compromise must be found
between a sufficiently high magnesium extraction and a
sufficiently low lithium co-extraction. Optimization of the
phase ratio was performed when constructing a McCabe−
Thiele diagram.
McCabe−Thiele Diagram. AMcCabe−Thiele diagram
could be constructed based on the results of magnesium
extraction with various O/A phase ratios. The McCabe−Thiele
diagram including the operating line can be found in Figure 8.
An O/A phase ratio between 1/1 and 2/1 should be chosen,
according to Figure 7b, because 1/1 phase ratio is not
sufficient to remove all of the magnesium (15 g L−1) with 1.0
mol L−1extractant, while the phase ratio of 2/1 would
coextract relatively high lithium. An O/A phase ratio of 1.75/1
was chosen and an operating line with a slope of 1.75/1 was
drawn in Figure 8 to estimate the number of stages required to
remove all the magnesium. Using this phase ratio, about 10%
of lithium would be coextracted (Figure 7b) in a single contact,
while 75% magnesium can be removed. By drawing horizontal
and vertical lines starting from the magnesium feed
concentration (15 g L−1) and using the operating line, it is
evident that three extraction stages would be necessary to
completely remove magnesium. This estimate was further
tested using a BCE test.
Extraction Rate. An experiment was carried out to
determine how long is needed to reach extraction equilibrium.
It was found that 20 min of shaking is enough to attain
equilibrium (Figure 9). In general, 30 min shaking was used
through all experiments to ensure that extraction equilibrium
was reached.
Validation of the Process. BCE Experiment. The
flowsheet of the three-stage BCE test is given in Figure 10,
and the results can be found in Figure 11. Sodium is not shown
for simplicity because the separation of sodium from lithium
and magnesium is easy as described above. In the first contact
of the fresh organic phase with the aqueous solution (EX 3),
50% of the lithium relative to the feed solution (determined to
be 14.8 g L−1Mg and 0.22 g L−1Li) was extracted. This high
percentage extraction of lithium can be explained by the
experiment with varying magnesium concentration in the feed
(Figure 6b). A low magnesium concentration in the feed leads
to more lithium coextraction. The magnesium concentration in
the aqueous solution of EX 3 before extraction contained only
about 3.0 g L−1Mg, which is similar to the investigated 3.0 g
L−1Mg concentration in Figure 6, where 19% of lithium was
extracted. The phase ratio 1.75/1 in EX 3 is higher than 1/1 in
Figure 7. Loading (a) and percentage extraction (b) of magnesium,
lithium, and sodium from the original brine (15 g L−1Mg, 0.2 g L−1Li
and 80 g L−1Na) in the function of phase ratio. 1.0 mol L−1
[A336][V10] in p-cymene was used as the organic phase.
Figure 8. McCabe−Thiele diagram for the counter-current extraction
of magnesium using three extraction stages with the phase ratio of O/
A = 1.75/1. Data in this diagram is based on the results of Figure 7.
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19230

Figure 6, explaining the even higher coextraction percentage of
lithium. In the subsequent extraction stages, part of lithium was
scrubbed from the organic phase due to the higher amount of
magnesium loaded into the organic phase, competing for the
extractant in EX 2 and EX 1. In the end, a loaded organic phase
containing 8.54 g L−1Mg and 0.011 g L−1Li was obtained (EX
1), while 0.05 g L−1Mg and 0.19 g L−1Li was left in the
aqueous raffinate (EX 3) after three-stage extraction.
Magnesium was almost completely removed (>99.5%) with
as low as 9% co-extraction of lithium, which is very good
separation.
Because the loading of magnesium scrubs the coextraction of
lithium, it is possible to further reduce the coextraction of
lithium by using a slightly smaller phase ratio, such as O/A =
1.5/1. A similar BCE test was carried out using a phase ratio
1.5/1 (Figure S12). After three extraction stages, the loaded
organic phase (EX 1) had 8.85 g L−1Mg and 0.010 g L−1Li,
and the raffinate (EX 3) had 1.41 g L−1Mg and 0.19 g L−1Li.
