Separation and Purification Technology 274 (2021) 119051

Contents lists available at ScienceDirect

Separation and Purification Technology

journal homepage: www.elsevier.com/locate/seppur

Selective separation of lithium from the high magnesium brine by the

extraction system containing phosphate-based ionic liquids
Ruibing Bai a, b, c, Junfeng Wang a, b, *, Daoguang Wang a, b, Yanqiang Zhang a, b, Junjie Cui a, b
a
Beijing Key Laboratory of Ionic Liquids Clean Process/State Key Laboratory of Multiphase Complex Systems, Institute of Process Engineering, Chinese Academy of
Sciences, Beijing 100190, PR China
b
Innovation Academy for Green Manufacture, Chinese Academy of Sciences, Beijing 100190, PR China
c
Sino-Danish College, University of Chinese Academy of Sciences, Beijing 100049, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: The high acidity of stripping is one of the major problems that hinder the industrialization of the tributyl
Ionic liquid phosphate (TBP) + FeCl3 system for lithium extraction from the salt lake brines with high Mg/Li ratio. In this
Solvent extraction study, a novel solvent extraction system, trialkylmethylammonium di(2-ethylhexyl)orthophosphinate ([N1888]
Lithium
[P507]) + TBP + FeCl3, was developed to significantly reduce the difficulty of stripping and simplify the
Stripping
High Mg brine
extraction process. The conditions of extraction and stripping were systematically studied in the present work.
The results showed that Li+ extraction efficiency could reach 61% under the optimal extraction parameters and
Li+ stripping ratio was 90% with a relative mild conditions of O/A = 10:1, [HCl]aq = 1.5 mol⋅L–1. Also, combined
with experimental results, FT-IR analysis, and UV analysis, the possible mechanisms of lithium extraction and
stripping were put forward. Li+ was extracted from the aqueous phase through the formation of [Li⋅2TBP][FeCl4]
adduct, meanwhile [N1888][P507] was just slightly involved in the extraction step. In the stripping procedure, the
interaction between [FeCl4]− and [Li⋅2TBP]+ might be destroyed, and new possible complexes, [Li⋅2TBP]+, and
[FeCl4⋅TBP⋅P507]2− /[FeCl4⋅P507]2− were formed in the organic phase. Additionally, the extraction efficiency of
Li+ remained above 70% while the organic phase was reused for six cycles. Compared with the traditional
system, the addition of [N1888][P507] resulted in a slight decrease of Li+ extraction efficiency (about 5%), the
stripping acidity significantly reduced from 6 mol⋅L–1 HCl to 1.5 mol⋅L–1 HCl. This new extraction system showed
a promising performance in selective extracting and easy stripping in lithium separation from the salt lake brines
with high Mg/Li ratio.

1. Introduction
Lithium has been regarded as one kind of important strategic mate­
rial internationally for wide application in the new energy, nuclear
fusion, pharmaceuticals, modern materials industry, and other in­
dustries [1–5]. In recent years, lithium consumption has been expo­
nentially increased, driven by the demands for lithium-ion batteries [6].
In 2019, global lithium consumption was 57,700 tons, of which the
lithium battery industry accounted for 65% consumption [7]. Besides,
the stable isotopes of lithium, 6Li and 7Li, play critical roles in nuclear
fusion reactions, e.g., 7Li often used as neutron moderator and heat
transfer agent in the nuclear fusion reactors [8]. Therefore, lithium is the
primary player in the energy industry of the 21st century.
Lithium resources mainly store in brine and solid lithium mines
reservoirs [9]. Currently, lithium products from solid mines cannot meet
the current market demand [1,10]. Meanwhile, 69% of the world’s total
lithium reserves are distributed in brines which possesses about 55.2
million tons of lithium resources [7,11]. So lithium extraction from
brines has become an inevitable project for obtaining lithium products.
The main process difficulty is the brine’s chemical property of high Mg/
Li ratio and low Li+ concentration, which leads to the high extraction
costs. At present, the main methods for lithium extraction from salt lake
brines are calcination leaching [10,12], adsorption [13,14], precipita­
tion [15,16], nanofiltration [17–19], and electrolysis [20,21], solvent
extraction [22–25], etc. Among these methods, the solvent extraction
was considered to be a most promising technique for extracting lithium

* Corresponding author at: Beijing Key Laboratory of Ionic Liquids Clean Process/State Key Laboratory of Multiphase Complex Systems, Institute of Process En­
gineering, Chinese Academy of Sciences, Beijing 100190, PR China.
E-mail address: junfwang@ipe.ac.cn (J. Wang).

