Homogeneous liquid-liquid extraction of neodymium(iii) by choline hexafluoroacetylacetonate in the ionic liquid choline bis(trifluoromethylsulfonyl)imide

Article abstract of DOI:10.1039/c4dt01340a

The ionic liquid choline bis(trifluoromethylsulfonyl)imide, [Chol][Tf ₂N], was used for the extraction of neodymium(iii), in combination with choline hexafluoroacetylacetonate, [Chol][hfac], as the extractant. The binary mixture of [Chol][Tf₂N] and water shows temperature-dependent phase behavior, with an upper critical solution temperature of 72 °C. A novel extraction technique, homogeneous liquid-liquid extraction (HLLE), was applied to this solvent system. HLLE is based on the use of thermomorphic solvent mixtures and has the advantage of forming a homogeneous phase during mixing. Extraction is not kinetically hindered by an interface and the extraction equilibrium is reached faster than in the case of heterogeneous mixing in conventional solvent extraction. Several extraction parameters were studied for the extraction of neodymium(iii) with [Chol][hfac]: temperature, pH, extractant concentration and loading of the ionic liquid phase. A speciation study was performed to determine the stoichiometry of the extracted neodymium(iii) complex and a plausible extraction mechanism is proposed. Neodymium is extracted as a tetrakis hexafluoroacetylacetonate complex with one choline cation as counter ion. The crystal structure of the extracted complex showed the presence of a coordination bond between the choline counter ion and the neodymium(iii) center, resulting in a coordination number of nine. The stripping of the loaded neodymium and the influence of acid and extractant concentrations on the phase behavior of the [Chol][Tf₂N]-H ₂O system were investigated. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4dt01340a