The coextraction of lithium was reduced to 7%. However, the
extraction of magnesium was also reduced to 90%. Considering
both the extraction of magnesium and coextraction of lithium,
the phase ratio of 1.75/1 is more efficient for the separation
process.
Scrubbing and Stripping. Scrubbing with NaCl. To
reduce the amount of the coextracted lithium from the loaded
organic phase obtained after the BCE, a NaCl solution was
used to scrub lithium. A scrub solution containing 80 g L−1Na
was chosen because the same concentration of sodium was
presented in the feed brine solution. Different phase ratios
were investigated for scrubbing with 80 g L−1Na (Figure 12).
With an increasing phase ratio, the scrubbing percentage for
both magnesium and lithium decreased. The scrubbing with a
1/1 phase ratio had the best scrubbing efficiency, with 29%
magnesium and 50% lithium scrubbed. However, the high loss
of magnesium during scrubbing is a serious disadvantage.
Therefore, scrubbing could not further improve the separation
process.
Phase Ratio for Stripping with Water. Different phase
ratios were used to investigate the stripping of the loaded
organic obtained after the BCE test (Figure 13a). Stripping
with 1.0 mol L−1HCl was used as a comparison (100%
stripping of magnesium and lithium). Lithium (100%) and
magnesium (62%) were stripped by water with an O/A phase
ratio of 1/1. An increase in the O/A phase ratio led to a
Figure 9. Apparent kinetics of magnesium and lithium extraction by
[A336][V10].
Figure 10. Flowsheet of the three-stage BCE test. The organic phase
(O), the aqueous feed (F), the loaded organic phase (LO) and the
raffinate (R) are abbreviated in the scheme.
Figure 11. Concentration profiles of magnesium and lithium in the
aqueous phase (a) and in the organic phase (b), and percentage
extraction of magnesium and lithium relative to feed solution (c),
during the three-stage BCE. The phase ratio O/A was 1.75/1.
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19231

reduction in stripping efficiency for both lithium and
magnesium. An O/A phase ratio of 3/1 seems to be a good
compromise between stripping efficiency and consumption of
water.
Multistage Stripping with Water. An O/A phase ratio of 3/
1 was further tested for multistage cross-current stripping using
water. Four stripping stages were investigated. In the first three
stages, water was used as the stripping agent, followed by
complete stripping of the remaining magnesium, lithium, and
sodium using 1.0 mol L−1HCl (for analysis purpose). The
mass balance was used to calculate the percentage stripping,
and the results are shown in Figure 13b. The amounts stripped
can be found in Figure S14. The stripping percentages are
cumulative, and the sum is 100%. In the first stage, 90% of the
loaded lithium and 46% of the loaded magnesium were
stripped. In the second stage, most of the remaining lithium
was stripped, together with 22% of magnesium. In the
subsequent stages, no additional amounts of lithium were
stripped because it had already been completely stripped in the
first two stages. 13% magnesium was stripped in the third stage
and complete stripping of magnesium was achieved using 1.0
mol L−1HCl in the last stage. In short, 81% magnesium can be
stripped in three stages using water with an O/A phase ratio of
3/1, which is higher than a single-stage stripping using water
with an O/A phase ratio of 1/1 (62%). These two stripping
approaches consume the same amount of water. The stripping
of magnesium regenerates [A336][V10], which in principle
can be used for the next extraction cycle.
Mixer−Settler Experiments. To check the practical
potential of the separation of lithium and magnesium using
[A336][V10], mixer−settler experiments were carried out.
Pictures of the setup are given in Figure S15. Only the
extraction step (not including scrubbing or stripping) was
studied in mixer−settlers because it represents the core of the
separation process presented in this work. In the mixer−settler
test, coalescence plates were used in the settler to aid the
separation of the phases. Good phase disengagement was
achieved in each of the extraction stages and formation of
emulsions or the third phase was not observed during the total
duration of the experiment.