https://doi.org/10.1016/j.seppur.2021.119051
Received 24 February 2021; Received in revised form 23 May 2021; Accepted 28 May 2021
Available online 3 June 2021
1383-5866/© 2021 Published by Elsevier B.V.
R. Bai et al.
from high Mg/Li ratio brine because of its high selectivity, short equil­
ibration time, and low chemical and energy consumption.
In 1970, Nelli and Arthur [26] first proposed the Tributyl phosphate
(TBP) + FeCl3 system that performed a high separation factor of lithium
to magnesium in the lithium extraction from chloride media. After that,
numerous works have been carried out to extract lithium from high Mg/
Li ratio brine by using the TBP + FeCl3 based system due to its good
selective for lithium (separation factor >300), and low cost of the pro­
cess [22–24,27–31]. However, these processes suffer from the usage of
high HCl concentration (6–9 mol⋅L–1) for lithium stripping [32]. The
mechanism of the extraction and stripping obeyed the following
equations:
Li+ −
aq + 2TBPorg + FeCl4,aq ⇄LiFeCl4 ⋅2TBPorg (1)
+
Haq + LiFeCl4 ⋅2TBPorg ⇄Li+
aq + HFeCl4 ⋅2TBPorg (2)
The highly concentrated HCl corrodes the equipment and requires a
large amount of base for the organic phase’s neutralization and regen­
eration. Additionally, TBP can be hydrolyzed easily in the repeating
process of stripping and oil phase saponification. Therefore, the TBP +
FeCl3 based system cannot be widely used for industrial applications.
Companies would use other lithium extraction methods (adsorption or
membrane separation methods) instead of solvent extraction.
Compared with volatile organic solvents, ionic liquid (IL) is a kind of
promising solvent with low saturated vapor pressure and suitable for
liquid-liquid extraction. Additionally, due to its advantages of excellent
ion-exchange capabilities, unique pure ionic solvation environments,
and ion-recognition capabilities of complexing ligands [33,34], ILs bring
new opportunities to solve the difficult stripping problem in traditional
solvent extraction. 1-alkyl-3-methylimidazolium hexafluorophosphate
([CnMIm][PF6]), was used to replace the kerosene as a green medium to
form an IL + TBP + FeCl3 system to extract lithium from brine [35].
Under optimal conditions of 9:1 (v/v) for TBP/IL, 0.03 mol⋅L–1 HCl, O/A
of 1:1 and Fe/Li of 2:1, single extraction and back-extraction efficiencies
were 87% and 90%, respectively. However, the acidity of the back-
extraction was still 6.0 mol⋅L–1 HCl for above system [35]. Since then,
ILs, such as [Cnmim][PF6], [Cnmim][NTf2], [HOEmim][NTf2], were
used as the co-extractants to form an IL-TBP system to extract lithium
from high Mg brine in some studies, where it was found that the highest
Li+ extraction ratio reached 95%, and the single-stage separation factor
of Li/Mg reached 512.5 [36–41]. These studies also obtained a reduced
stripping acidity from 6.0 mol⋅L–1 HCl to 0.5 mol⋅L–1 HCl [36–41].
Meanwhile, these ILs were easily lost to water phase due to the cation
exchange process during extraction and their structures are destroyed.
Although the lost ionic liquids could be effectively recycled, the cost of
the extraction process with such system would be too high at an in­
dustrial scale. Functionalized ionic liquids, such as tetrabutylphospho­
nium bis(2,4,4-trimethylpentyl)phosphinate ([P4444][BTMPP]) [42],
tetrabutylammonium bis(2-ethylhexyl)-phosphate ([N4444][DEHP]) and
tetraoctylammonium bis(2-ethylhexyl)-phosphate ([N8888][DEHP])
[43], tetrabutylammonium 2-ethylhexyl hydrogen-2-
ethylhexylphosphonate ([N4444][EHEHP]) [44,45], etc. were synthe­
sized and used for lithium extraction. Results shown that above 98% Li
was stripped with 0.5 mol⋅L–1 HCl. Unfortunately, these functionalized
ILs systems could only effectively separate lithium from alkali metal
mixed solution (LiCl, NaCl, KCl, RbCl, and CsCl). Recently, a new ternary
synergistic solvent extraction system composed by TBP, FeCl3, and 2-
ethylhexyl hydrogen -2-ethylhexylphosphonate (P507) was developed.
With the aid of the new system, the use of pure water instead of the high
concentration of HCl was possible in the lithium stripping process [47].
The results also indicated that the addition of trialkylmethylammonium
di(2-ethylhexyl)orthophosphinate ([N1888][P507]) as the co-extractant
to the system might be more efficient for simplifying the Li+ stripping
with keeping the extraction efficiency as high as possible. Encouraged
by above results, this work will be carried out to explore the extraction
2
Separation and Purification Technology 274 (2021) 119051
of Li+ with [N1888][P507] deeply to promote the industrial application of
[N1888][P507]-based system.
In this work, phosphate-based IL, [N1888][P507], was selected and
added into the TBP + FeCl3 system as a co-extractant to solve the high
stripping acidity problem and maintain the stability of the extraction
system. The experiment condition of extraction and stripping were
examined systemically, and the regeneration process was optimized.
The mechanisms of lithium extraction and stripping by using this co-
extraction system were studied in detail. And the structures of extrac­
ted complexes were analyzed using the slope method, spectroscopic
characterization (FT-IR and UV–visible). Additionally, the reusability of
the organic phase in this phosphate-based IL contained system was
tested for 6 extraction cycles. All of the results showed that the newly
phosphate-based IL contained system was easy-stripping for lithium ions
from uploaded the organic phase and promising for selective lithium
extraction from brines with high Mg/Li ratio.
2. Experimental
2.1. Reagents
LiCl⋅H2O (>97%), MgCl2⋅6H2O (>99%), FeCl3⋅6H2O (>98%), 2-eth­
ylhexyl hydrogen-2-ethylhexylphosphonate (P507) (>95%), Tetraocty­
lammonium bromide ([N8888]Br, 98%) were purchased from Shanghai
Macklin Biochemical Co., Ltd. Methyltributylammonium chloride
([N1444]Cl, 98%), Tetrabutylammonium chloride hydrate ([N4444]
Cl⋅xH2O, 99%), tri-n-octylmethylammonium chloride ([N1888]Cl, 98%)
were obtained from Shanghai Energy Chemicals Co. Ltd. Tributyl
phosphate (TBP) (>98.5%), and Kerosene (MW = 142–254) were pro­
vided by Aladdin Chemical Co., Ltd. AgNO3 (99%) and HNO3 (68%)
were purchased from Sinopharm Group. Additionally, HCl (37 wt%) was
purchased from Beijing Chemical Works. The simulated brine solutions
with high Mg concentration were prepared by dissolving chloride salts
into the deionized water with certain amounts according to the real
brine solution’s composition. Fig. 1 shows the structures and their ab­
breviations of different FILs and TBP used in the present work.
2.2. Synthesis of phosphate-based ionic liquid
Phosphate-based ionic liquids were prepared via ion-exchange and
neutralizing reactions [46]. The ethanol solution of [Nxxxx]OH was
prepared from [Nxxxx]Cl (Br) (0.025 mol) using an Ambersep 900 (OH)
anion exchange resin (exchange efficiency = 33% based on the calcu­
lation). P507 (7.66 g, 0.025 mol) was added to the [Nxxxx]OH solution.
The mixture was then stirred at room temperature for 24 h till the so­
lution became neutral. Ethanol and water were distilled via rotary
evaporation, and the product was dried at 70 ◦ C under vacuum over­
night. [Nxxxx][P507] was shadow yellow viscous liquid. The detailed
synthesis methods and ILs NMR characterization were provided in the
supporting information.
2.3. Extraction procedure
The synthetic brine used in this work containing 0.10 mol⋅L–1 LiCl,
4.0 mol⋅L–1 MgCl2, 0.13 mol⋅L–1 FeCl3, and 0.03 mol⋅L–1 HCl, which
closely simulated the composition of concentrated West Tajinar Salt
Lake brines (China) [47]. The Fe/Li ratio of 1.3/1 was selected on the
basis of the reference [47]. And HCl (0.03 mol⋅L–1) in the aqueous phase
was used to prevent the hydrolysis of Fe3+. The organic phase contained
a certain concentrations of TBP and [N1888][P507] with kerosene as
dilution. The synthetic brine was mixed with the organic phase in a glass
bottle using a heating magnetic stirrer (HS-7, IKA, German) with an O/A
ratio of 1:1 (5.0 mL/5.0 mL) at 25 ± 0.5 ◦ C for 20 min (the equilibrium
time was discussed in the supporting information). After stirring, the
mixtures were centrifuged for 10 min (5000 rpm) for phase separation.
Subsequently, aqueous or organic samples were taken from the bottom
R. Bai et al. Separation and Purification Technology 274 (2021) 119051

Fig. 1. Structure and abbreviation of principal components in this work.

phases by using a syringe to determine metal concentrations. The con­ were analyzed by inductively coupled plasma–atomic emission spec­
centrations of Li+ and Mg2+ in organic phases were determined after trometry (ICP-AES, ICPE-9000, Shimadzu, Japan). The mean relative
stripping with 6 mol⋅L–1 HCl. Furthermore, [Fe3+]org was calculated by deviation of concentration analysis by ICP-AES was within 5%. The
mass balance based on the aqueous concentration because Fe3+ could extraction complexes (like [FeCl4]− ) in the organic phase was examined
not be stripped by concentrated HCl. by using Ultraviolet–visible spectrophotometer (UV-2550; Shimadzu
The extraction efficiency of different metal ions E%, stripping per­ Corporation, Japan) to evaluate the relative content of [FeCl4]− formed
centage S%, distribution ratio D, and separation factor β were defined as after mixed with the aqueous phase with different concentration of
MgCl2. The UV curves were obtained by wave scanning from 100 to 900
Corg Vorg
E/% = × 100% (3) nm with 0.5 nm of resolution [47]. The Fourier Transform Infrared
Corg Vorg + Caq Vaq
Spectroscopy (FT-IR) spectrophotometer (Nicolet iS50; Thermo Nicolet
Corg − Corg,s Corporation, USA) was employed to examine the organic phase for
S/% = × 100% (4) extraction complexes analysis. The spectra were performed in the mid-
Corg
infrared region (4000–400 cm− 1) using a Thermal Nicolet 380 spec­
Corg,M trometer with a resolution of 1 cm− 1 [25].
DM = (5)
Caq,M
3. Results and discussion
DLi
βLi/Mg = (6) 3.1. Effect of ILs cations
DMg

where Corg and Caq (g⋅L–1) and Vorg and Vaq (mL) were equilibrated
concentration of metal ions and volumes of the organic and the aqueous
phase after extraction, respectively; Corg,s was the concentration of metal
ions in the organic phase after stripping; DLi and DMg were the distri­
bution ratios of Li+ and Mg2+, respectively.
Different IL cations with changing carbon chain length, i.e., [N1444]
[P507], [N4444][P507], [N1888][P507], and [N8888][P507] were investi­
gated to determine the effect of IL hydrophobicity on the extraction
efficiency of Li+. The experimental result was shown in Fig. 2, in which
the extraction efficiencies of Li+ with the different IL-based extraction
systems was in range from 58% to 60%. Mg2+ extractions were really

2.4. Reusability of the organic phase

The experiments for evaluating the organic phase reusability was

subject to the following procedure: (1) Extraction: the fresh organic
phase was mixed with the initial brine phase, under the conditions of O/
A = 1:1, temperature = 25 ◦ C, extraction time = 30 min; (2) Stripping:
Li+ stripping efficiency was 87.5% and that of Mg2+ 90.4% under the
conditions of O/A = 10:1 and temperature = 25 ◦ C for one single-stage.
For the stripping of Li+, the loaded organic phase was stripped with the
1.5 mol⋅L–1 HCl solution for three times under the conditions of O/A =
10:1 and temperature = 25 ◦ C. (3) The stripped organic phase was mixed
with the fresh brine in the next extraction cycle. Then the extraction/
stripping cycle was repeated for the following five times.

2.5. Characterization

The concentrations of Li+, Mg2+, and Fe3+ in the aqueous samples Fig. 2. Effect of IL cations on the extraction of metal ions.

3
R. Bai et al.
limited, for which the extraction efficiency was maintained at a low level
of about 1%. Generally, the lithium selectivity slightly decreased as IL
hydrophobicity increased. The main reason was that a lithium com­
plexes formed in the organic phase showing a particular weak polarity
was more stable structurally in a relatively low hydrophobicity envi­
ronment. Therefore, Li+ extraction became slightly difficult as the hy­
drophobicity increased from [N1444][P507] to [N8888][P507], and βLi/Mg
was also decreased slightly. Combined the Li+ extraction capability and
hydrophobicity of the ILs, [N1888][P507] was selected as the co-
extractant for the further investigation. Additionally, the result was
compared with that of P507-based systems [47]. It was shown that P507
slightly suppresses the extraction of Li+, as the extraction of Li+ slightly
decreases with increasing P507 concentration. When [P507]org = 0.025
mol⋅L–1 (4–5 vol%), Li+ extraction efficiency was 65%, and Mg2+
extraction efficiency was ~2%. Compared with IL-based systems, the
P507-contained system has a relative higher Li+ extraction efficiency, but
for decreasing the stripping acidity, the P507 concentration need to reach
0.125 mol⋅L–1 (20 vol%). Under this condition, Li+ extraction efficiency
was only 57%. The results also indicated that IL anion ([P507]− ) played
the critical role in this extraction process.
3.2. Extraction parameters
Extraction parameters, such as the concentrations of [N1888][P507]
and TBP in the organic phase ([N1888][P507]org and [TBP]org), Mg/Li
molar ratio, and organic phase to aqueous phase ratio (O/A) have sig­
nificant effects on Li+ extraction from the salt lake brine.
3.2.1. Effect of [N1888][P507] concentration
In the present study, the concentration of [N1888][P507]org varied
over a range from 0 to 0.20 mol⋅L–1 for the [N1888][P507] + TBP + FeCl3
system. The results were revealed by the plot of the [N1888][P507]org
versus the extraction ratio of different metal ions at the equilibrium
state. As shown in Fig. 3, lithium extraction efficiency continuously
decreased with the increase of [N1888][P507]org concentration. It also
could be seen from this figure that the Fe3+ extraction efficiency reached
nearly 100%, and the Mg2+ extraction efficiency was kept at ~1%.
There were two possible reasons for this phenomenon. First, [P507]− in
the organic phase had a strong tendency to interact with [FeCl4]– (as
shown in Fig. 12) which was the crucial counter ions for lithium
extraction complex ([Li⋅nTBP]+) that transferred from the aqueous
phase to the organic phase. Thus, a high concentration of [P507]− led to
the decrease of [FeCl4]− concentration in the aqueous phase. Second,
[P507]− in the organic phase might destroy the coordination structure of
100
Li+
80 Mg2+
Fe3+
60
E /%
40
20
0.00 0.05 0.10 0.15 0.20
-1
Concentration of [N1888][P507]org /mol·L
Fig. 3. Effect of [N1888][P507] concentration in the organic phase on the metal
ions extractions.
4
Separation and Purification Technology 274 (2021) 119051
[Li⋅nTBP]+ formed in the extraction duration, because its coordination
ability with Li+ was stronger than that of TBP. Subsequently,
[Li⋅nP507](1–n) was easily formed in water, leading to a decrease in
lithium extraction efficiency. Therefore, 0.025 mol⋅L–1 IL was chosen for
further research.
3.2.2. Effect of [TBP]org concentration
To investigate the effect of [TBP]org on the lithium extraction effi­
ciency and selectivity of this extraction system, initial [TBP]org (volume
concentration) varied over a range from 10% to 90% and [N1888]
[P507]org kept at 0.025 mol⋅L–1 was employed for the [N1888][P507] +
TBP + FeCl3 system. It was shown in Fig. 4 that a third phase formed
between the aqueous phase and the organic phase when the [TBP]org <
50%. The third phase containing a high concentration of Fe(III), resulted
from the formation of Fe(III)-complexes with a low solubility in the
organic phase. Thus, the following tests were just carried out for 50% <
[TBP]org < 90%.
The extraction ratio of metal ions and separation factor of lithium to
magnesium (βLi/Mg) was plotted against the concentration of TBP in the
organic phase ([TBP]org, 50–90% or 1.82 mol⋅L–1–3.28 mol⋅L–1). As
shown in Fig. 5, lithium extraction efficiency decreased from 63.84% to
51.23%. Whereas, magnesium extraction efficiency increased slightly
with the increase of TBP concentration in the organic phase. So βLi/Mg
continuously decreased from 190 to 33. With the combined consider­
ation of both eliminating emulsification and achieving the highest
possible extraction efficiency, the concentration of TBP of 60% (2.18
mol⋅L–1) was selected.
3.2.3. Effect of O/A phase ratio
The O/A ratio in range of 4:1–1:10 were investigated. The experi­
mental results were presented in Fig. 6. The results indicated that the Li+
extraction efficiency increased first, then decreased, and reached the
highest level (63.4%) finally when the O/A phase ratio was 1:2. When
the volume of organic phase was larger than the aqueous phase (O/A
phase ratio was 4:1), the content of [N1888][P507] in the organic phase
was more than that of Fe3+ in the aqueous phase. Then most of Fe3+
would be extracted firstly with the formation of [FeCl4⋅TBP⋅P507]2–/
[FeCl4⋅P507]2–, leading to the amount of [FeCl4]– in the aqueous phase
not enough for the formation of [Li⋅2TBP][FeCl4] complexes and
consequently the very low extraction efficiency (~20%) of Li+. Mean­
while, the content of Fe3+ increased with the decrease of O/A phase
ratio from 4:1 to 1:2. Then enough [FeCl4]– formed to coordinate and
extract Li+ from the aqueous phase, which contributed to the increase of
Li+ extraction efficiency as the O/A varying from 4:1 to 1:2. When the
O/A phase ratio decreased continuously from 1:2 to 1:10, the emulsion
phase appeared. The emulsion was harmful for the lithium extraction,
leading to the decrease of the lithium extraction efficiency. Additionally,
when the O/A phase ratio was lower than 1:1, the consumption of
simulated brine became quite considerable and the extraction efficiency
of Li+ ions was just slightly increased about 3.4%. Therefore, 1:1 was