After running the experiment for 7 h, 94% magnesium was
extracted with 11% lithium co-extraction. The results are
presented in Figure 14 and they are consistent with the results
obtained for the batch counter-current test (Figure 11). The
results show that the aqueous lithium concentration remained
practically constant in the extraction stages EX 1 and EX 2 and
slightly decreased in stage EX 3. The aqueous magnesium
concentration decreased with the increasing number of
extraction stages: from 14.51 g L−1in the feed until 0.19 g
L−1Mg left in the raffinate at EX 3, with >98% removal of
magnesium. The loading of magnesium in the organic phase
increased from 1.32 g L−1at EX 3 to 7.81 g L−1at EX 1. While
the loading of lithium in the organic phase decreased from
0.032 g L−1at EX 3 to 0.014 g L−1at EX 1, due to the
scrubbing effect by loaded magnesium.
After 7 h of extraction, the system was shut down and
restarted again the next day. Analysis of the samples taken at
the 15th h show that 97% of magnesium had been extracted
with 13% of lithium coextraction (Figure S16). These results
prove that the binary extractant [A336][V10] extracts
magnesium efficiently and selectively from the lithium-
containing brine solutions and the process can be safely and
easily carried out in a continuous mode. The complete metal
loading profile as a function of time is given in Figure S17.
Lithium in the extraction raffinate after removal of
magnesium could be concentrated by a synergistic solvent
extraction system containing a beta-diketone (e.g., LIX 54) and
a neutral ligand (e.g., Cyanex 923). Plenty of studies have
investigated this extraction system.
39,40
The loaded lithium in
the organic phase after extraction from the raffinate can be
stripped and concentrated in an aqueous solution, which can
subsequently be treated by Na2CO3to produce Li2CO3. The
concentration of lithium by solvent extraction would require
base and acids for adjustment of pH. However, the amount of
chemicals required is much smaller than the precipitation of
magnesium in the existing process. Besides, the MgCl2
Figure 12. Effect of phase ratio on the percentage scrubbing (% Scr)
of the loaded organic phase obtained after the BCE.
Figure 13. Percentage stripping of magnesium and lithium with HCl
solution or water using different phase ratios in single-stage stripping
(a); percentage stripping of magnesium and lithium with water (ST 1,
ST 2 and ST 3) and 1.0 mol L−1HCl (ST 4) with an O/A phase ratio
of 3/1 in multistage cross-current stripping (b).
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19232

obtained in the new process has a high purity (>99%) and
hence might be a marketable byproduct.
Sustainability of the Process. The proposed new process
for separation of magnesium and lithium by selective removal
of magnesium using the binary extractant [A336][V10] is
more sustainable than the state-of-the-art processes. The
sustainability of the new process includes (1) no consumption
of bases during extraction, (2) no consumption of acids for
stripping of magnesium from the loaded organic phase, (3)
highly efficient separation of magnesium from lithium, with
almost complete removal of magnesium and about 10% co-
extraction of lithium, and (4) no generation of waste (such as
Mg(OH)2by precipitation). Besides, the process is able to deal
with the original brine directly, largely shortening the duration
of the process because the slow preconcentration of brine by
solar evaporation is not necessary.
■CONCLUSIONS
The binary extractant [A336][V10] was investigated for the
removal of magnesium from both the (synthetic) concentrated
brine (106 g L−1Mg and 10 g L−1Li) and the (synthetic)
original brine (15 g L−1Mg, 0.2 g L−1Li and 80 g L−1Na) with
the purpose of separating magnesium and lithium in a more
sustainable way than conventionally. The separation was poor
for the concentrated brine due to too high magnesium
concentrations, whereas excellent separation was achieved for
the original brine, with almost complete removal of magnesium
and about 10% coextraction of lithium in a three-stage counter-
current extraction. The process was further demonstrated in a
mixer−settler battery in a continuous mode and the results
were consistent with the BCE. After loading, 81% magnesium
and 100% lithium in the loaded solution can be stripped in
three cross-current stages using water with an O/A phase ratio
of 3/1. The removal of magnesium by [A336][V10] does not
consume acid or base, while having both high extraction
efficiency and high selectivity. The new process has the
potential for producing Li2CO3from brines in a more
sustainable manner.