chosen and suitable for the further study.
Fig. 4. The third phase formed when [TBP]org < 50 v%. (1) [TBP]org = 10%
(0.36 mol⋅L–1), (2) [TBP]org = 20% (0.73 mol⋅L–1), (3) [TBP]org = 30% (1.09
mol⋅L–1), (4) [TBP]org = 50% (1.82 mol⋅L–1), (5) [TBP]org = 60% (2.18 mol⋅L–1).
R. Bai et al.
Fig. 5. Effect of TBP concentration in the organic phase on the metal
ions extraction.
120
Li+
Mg2+
100
Fe3+
80
60
E /%
40
20
0 [P507] + TBP + FeCl3 system, about 54% Li+ in the loaded organic phase
4:1 2:1 1:1 1:2 1:3 1:5 1:7 1:10
O/A Phase ratio
Fig. 6. Metals extraction by 60 v% TBP + 0.025 mol⋅L–1 [N1888][P507] at
different O/A phase ratio.
3.3. Effect of stripping conditions
For the lithium extraction system of TBP + FeCl3, the high concen­
tration of HCl (6–9 mol⋅L–1) were used for lithium stripping. With the
high stripping acidity, Fe3+would be kept in the oil phase to avoid any
loss accompanied with a back extraction efficiency of Li+ reaching for
nearly 100%. However, the HCl of so high concentration brought many
practical problems, such as the corrosion of the reactor, consuming a
large amount of soda ash as neutralizer and precipitant to obtain the
lithium carbonate and the saponification of oil phase after stripped. All
these would increase the cost of extracting lithium and hinder the large-
scale industrial application of the extraction method.
In order to obtain the lowest possible stripping acidity, following the
extraction performance of the IL + TBP + FeCl3 system, Li+ stripping of
this system was systematically studied. Stripping conditions, such as, the
O/A phase ratio and hydrophobicity of phosphonate IL, temperature,
initial IL concentration in organic phase ([IL]org), HCl concentration in
stripping aqueous phase ([HCl]aq) were discussed one by one.
3.3.1. O/A phase ratio and hydrophobicity of phosphonate IL
The organic phase was mixed with the simulated brine solution to
load metal ions under the optimal conditions: O/A = 1/1 (230 mL:230
mL), [TBP] = 60%, [N1888][P507] = 0.025 mol⋅L–1, extract time = 20
min, temperature = 25 ◦ C. The results showed that 6.90 g⋅L–1 Fe3+
5
Separation and Purification Technology 274 (2021) 119051
(100%), 0.432 g⋅L–1Li+ (62%), and 1.67 g⋅L–1 Mg2+ (1.7%) were loaded
to the organic phase. The loaded organic phase was stripped with pure
water at different O/A ratios. Results were shown in Fig. 7c. Compared
with the TBP + FeCl3 system of which the stripping efficiency curves of
Li+ and Fe3+ were almost consistent, stripping efficiency curves of Li+
and Fe3+ were separated in the new [N1888][P507] + TBP + FeCl3 system
when the O/A ratio was less than 20. It was indicated that Li+ could be
stripped into the aqueous phase while Fe3+ left in the organic phase.
However, the stripping efficiency of Li+ decreased rapidly and that of
Fe3+ dropped to near 0 with the further increase of O/A ratio. The
reduction of Li+ and Fe3+ stripping efficiency were mainly due to the
rapidly increase of Cl− concentration in the aqueous phase. The
increased Cl− concentration mainly came from MgCl2 which was back-
extracted from the organic phase. Although the extraction ratio of
Mg2+ was just about 1.7%, the total amount of Mg2+ in loaded organic
phase and finally in stripped aqueous phase was large because Mg2+
concentration in the feed solution was too high.
Additionally, other three kinds of IL with different cations, [N1444]
[P507], [N4444][P507], and [N8888][P507] were used to investigate the
effect of the hydrophobicity of ILs on the stripping of the metal ions. As
shown in Fig. 7 a-d, the system of [N1888][P507] + TBP + FeCl3 exhibited
the best stripping performance among above ILs. The gap between the
stripping efficiency curves of Li+ and Fe3+ by different ILs increased
firstly and then decreased as the increasing of alkyl length of their cat­
ions. The possible reason was that the more hydrophobic the IL cation
was, the weaker interaction between the cation and the anion was, due
to the effect of the steric hindrance. Therefore, more free [P507]− existed
in the system consisting of more hydrophobic IL, which was favorable
for stripping of Li+ and holding of Fe3+ in the organic phase. However,
the lower Li+ stripping efficiency for [N8888][P507]-based system than
that of [N1888][P507]-based system was noted in Fig. 7c and 7d. The
main reason was that a new complex [Li⋅nTBP][P507] was formed in the
[N8888][P507]-based system to hinder the stripping of Li+. In the [N1888]
could be stripped by water at the O/A ratio of 10, and slight Fe3+ was
lost. The stripped organic phase could be used directly in the next
extraction cycle without saponification.
3.3.2. Temperature
Effect of temperature on the stripping efficiencies of Li+, Mg2+, and
Fe3+ was investigated. In the experiment, the pure water was used as the
stripping phase with the O/A phase ratio fixed at 10:1. The results were
shown in Table 1.
It was obvious that the stripping efficiency of Mg2+ was maintained
at a very high level (around 98%) with an increase in temperature. And
there was no varying process during the temperature increasing from 15
to 65 ◦ C. However, the stripping efficiency of Li+ and Fe3+ reduced very
slightly as the temperature increased. It was indicated that temperature
showed a little effect on Li+ stripping and high temperature was favor­
able for Mg2+ stripping. When the temperature was 25 ◦ C, the stripping
efficiency of Fe3+ was about 12% and that of Li+ was about 54%. So
25 ◦ C was used as the actual stripping temperature.
3.3.3. IL concentration in the organic phase
The effect of [N1888][P507] concentration on stripping efficiency of
Li+, Mg2+, Fe3+ was studied. It was obvious from Fig. 8 that the stripping
efficiency of Li+ increased with [N1888][P507] concentration. When
[N1888][P507]org concentration was higher than 0.10 mol⋅L–1, the strip­
ping efficiency of Li+ reached up to about 88%. However, the extraction
efficiency of Li+ was just 51% when the [N1888][P507]org was 0.1
mol⋅L–1. Therefore, 0.025 mol⋅L–1 was selected as suitable [N1888][P507]
concentration in organic phase.
3.3.4. HCL concentration in the aqueous phase ([HCl]aq)
A series of stripping experiments with different HCl concentration in
aqueous phase were carried out for obtain the lowest [HCl]aq as the
R. Bai et al. Separation and Purification Technology 274 (2021) 119051