■ASSOCIATED CONTENT
*
SSupporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acssusche-
meng.9b05436.
Characterizations of [A336][V10] and [A336][C272];
properties of [A336][V10]; more tests data on
magnesium extraction from concentrated brine; loading
of magnesium, lithium and sodium in the organic phase;
scrubbing and stripping of magnesium and lithium under
various conditions; batch counter-current tests with O/
A ratio of 1.5; and results of mixer−settler test (PDF)
■AUTHOR INFORMATION
Corresponding Author
*E-mail: zheng.li@kuleuven.be.
ORCID
Zheng Li: 0000-0002-7882-5999
Xiaohua Li: 0000-0003-4555-8705
Sofía Riaño: 0000-0002-1049-6156
Koen Binnemans: 0000-0003-4768-3606
Notes
The authors declare no competing financial interest.
■ACKNOWLEDGMENTS
The research was supported by the European Research
Council (ERC) under the European Union’s Horizon 2020
Research and Innovation Programme: grant agreement
694078solvometallurgy for critical metals (SOLCRIMET).
■REFERENCES
(1) Tollefson, J. Car industry: charging up the future. Nature 2008,
456, 436−440.
(2) Coonan, T. G. Lithium Use in Batteries: U.S. Geological Survey
Circular 1371, 2012; p 14. http://pubs.usgs.gov/circ/1371/.
(3) Yoshio, M.; Brodd, R.; Kozawa, A. Lithium-Ion Batteries: Science
and Technologies; Springer-Verlag: New York, NY, 2009.
(4) Etacheri, V.; Marom, R.; Elazari, R.; Salitra, G.; Aurbach, D.
Challenges in the development of advanced Li-ion batteries: a review.
Energy Environ. Sci. 2011,4, 3243−3262.
(5) Fröhlich, P.; Lorenz, T.; Martin, G.; Brett, B.; Bertau, M.
Valuable metalsrecovery processes, current trends, and recycling
strategies. Angew. Chem., Int. Ed. 2017,56, 2544−2580.
(6) U.S. Geological Survey. Mineral Commodity Summaries 2018:
U.S. Geological Survey 2018, 2018; p 200.
Figure 14. Metal loading profiles of magnesium and lithium in the
aqueous phase (a) and in the organic phase (b), and percentage
extraction of magnesium and lithium relative to the feed solution (c),
during the three-stage counter-current mixer−settler experiment at
the 7th h. The phase ratio O/A was 1.7/1.
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19233

(7) Speirs, J.; Contestabile, M.; Houari, Y.; Gross, R. The future of
lithium availability for electric vehicle batteries. Renewable Sustainable
Energy Rev. 2014,35, 183−193.
(8) Loganathan, P.; Naidu, G.; Vigneswaran, S. Mining valuable
minerals from seawater: a critical review. Environ. Sci.: Water Res.
Technol. 2017,3,37−53.
(9) Kesler, S. E.; Gruber, P. W.; Medina, P. A.; Keoleian, G. A.;
Everson, M. P.; Wallington, T. J. Global lithium resources: Relative
importance of pegmatite, brine and other deposits. Ore Geol. Rev.
2012,48,55−69.
(10) Swain, B. Recovery and recycling of lithium: A review. Sep.
Purif. Technol. 2017,172, 388−403.
(11) Cabeza, L. F.; Gutierrez, A.; Barreneche, C.; Ushak, S.;
Fernández, Á. G.; Inés Fernádez, A.; Grágeda, M. Lithium in thermal
energy storage: A state-of-the-art review. Renewable Sustainable Energy
Rev. 2015,42, 1106−1112.