100 (a) 100 (b)

80 80

S /% 60 60

S /%
40 40
Li+ Li+
20 Mg2+ 20 Mg2+
Fe3+ Fe3+
0 0

0 20 40 60 80 0 20 40 60 80
O/A ratio O/A ratio

100 (c)
100 (d)
80
80

60
60
S /%

S /%
40 40
Li+ Li+
Mg2+
20 20 Mg2+
Fe3+
Fe3+
0 0

0 20 40 60 80 0 20 40 60 80
O/A ratio O/A ratio

Fig. 7. Effect of IL cations on stripping efficiency, (a) [N1444][P507]; (b) [N4444][P507]; (c) [N1888][P507]; (d) [N8888][P507].

Table 1 120
Effect of temperature on the stripping ratios of metal ions. The organic phase: Li+ Mg2+ Fe3+
[N1888][P507]org = 0.025 mol⋅L–1, [TBP]org = 2.185 mol⋅L–1 (60%); The unloa­
ded organic phase obtained by extraction with the simulated brine: 0.10 mol⋅L–1 100
LiCl, 4.0 mol⋅L–1 MgCl2, 0.13 mol⋅L–1 FeCl3, and 0.03 mol⋅L–1 HCl; The aqueous
phase: pure water; O/A = 10:1, and T = 15–65 ◦ C.
80
Temperature/oC Metal stripping/%

Mg2+ Fe3+
S /%

Li+
60
15 54.8 ± 0.2 98.2 ± 2.0 10.4 ± 2.0
25 53.8 ± 0.2 94.4 ± 3.0 11.8 ± 0.4
35 53.5 ± 0.2 96.4 ± 0.2 11.5 ± 0.2 40
45 51.5 ± 0.2 99.9 ± 0.2 10.9 ± 0.2
55 51.5 ± 0.2 97.8 ± 0.2 10.6 ± 0.1
65 52.6 ± 0.4 97.6 ± 2.0 8.90 ± 0.1 20

stripping solution. The effect of HCl concentration on stripping effi­ 0

0 0.025 0.05 0.10 0.15
ciency of Li+, Mg2+, Fe3+ was studied. It could be seen from Fig. 9 that
[IL]org /mol⋅L-1
the stripping efficiency of Li+ increased with H+ concentration. Though
the stripping efficiency of Li+ reached up to about 99% by the single-
Fig. 8. Effect of IL concentration on stripping ratios of metal ions. [TBP]org =
stage of stripping process when [HCl]aq in the stripping solution was 60%, O/A = 10, T = 25 ◦ C, and stripping time = 20 min.
higher than 2.50 mol⋅L–1, a much lower acidity of 1.5 mol⋅L–1 for [HCl]aq
was selected as suitable H+ concentration in the stripping solution.
3.4.1. Extraction of metal ions by the [N1888][P507]
Under the selected acidity, the stripping efficiency of Li+ and Mg2+ were
The extraction of metal ions by the [N1888][P507] alone system was
about 90%, and the loss of Fe3+ was lower than 1.1%.
firstly studied to investigate the mechanism of this extraction system.
The aqueous solutions containing 0.10 mol⋅L–1 LiCl, 0.13 mol⋅L–1 FeCl3,
3.4. Mechanisms of lithium extraction and stripping
and 4.0 mol⋅L–1 MgCl2 were mixed with the organic phase ([N1888]
[P507] + kerosene). The effect of [N1888][P507] concentration on metal
Fe3+ played a critical role in the lithium extraction and stripping
ions extraction has been investigated. As shown in Fig. 10, Fe3+ was
process with the [N1888][P507] + TBP + FeCl3 system. So Fe3+ extraction
efficiently extracted with the increase of [N1888][P507] concentration
was discussed in detail and the possible mechanisms of lithium extrac­
and tended to equilibrium finally. However, both of Li+ and Mg2+ in the
tion and stripping were proposed in this part.
aqueous phase were slightly extracted. The results indicated that the

6
R. Bai et al. Separation and Purification Technology 274 (2021) 119051

120 The distribution ratio of Fe3+ (DFe) was calculated as

2+ 3+
Li +
Mg Fe 3)−
[FeClm ⋅n(P507 )⋅kTBP](m+n−
(9)
org
100 DFe =
[Fe3+ ]aq

Combined Eqs. (8) and (9), and taken the logarithm on both sides of
80
the resulting equation:
[ ]
(10)
S /%

60 lgDFe = lgKe + mlg[Cl− ]aq + nlg P−507 org + klg[TBP]org

The concentration of Cl− in the solution was adjusted by adding

40 MgCl2, at the same time maintaining the concentration of [TBP] and
[N1888][P507] constant respectively. After extraction, plotting the lgDFe
against the lg[Cl− ], a fitting line was obtained, as shown in Fig. 11a. The
20
slope was close to 4. It meant that 4 mol Cl− were also extracted into the
oil phase simultaneously when 1 mol of Fe3+ was extracted. The similar
0 procedures were used to plot the line of lgDFe against the lg[P−507]org and
0 0.25 0.5 1.0 1.5 2.0 2.5 3.0
-1
lg[TBP]org. The results were shown in Fig. 11b, c, indicating the slope
[HCl]aq /mol⋅L were 1.1 and 0.25, respectively. The slopes meant that n = 1, k = 0.25. In
other words, 1 mol [P507]− and 1 or 0 mol TBP were involved in the
Fig. 9. Effect of [HCl]aq on the stripping ratios of metal ions. [TBP]org = 60%,
extraction of 1 mol Fe3+. Therefore, Eq. (7) could be expressed as Eq.
[IL]org = 0.025 mol⋅L–1, O/A = 10, T = 25 ◦ C, and stripping time = 20 min.
(11) or Eq. (12). And the possible extraction adducts of Fe were shown in
Fig. 12. The results were different from P507 as co-extractant in TBP +
FeCl3 system, of which the possible Fe3+ complexes was FeCl2L⋅(HL)⋅
100 2TBP (HL denotes P507) [48].

Fe3+ − 2− +
aq + 4Claq + [N1888 ][P507 ]org ⇄[FeCl4 ⋅P507 ]org + [N1888 ]org (11)
80 Li+
Mg2+ Fe3+ − 2− +
aq + 4Claq + [N1888 ][P507 ]org + TBPorg ⇄[FeCl4 ⋅TBP⋅P507 ]org + [N1888 ]org

60 Fe3+ (12)
E /%

3.4.3. Effect of [Cl− ]aq on Fe3+ extraction by [N1888][P507] + TBP +

40
20
0 same time with the continuously adding MgCl2. Fe3+ transferred into oil
0.00 0.04 0.08 0.12
[N1888][P507]org /mol⋅L-1
Fig. 10. Effect of [N1888][P507] concentration on metal extraction in [N1888]
[P507] alone system. O/A = 1:1, T = 25 ◦ C, and extraction time = 20 min.
Fe3+ could be extracted by [N1888][P507] preferentially.
3.4.2. Extraction of Fe3+ by [N1888][P507] + TBP + FeCl3 system
In order to determine the quantity relationships of Fe3+ versus Cl–,
Fe3+ versus [N1888][P507], and Fe3+ versus TBP in the extraction, ex­
periments were carried out. The extraction reaction of Fe3+ from the
brine with [N1888][P507] + TBP + kerosene could be described as Eq. (7).
FeCl3 system
The metal ions extraction ratio at different concentrations of MgCl2
was shown in Fig. 13. Without MgCl2, the extraction of Li+ was negli­
gible (0.5%). Meanwhile, about 12% of Fe3+ was extracted, assumed to
be extracted by TBP and IL anion ([P507]− ) with the formation of
[FeCl3⋅P507⋅nTBP]− . The extraction ratio of Li+ and Fe3+ increased at the
phase mainly in two formations, one was [FeCl3⋅P507⋅nTBP]− , the other
0.16
was [Li⋅2TBP⋅FeCl4] that could be supported by the increasing extrac­
tion of Li+. When the initial [MgCl2]aq was higher than 3.0 mol⋅L–1, Li+
and Fe3+ would be extracted dominantly in the [Li⋅2TBP⋅FeCl4] species
because of the ultrahigh concentration of Cl− in aqueous phase. TBP and
[P507]− were hardly involved in Fe3+ extraction.
The effect of [Cl–]aq on the stoichiometry of TBP in the Fe3+
extraction was also investigated by varying TBP concentration in the
organic phase, and maintaining the [N1888][P507]org constant at three
different levels of [Cl–]aq (0.4, 4.0, 7.9 mol⋅L–1). The results were plot­
ting in Fig. 14, which showed that the slopes were 0.251, 0.899, 3.46,
respectively. It was indicated that slopes increased gradually with the
increasing concentration of Cl–. It meant that when [Cl− ]aq = 7.9
mol⋅L–1, one mole of Fe3+ was extracted, involving 3–4 mol of TBP to be

Fe3+ − (m+n−
aq + mClaq + n[N1888 ][P507 ]org + kTBPorg ⇄[FeClm ⋅kTBP⋅nP507 ]org
3)−
+ n[N1888 ]+
org (7)

coordinated in the organic phase.

where m, n, k were the stoichiometry of Cl–, [N1888][P507], and TBP in The FT-IR spectra of the oil phase after the extraction were examined
the reaction, respectively. The equilibrium constant (Ke) of the extrac­ to observe the change of the functional groups in molecules. The results
tion was shown in Eq. (8): were shown in Fig. 15a. According to the literature and our former study
3)− n
[25,32], the characteristic peak of TBP was the stretching vibration
[FeClm ⋅n(P507 )⋅kTBP](m+n− [N1888 + ]org absorption of P– – O, which was a medium-intensity absorption peak at
(8)
org
Ke = − 1
n
[Fe3+ ]aq [Cl− ]morg [N1888 + ]org [P507 − ]norg [TBP]korg near 1280 cm . It was reported that the peak of P– – O at 1280 cm− 1