(12) Meshram, P.; Pandey, B. D.; Mankhand, T. R. Extraction of
lithium from primary and secondary sources by pre-treatment,
leaching and separation: A comprehensive review. Hydrometallurgy
2014,150, 192−208.
(13) Choubey, P. K.; Kim, M.-S.; Srivastava, R. R.; Lee, J.-C.; Lee, J.-
Y. Advance review on the exploitation of the prominent energy-
storage element: Lithium. Part I: From mineral and brine resources.
Miner. Eng. 2016,89, 119−137.
(14) Yu, J.; Zheng, M.; Wu, Q. Research progress of lithium
extraction process in lithium-containing salt lake. Chem. Ind. Eng. Prog.
2013,32,13−21.
(15) Zhou, Y.; Li, L.; Wu, Z.; Li, X. Exploitation and Comprehensive
Utilization for Qinghai Salt Lakes. Prog. Chem. 2013,25, 1613−1624.
(16) Su, H.; Li, Z.; Zhu, Z.; Wang, L.; Qi, T. Extraction relationship
of Li+and H+using tributyl phosphate in the presence of Fe(III). Sep.
Sci. Technol. 2019,1−9.
(17) Nelli, J. R.; Arthur, T. E., Jr. Recovery of Lithium from Bitterns.
U.S. Patent 3,537,813 A, 1968.
(18) Zhou, Z.; Qin, W.; Liang, S.; Tan, Y.; Fei, W. Recovery of
lithium using tributyl phosphate in methyl isobutyl ketone and FeCl3.
Ind. Eng. Chem. Res. 2012,51, 12926−12932.
(19) Wang, Y.; Liu, H.; Fan, J.; Liu, X.; Hu, Y.; Hu, Y.; Zhou, Z.;
Ren, Z. Recovery of Lithium Ions from Salt Lake Brine with a High
Magnesium/Lithium Ratio Using Heteropolyacid Ionic Liquid. ACS
Sustainable Chem. Eng. 2019,7, 3062−3072.
(20) Zhang, G.; Qin, W.; Tan, Y.; Dai, Y. Separation of magnesium
and lithium by solvent extraction using di-(2-ethylhexyl) phosphoric
acid (D2EHPA). Qinghua Daxue Xuebao 2010,50, 430−433.
(21) Virolainen, S.; Fallah Fini, M.; Miettinen, V.; Laitinen, A.;
Haapalainen, M.; Sainio, T. Removal of calcium and magnesium from
lithium brine concentrate via continuous counter-current solvent
extraction. Hydrometallurgy 2016,162,9−15.
(22) Grinstead, R. R.; Davis, J. C.; Lynn, S.; Charlesworth, R. K.
Extraction by phase separation with mixed ionic solvents. Ind. Eng.
Chem. Prod. Res. Dev. 1969,8, 218−227.
(23) Grinstead, R. R.; Davis, J. C. Extraction by phase separation
with mixed ionic solvents. Recovery of magnesium chloride from sea
water concentrates. Ind. Eng. Chem. Prod. Res. Dev. 1970,9,66−72.
(24) Hanson, C.; Hughes, M. A.; Murthy, S. L. N. Extraction of
magnesium chloride from brines using mixed ionic extractants. J.
Inorg. Nucl. Chem. 1975,37, 191−198.
(25) Hernandez, R. A.; Martinez, J. M. Extraction of lithium and
magnesium salts with a mixed ionic extractant. Ind. Eng. Chem. Process
Des. Dev. 1981,20, 698−704.
(26) Eyal, A. M.; Bressler, E.; Bloch, R.; Hazan, B. Extraction of
metal salts by mixtures of water-immiscible amines and organic acids
(acid-base couple extractants). 1. A review of distribution and
spectroscopic data and of proposed extraction mechanisms. Ind. Eng.
Chem. Res. 1994,33, 1067−1075.