7
R. Bai et al. Separation and Purification Technology 274 (2021) 119051

3.0
(a)
2.5

2.0

1.5

lgDFe
1.0 Slope = 3.95
R2 = 0.986
0.5

0.0

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

lg[Cl-]aq

0.2
(b) 0.00 (c)

0.0
-0.05 Slope = 0.251
R2 = 0.974
lgDFe

lgDFe
-0.2
-0.10
Slope = 1.10
R2 = 0.999
-0.4 -0.15

-0.6 -0.20

-1.76 -1.60 -1.44 -1.28 -1.12 -0.96 -0.6 -0.4 -0.2 0.0 0.2
lg[N1888][P507]org lg[TBP]org

Fig. 11. Effect of Cl–, [N1888][P507], TBP concentration on the extraction of Fe3+. (a) lgDFe vs lg[Cl–]aq; (a) lgDFe vs lg[N1888][P507]org; (a) lgDFe vs lg[TBP]org.

100
Li+
80 Mg2+
Fe3+

Fig. 12. Possible structures of adducts containing Fe3+. (a) Fe3+ coordination 60
in Eq. (11); (b) Fe3+ coordination in Eq. (12); R = − CH2(C2H5)CH2CH2CH2CH3;
E /%

R’ = − CH2CH2CH2CH3.
40
underwent a significant red-shift when the TBP molecule coordinated
with metal ions. As shown in Fig. 15a, the specific absorption peak 20
performed significantly red shifted from 1285 cm− 1 to 1267 cm− 1 after
the extraction for Li+ and Fe3+, indicating that the coordination effect
0
was significantly increased when the MgCl2 concentration increased
from 0 to 4.0 mol⋅L–1. The reason was that large amount of Cl− will 0 1 2 3 4
urgent the formation of [FeCl4]− , the co-extractant of TBP for lithium
[Mg2+]initial /mol⋅L-1
extraction. What’s more, the UV–vis spectrum (Fig. 15b) did not show
characteristic bands when there was no MgCl2 in the aqueous solution. Fig. 13. Metals extraction by 60 v% TBP-0.025 mol⋅L–1 IL at different con­
With the addition of MgCl2, three main absorption peaks appeared at centration of Mg2+. O/A = 1:1, T = 25 ◦ C, and extraction time = 20 min.
530, 618, and 685.5 nm, which well matched the characteristic ab­
sorption peaks of [FeCl4]− [49,50]. The results suggested that Fe3+
process of Li+ and Fe3+ by the [N1888][P507] + TBP + Kerosene system
presented in the organic was in the form of [FeCl4]− . Additionally, ac­
might be comply with the following scheme: (1) In the brine with high
cording to Fig. 15b, the strength of the specific peak will don’t change
concentration of MgCl2, [FeCl4]− was formed and extracted together
when the [Mg2+]aq exceed 2.0 mol⋅L–1 ([Cl− ]aq = 4.4 mol⋅L–1), indi­
with [Li⋅2TBP]+, meanwhile IL, [N1888][P507], was just slightly involved
cating the [FeCl4]− concentration will reached the maximum value.
in this step. The reaction was shown in Eq. (1); (2) In the scrubbing step,
most of extracted Mg2+ was eluted into the aqueous phase together with
3.4.4. Possible mechanisms of lithium extraction and stripping
Cl− ; (3) In the stripping procedure, the structure of [FeCl4]− might be
According to the above discussion, the extraction and stripping
damaged and new complexes were formed. The new possible

8
R. Bai et al. Separation and Purification Technology 274 (2021) 119051

2.5
[Cl-] = 0.4 mol⋅L-1
[Cl-] = 4.0 mol⋅L-1
2.0
[Cl-] = 7.9 mol⋅L-1
Slope = 3.46
1.5 R2 = 0.998

lgDFe
1.0

0.5
Slope = 0.899 Slope = 0.251
R2 = 0.957 R2 = 0.974
0.0

-0.5
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
lg[TBP]org

Fig. 14. Effect of Cl– concentration on the extraction of Fe3+.

coordination of Fe3+ was [FeCl4⋅TBP⋅(P507)]2− , as shown in Eq. (13). stripping don’t need to be regenerated by the NaOH solution and could

[Li⋅2TBP][FeCl4 ]org + [P507 ]−org + Haq

+
⇄Li+ 2−
aq + [FeCl4 ⋅TBP⋅(P507 )]org + [TBP]org + [H⋅P507 ]org (13)

0.3
(a) (b) [Mg2+] = 4.0 mol⋅L-1
[Mg2+] = 3.0 mol⋅L-1
[Mg2+] = 2.0 mol⋅L-1
530 nm
[Mg2+] = 1.0 mol⋅L-1
0.2 [Mg2+] = 0.5 mol⋅L-1
[Mg2+] = 0 mol⋅L-1
Transmittance

Abs.

618 nm 685.5 nm
0.1
[Mg2+] = 4.0 mol⋅L-1
[Mg2+] = 3.0 mol⋅L-1
[Mg2+] = 2.0 mol⋅L-1
1285 cm-1 [Mg2+] = 1.0 mol⋅L-1
1267 cm-1 [Mg2+] = 0.5 mol⋅L-1 0.0
[Mg2+] = 0 mol⋅L-1

1500 1250 1000 750 500 500 600 700 800

-1
Wavenumber /cm Wavelength /nm
Fig. 15. Spectral analysis of the equilibrium organic phase. (a) FTIR spectra; (b) UV–visible spectra.

3.5. Reusability of the [N1888][P507] + TBP + FeCl3 system
In the extraction system (TBP + FeCl3), the organic phase after
stripping with high concentration HCl solution couldn’t be directly used
for the next extraction cycle. And the saponification process was
essential step which consumed large amount of NaOH and increased the
whole cost. However, in the new extraction system ([N1888][P507] +
TBP + FeCl3), the concentration of HCl solution used in the lithium
striping was much lower than that of the traditional system. The H+
produced by ionization could be used to prevent the hydrolysis of Fe3+
since Fe3+ would be hydrolyzed and formed Fe(OH)3 in the TBP + FeCl3
system without enough acidity. Therefore, the organic phase after
be used directly in the next extraction cycle. To test the former view and
to study the stability and reusability of the extraction system, six com­
plete extraction cycles (extraction1-washing-stripping-extraction2-
⋅⋅⋅-extraction6) were carried out. The extraction efficiency of Li+ just
fluctuated around 70% with the increase of regeneration times (Fig. 16),
the stripping efficiency of metal ions along extraction/stripping cycles
was presented in Table 2. It was indicated that the extraction system
showed relatively high stability and reusability.
If FeCl3 wasn’t added to the feed brine from the second cycle, the
extraction efficiency of Li+ significantly decreased from 62.07% to 31%.
And in the following 4 cycles the extraction efficiency of Li+ slowly
decreased to 26.7% (Table S1). This result indicated that [FeCl4]−
formed in the high Mg brine performed the essential role on the selective
extraction of Li+, according to the ion-pair induced solvent extraction
principle [24].

9
R. Bai et al. Separation and Purification Technology 274 (2021) 119051

Acknowledgements
Fe3+ Li+ Mg2+
100 Sincerely appreciate Prof. Suojiang Zhang (IPE, CAS) for his careful
academic guidance and great support. This work was supported by the
National Natural Science Foundation of China (21978302, U20A20149),
80 Innovation Academy for Green Manufacture, Chinese Academy of Sci­
ences (IAGM2020DB05), and State Key Laboratory of Multiphase
60 Complex Systems (MPCS-2019-D-15).
E /%

Author Contributions
40
The paper was written through the contributions of all authors. All
20 authors have approved the final version of the paper. Ruibing Bai
executed experiments and wrote the paper. Daoguang Wang, Yanqiang
Zhang and Junjie Cui gave some good suggestions about the experi­
0 ments. Junfeng Wang guided the project and revised the paper.
1 2 3 4 5 6
Reganeration times of organic phase Appendix A. Supplementary material
Fig. 16. Cycle times of organic phase with FeCl3 contained in the
aqueous phase. Supplementary data to this article can be found online at https://doi.
org/10.1016/j.seppur.2021.119051.

References
Table 2
Stripping efficiency along extraction/stripping cycles.
[1] S.E. Kesler, P.W. Gruber, P.A. Medina, G.A. Keoleian, M.P. Everson, T.
Cycle time SLi/% SMg/% SFe/% J. Wallington, Global lithium resources: Relative importance of pegmatite, brine
and other deposits, Ore Geol. Rev. 48 (2012) 55–69.
1 97.6 ± 0.06 100 ±1 6.80 ± 0.04 [2] L.T. Peiró, G.V. Méndez, R.U. Ayres, Lithium: Sources, Production, Uses, and
2 97.0 ± 0.04 100 ±2 12.1 ± 0.07 Recovery Outlook, JOM 65 (2013) 986–996.
3 95.0 ± 0.07 100 ±2 12.6 ± 0.07 [3] P.K. Choubey, K.S. Chung, M.S. Kim, J.C. Lee, R.R. Srivastava, Advance review on
4 97.6 ± 0.03 100 ±2 9.40 ± 0.05 the exploitation of the prominent energy-storage element Lithium. Part II: From sea
5 96.4 ± 0.05 100 ±2 9.20 ± 0.05 water and spent lithium ion batteries (LIBs), Miner. Eng. 110 (2017) 104–121.
[4] S. Gupta, V. Taupin, C. Fressengeas, J. Chevy, Crystal plasticity modeling of the
effects of crystal orientation and grain-to-grain interactions on DSA-induced strain
localization in Al–Li alloys, Materialia 8 (2019), 100467.
4. Conclusion
[5] A. Lu, W. Niu, Y. Dai, H. Xu, J. Dong, Tribological Properties of ZnS
(NH2CH2CH2NH2)0.5 and ZnS as Additives in Lithium Grease, Lubricants 7 (2019)
A novel solvent extraction system of [N1888][P507] + TBP + FeCl3 26.
was developed to extract lithium from the brine with high Mg/Li ratio. [6] G. Zubi, R. Dufo-López, M. Carvalho, G. Pasaoglu, The lithium-ion battery: State of
the art and future perspectives, Renew. Sustain. Energy Rev. 89 (2018) 292–308.
Lithium in the loaded organic phase was easily stripped by the diluted [7] Mineral commodity summaries 2020, Mineral Commodity Summaries Reston, VA,
HCl solution (1–1.5 mol⋅L–1). Stripping efficiency of Li+ was about 90% 2020, pp. 204.
and the loss of Fe3+ was lower than 1.1%, under the condition of O/A = [8] S. Jakhar, M. Abhangi, S. Tiwari, R. Makwana, V. Chaudhari, H.L. Swami,
C. Danani, C.V.S. Rao, T.K. Basu, D. Mandal, S. Bhade, R.V. Kolekar, P.J. Reddy,
10:1 and [HCl] = 1.5 mol⋅L–1. The stripped organic phase could be used R. Bhattacharyay, P. Chaudhuri, Tritium breeding mock-up experiments containing
directly for the next extraction cycle. After six extraction-stripping- lithium titanate ceramic pebbles and lead irradiated with DT neutrons, Fusion Eng.
extraction cycles, it was found that the Li extraction efficiency by the Des. 95 (2015) 50–58.
[9] B. Swain, Recovery and Recycling of Lithium: A Review, Sep. Purif. Technol. 172
proposed system remained above 70%. Li+ and Fe3+ extraction and (2017) 388–403.
stripping mechanism were studied in detail. The results showed that [10] P. Meshram, B.D. Pandey, T.R. Mankhand, Extraction of lithium from primary and
[FeCl4]– was formed and complexed together with [Li⋅2TBP]+ to form secondary sources by pre-treatment, leaching and separation: A comprehensive
review, Hydrometallurgy 150 (2014) 192–208.
[Li⋅2TBP][FeCl4]. In the stripping procedure, the structure of [Li⋅2TBP] [11] Mineral Commodity Summaries 2019, Mineral Commodity Summaries Reston, VA,
[FeCl4] might be destroyed, and new possible coordination, [Li⋅2TBP]+, 2019.
and [FeCl4⋅TBP⋅P507]2–/[FeCl4⋅P507]2– to be formed simultaneously. [12] X. Liu, H. Zhong, Z. Tang, Current status and existing problems of lithium
extraction technology from salt lake, Inorg. Chem. Ind. 41 (2009) 4–6.
Although the addition of [N1888][P507] to the TBP + FeCl3 system made
[13] X. Li, Y. Chao, L. Chen, W. Chen, J. Luo, C. Wang, P. Wu, H. Li, W. Zhu, Taming
the Li+ extraction efficiency decrease slightly (about 5%), this novel wettability of lithium ion sieve via different TiO2 precursors for effective Li
solvent extraction system significantly reduced the stripping acidity recovery from aqueous lithium resources, Chem. Eng. J. 392 (2020), 123731.
comparing to the traditional system of TBP + FeCl3 from HCl 6 mol⋅L–1 [14] G. Yang, F. Wu, Z. Wei, K. Yang, C. Zhang, Y. Ma, J. Pan, Capturing lithium using
functional macroporous microspheres with multiple chambers from one-step
to 1.5 mol⋅L–1. The new system was promising for lithium recovery from double emulsion via a tailoring supramolecular route and postsynthetic interface
the high-magnesium brine. The method proposed in this work could modification, Chem. Eng. J. 389 (2020), 124372.
provide a reference for reducing the stripping acidity in the process of [15] K. Ooi, A. Sonoda, Y. Makita, R. Chitrakar, Y. Tasaki-Handa, T. Nakazato, Recovery
of lithium from salt-brine eluates by direct crystallization as lithium sulfate,
lithium extraction with the TBP + FeCl3 system. Hydrometallurgy 174 (2017) 123–130.
[16] H. Wang, Y. Zhong, B. Du, Y. Zhao, M. Wang, Recovery of both magnesium and
Declaration of Competing Interest lithium from high Mg/Li ratio brines using a novel process, Hydrometallurgy 175
(2018) 102–108.
[17] A. Razmjou, M. Asadnia, E. Hosseini, A. Habibnejad Korayem, V. Chen, Design
The authors declare that they have no known competing financial principles of ion selective nanostructured membranes for the extraction of lithium
interests or personal relationships that could have appeared to influence ions, Nat. Commun. 10 (2019) 5793.
[18] H. Wu, Y. Lin, W. Feng, T. Liu, L. Wang, H. Yao, X. Wang, A novel nanofiltration
the work reported in this paper. membrane with [MimAP][Tf2N] ionic liquid for utilization of lithium from brines
with high Mg2+/Li+ ratio, J. Membr. Sci. 603 (2020), 117997.
[19] Y. Zhang, L. Wang, W. Sun, Y. Hu, H. Tang, Membrane technologies for Li+/Mg2+
separation from salt-lake brines and seawater: A comprehensive review, J. Ind.
Eng. Chem. 81 (2020) 7–23.