(27) Eyal, A. M.; Kogan, L.; Bressler, E. Extraction of metal salts by
mixtures of water-immiscible amines and organic acids (acid-base
couple extractants). 2. Theoretical treatment and analysis. Ind. Eng.
Chem. Res. 1994,33, 1076−1085.
(28) Kholkin, A. I.; Pashkov, G. L.; Fleitlikh, I. Y.; Belova, V. V.;
Luboshnikova, K. S.; Sergeev, V. V.; Kopanev, A. M.; Kulmuhamedov,
G. K. Application of binary extraction in hydrometallurgy. Hydro-
metallurgy 1994,36, 109−125.
(29) Pospiech, B.; Chagnes, A. Highly selective solvent extraction of
Zn (II) and Cu (II) from acidic aqueous chloride solutions with
mixture of Cyanex 272 and Aliquat 336. Sep. Sci. Technol. 2015,50,
1302−1309.
(30) Liu, Y.; Zhu, L.; Sun, X.; Chen, J. Toward greener separations
of rare earths: bifunctional ionic liquid extractants in biodiesel. AIChE
J. 2010,56, 2338−2346.
(31) Kumari, A.; Sinha, M. K.; Sahu, S. K.; Pandey, B. D. Solvent
extraction and separation of trivalent lanthanides using Cyphos IL
104, a novel phosphonium ionic liquid as extractant. Solvent Extr. Ion
Exch. 2016,34, 469−484.
(32) Swain, B.; Jeong, J.; Lee, J.; Lee, G.-H. Separation of cobalt and
lithium from mixed sulphate solution using Na-Cyanex 272.
Hydrometallurgy 2006,84, 130−138.
(33) Bradaric, C. J.; Downard, A.; Kennedy, C.; Robertson, A. J.;
Zhou, Y. Industrial preparation of phosphonium ionic liquids. Green
Chem. 2003,5, 143−152.
(34) Gao, J.; Guo, Y.; Wang, S.; Deng, T.; Chen, Y.-W.; Belzile, N.
Interference of Lithium in Measuring Magnesium by Complexometry:
Discussions of the Mechanism. J. Chem. 2013,2013,1−4.
(35) Li, Z.; Smith, K. H.; Stevens, G. W. The use of environmentally
sustainable bio-derived solvents in solvent extraction applicationsa
review. Chin. J. Chem. Eng. 2016,24, 215−220.
(36) Torkestani, K.; Goetz-Grandmont, G. J.; Brunette, J. P.
Synergistic extraction of cadmium and zinc from nitrate medium with
3-phenyl-4-benzoylisoxazol-5-one in the presence of methyl-tri-n-
octylammonium nitrate in chloroform and toluene. Solvent Extr. Ion
Exch. 1997,15, 819−835.
(37) Atanassova, M.; Dukov, I. L. Synergistic solvent extraction and
separation of trivalent lanthanide metals with mixtures of 4-benzoyl-3-
methyl-1-phenyl-2-pyrazolin-5-one and aliquat 336. Sep. Purif.
Technol. 2004,40, 171−176.
(38) Zhu, M.; Zhao, J.; Li, Y.; Mehio, N.; Qi, Y.; Liu, H.; Dai, S. An
ionic liquid-based synergistic extraction strategy for rare earths. Green
Chem. 2015,17, 2981−2993.
(39) Pranolo, Y.; Zhu, Z.; Cheng, C. Y. Separation of lithium from
sodium in chloride solutions using SSX systems with LIX 54 and
Cyanex 923. Hydrometallurgy 2015,154,33−39.
(40) Swain, B. Separation and purification of lithium by solvent
extraction and supported liquid membrane, analysis of their
mechanism: a review. J. Chem. Technol. Biotechnol. 2016,91, 2549−
2562.
ACS Sustainable Chemistry & Engineering Research Article
DOI: 10.1021/acssuschemeng.9b05436
ACS Sustainable Chem. Eng. 2019, 7, 19225−19234
19234