10
R. Bai et al.
[20] Z. Ji, Q. Chen, J. Yuan, J. Liu, Y. Zhao, W. Feng, Preliminary study on recovering
lithium from high Mg2+/Li+ ratio brines by electrodialysis, Sep. Purif. Technol.
172 (2017) 168–177.
[21] X. Zhao, H. Yang, Y. Wang, Z. Sha, Review on the electrochemical extraction of
lithium from seawater/brine, J. Electroanal. Chem. 850 (2019), 113389.
[22] D. Shi, B. Cui, L. Li, X. Peng, L. Zhang, Y. Zhang, Lithium extraction from low-grade
salt lake brine with ultrahigh Mg/Li ratio using TBP – kerosene – FeCl3 system, Sep.
Purif. Technol. 211 (2019) 303–309.
[23] X. Yu, X. Fan, Y. Guo, T. Deng, Recovery of lithium from underground brine by
multistage centrifugal extraction using tri-isobutyl phosphate, Sep. Purif. Technol.
211 (2019) 790–798.
[24] L. Cui, L. Wang, M. Feng, L. Fang, Y. Guo, F. Cheng, Ion-pair induced solvent
extraction of lithium(I) from acidic chloride solutions with tributyl phosphate,
Green Energy Environ. (2020), https://doi.org/10.1016/j.gee.2020.05.002.
[25] R. Bai, J. Wang, L. Cui, S. Yang, W. Qian, P. Cui, Y. Zhang, Efficient extraction of
lithium ions from high mg/li ratio brine through the synergy of tbp and hydroxyl
functional ionic liquids, Chin. J. Chem . 38 (2020) 1743–1751.
[26] J.R. Nelli, T.E. Arthur, J. Gastonia, Recovery of Lithium from Bitterns. US patent
US3537813DA, 1970.
[27] L. Ji, Y. Hu, L. Li, D. Shi, J. Li, F. Nie, F. Song, Z. Zeng, W. Sun, Z. Liu, Lithium
extraction with a synergistic system of dioctyl phthalate and tributyl phosphate in
kerosene and FeCl3, Hydrometallurgy 162 (2016) 71–78.
[28] J. Song, T. Huang, H. Qiu, X. Li, T. He, Recovery of lithium from salt lake brine of
high Mg/Li ratio using Na[FeCl4⋅2TBP] as extractant: Thermodynamics, kinetics
and processes, Hydrometallurgy 173 (2017) 63–70.
[29] W. Xiang, S. Liang, Z. Zhou, W. Qin, W. Fei, Lithium recovery from salt lake brine
by counter-current extraction using tributyl phosphate/FeCl3 in methyl isobutyl
ketone, Hydrometallurgy 171 (2017) 27–32.
[30] D. Shi, L. Zhang, X. Peng, L. Li, F. Song, F. Nie, L. Ji, Y. Zhang, Extraction of lithium
from salt lake brine containing boron using multistage centrifuge extractors,
Desalination 441 (2018) 44–51.
[31] H. Li, L. Li, X. Peng, L. Ji, W. Li, Selective recovery of lithium from simulated brine
using different organic synergist, Chin. J. Chem. Eng. 27 (2) (2019) 335–340.
[32] H. Su, Z. Li, Z. Zhu, L. Wang, T. Qi, Extraction relationship of Li+ and H+ using
tributyl phosphate in the presence of Fe(III), Sep. Sci. Technol. 55 (9) (2020)
1677–1685.
[33] M.L. Dietz, D.C. Stepinski, A ternary mechanism for the facilitated transfer of metal
ions into room-temperature ionic liquids (RTILs): implications for the “greenness”
of RTILs as extraction solvents, Green Chem. 7 (2005) 747–750.
[34] H. Luo, S. Dai, P.V. Bonnesen, A.C.B. Iii, Separation of fission products based on
ionic liquids: task-specific ionic liquids containing an aza-crown ether fragment,
J. Alloy. Compd. 418 (2006) 195–199.
[35] X. Song, S. Wang, Z. Li, Application of ionic liquid as green medium for extraction
of lithium from salt lake brine, Chin. J. Rare Met. 35 (2011) 434–439.
11
Separation and Purification Technology 274 (2021) 119051
[36] S. Yang, G. Liu, J. Wang, L. Cui, Y. Chen, Recovery of lithium from alkaline brine
by solvent extraction with functionalized ionic liquid, Fluid Phase Equilib. 493
(2019) 129–136.
[37] D. Gao, Y. Guo, X. Yu, S. Wang, T. Deng, Extracting lithium from the high
concentration ratio of magnesium and lithium brine using imidazolium-based ionic
liquids with varying alkyl chain lengths, J. Chem. Eng. Jpn. 49 (2) (2016) 104–110.
[38] C. Shi, Y. Jing, Y. Jia, Solvent extraction of lithium ions by tri-n-butyl phosphate
using a room temperature ionic liquid, J. Mol. Liq. 215 (2016) 640–646.
[39] S. Chen, D. Gao, X. Yu, Y. Guo, T. Deng, Thermokinetics of lithium extraction with
the novel extraction systems (tri-isobutyl phosphate + ionic liquid + kerosene),
J. Chem. Thermodyn. 123 (2018) 79–85.
[40] Y. Wang, H. Liu, J. Fan, X. Liu, Y. Hu, Y. Hu, Z. Zhou, Z. Ren, Recovery of lithium
ions from salt lake brine with a high magnesium/lithium ratio using
heteropolyacid ionic liquid, ACS Sustain. Chem. Eng. 7 (3) (2019) 3062–3072.
[41] A. Masmoudi, G. Zante, D. Trébouet, R. Barillon, M. Boltoeva, Understanding the
mechanism of lithium ion extraction using tributyl phosphate in room temperature
ionic liquid, Solvent Extr. Ion Exch. 38 (7) (2020) 777–799.
[42] C. Shi, Y. Jing, J. Xiao, X. Wang, Y. Jia, Liquid-liquid extraction of lithium using
novel phosphonium ionic liquid as an extractant, Hydrometallurgy 169 (2017)
314–320.
[43] C. Shi, Y. Jing, J. Xiao, X. Wang, Y. Yao, Y. Jia, Solvent extraction of lithium from
aqueous solution using non-fluorinated functionalized ionic liquids as extraction
agents, Sep. Purif. Technol. 172 (2017) 473–479.
[44] C. Shi, H. Li, B. Liu, Y. Qin, G. Song, Solvent extraction of lithium from aqueous
solution using an ammonium ionic liquid, J. Mol. Liq. 304 (2020), 112756.
[45] X. Zhao, H. Wu, M. Duan, X. Hao, Q. Yang, Q. Zhang, X. Huang, Liquid-liquid
extraction of lithium from aqueous solution using novel ionic liquid extractants via
COSMO-RS and experiments, Fluid Phase Equilib. 459 (2018) 129–137.
[46] X. Sun, C. Do-Thanh, H. Luo, S. Dai, The optimization of an ionic liquid-based
TALSPEAK-like process for rare earth ions separation, Chem. Eng. J. 239 (2014)
392–398.
[47] H. Su, Z. Li, J. Zhang, W. Liu, Z. Zhu, L. Wang, T. Qi, Combining selective
extraction and easy stripping of lithium using a ternary synergistic solvent
extraction system through regulation of Fe3+ coordination, ACS Sustain. Chem.
Eng. 8 (4) (2020) 1971–1979.
[48] C. Tan, X. Zhang, S. Cao, S. Li, H. Guo, Y. Tian, D. Chen, W. Tian, L. Wang, Z. Qin,
Solvent extraction of americium(III) and europium(III) with 2,6-bis(5,6-diethyl-
1,2,4-triazin-3-yl) pyridine in ionic liquids: experimental study and molecular
dynamics simulation, Sep. Purif. Technol. 192 (2018) 302–308.
[49] D.E. Metzler, R.J. Myers, The distribution of ferric iron between hydrochloric acid
and isopropyl ether solutions. III. Spectral, isopiestic and magnetic susceptibility
studies, J. Am. Chem. Soc. 72 (8) (1950) 3776–3780.
[50] H.L. Friedman, The visible and ultraviolet absorption spectrum of the
tetrachloroferrate (III) ion in various media 1, J. Am. Chem. Soc. 74 (1952) 5–10.