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Homogeneous liquid–liquid extraction
of neodymium(III) by choline
hexafluoroacetylacetonate in the ionic liquid
choline bis(trifluoromethylsulfonyl)imide†
Cite this: Dalton Trans., 2014, 43,
11566
Bieke Onghena, Jeroen Jacobs, Luc Van Meervelt and Koen Binnemans*
The ionic liquid choline bis(trifluoromethylsulfonyl)imide, [Chol][Tf2N], was used for the extraction of
neodymium(III), in combination with choline hexafluoroacetylacetonate, [Chol][hfac], as the extractant.
The binary mixture of [Chol][Tf2N] and water shows temperature-dependent phase behavior, with an
upper critical solution temperature of 72 °C. A novel extraction technique, homogeneous liquid–liquid
extraction (HLLE), was applied to this solvent system. HLLE is based on the use of thermomorphic solvent
mixtures and has the advantage of forming a homogeneous phase during mixing. Extraction is not kineti-
cally hindered by an interface and the extraction equilibrium is reached faster than in the case of hetero-
geneous mixing in conventional solvent extraction. Several extraction parameters were studied for the
extraction of neodymium(III) with [Chol][hfac]: temperature, pH, extractant concentration and loading of
the ionic liquid phase. A speciation study was performed to determine the stoichiometry of the extracted
neodymium(III) complex and a plausible extraction mechanism is proposed. Neodymium is extracted as a
tetrakis hexafluoroacetylacetonate complex with one choline cation as counter ion. The crystal structure
of the extracted complex showed the presence of a coordination bond between the choline counter ion
and the neodymium(III) center, resulting in a coordination number of nine. The stripping of the loaded
neodymium and the influence of acid and extractant concentrations on the phase behavior of the
[Chol][Tf2N]–H2O system were investigated.
Received 5th May 2014,
Accepted 7th June 2014
DOI: 10.1039/c4dt01340a
www.rsc.org/dalton
Extractants can be termed acidic, basic (anion exchange) or
solvating, depending on the type of complex that is formed.1
Introduction
Liquid–liquid extraction or solvent extraction is widely used in Solvent extraction has the advantage that it can be
hydrometallurgy and is applied to the separation and proces- implemented in a continuous mode and it is suitable for the
sing of a wide range of metals, e.g. rare earths, hafnium–zirco- processing of high metal feed concentrations.3 Ionic liquids
nium, niobium–tantalum and nickel–cobalt.1–3 Solvent are intensively studied as an alternative organic solvent for
extraction is based on the preferential distribution of a solute solvent extraction.4–6 Ionic liquids (ILs) are solvents that
between two immiscible phases, an aqueous phase and an consist entirely of ions.7–14 Properties such as low volatility,
organic phase, that are in contact with each other. The organic low flammability, wide liquidus range, good chemical stability
phase usually consists of a molecular solvent such as kerosene and adjustable miscibility and polarity, make ionic liquids
or toluene. Metal ions in the aqueous solution are generally interesting replacements for the currently used noxious volatile
hydrated and prefer to stay in the aqueous phase during organic solvents (VOCs). The specific properties of an ionic
solvent extraction because of the low solubility of the hydrated liquid depend both on the structure of the cation and the
metal ion in the organic phase. Therefore, an extractant is anion. Different cation–anion combinations result in different
added to the organic phase, which will coordinate to the metal properties. Thus, the structure can be tuned according to the
ion to form a more hydrophobic and extractable complex. application. Ionic liquids are often highly viscous and this
hampers their use in solvent extraction.15–18 In a stirred two-
phase mixture during solvent extraction, two stagnant films
KU Leuven, Department of Chemistry, Celestijnenlaan 200F, P.O. Box 2404,
are formed near the interface, one at each side, in which move-
3001 Leuven, Belgium. E-mail: Koen.Binnemans@chem.kuleuven.be;
ment of solute solely occurs via diffusion. The thickness of
Fax: +32 16 32 7446
these films is correlated to the viscosity of the liquid, i.e.
†CCDC 1000170. For crystallographic data in CIF or other electronic format see
DOI: 10.1039/c4dt01340a
a higher viscosity results in a thicker film.1 Since diffusion of
11566 | Dalton Trans., 2014, 43, 11566–11578
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
the solute is also slower when the viscosity is higher, a higher for the aqueous phase. As the β-diketone hexafluoroacetylace-
viscosity slows down the extraction kinetics in two ways. Thus, tone meets these requirements, it was used as extractant. The
longer stirring times are required for highly viscous solvent ionic liquid [Chol][hfac] was preferred instead of molecular
mixtures to reach equilibrium. Moreover, mixing of highly hexafluoroacetylacetone (Hhfac) since the ionic liquid is much
viscous ionic liquids requires a large energy input.
less volatile than Hhfac.56 Moreover, the extractant moiety is
In our previous work, homogeneous liquid–liquid extraction already present in its deprotonated form, making it immedi-
(HLLE) was introduced as an alternative technique to circum- ately available for coordination and no pH adjustments have to
vent the issue of the high viscosity of ionic liquids.19,20 It is a be made. In principle, hexafluoroacetylacetone ionic liquids
novel extraction technique that makes use of ionic liquid with cations other than Chol+ could also be used, such as
systems with thermomorphic behavior, i.e. with a temperature- 1-butyl-3-methylimidazolium or tetrabutylphosphonium.56,57
dependent phase behavior and mutual solubility.21–26 In an However, the composition of the extraction system was kept as
HLLE process, an aqueous feed solution containing the metal simple as possible by using the same Chol+ cation for both
to be extracted is mixed with a thermomorphic ionic liquid extractant and diluent. β-Diketones, such as hexafluoroacetyl-
and heated above the upper critical solution temperature acetone, are known for their coordination behavior with respect
(UCST). At this temperature, the organic phase is completely to lanthanide ions and are of general interest because of their
miscible with water and a homogeneous solution is obtained luminescence properties.58 Lanthanide β-diketonates have also
(homogeneous stage). Complex formation can now occur over been studied for liquid–liquid extraction, e.g. in the context of
the entire volume of the solution and is no longer hindered by the separation of fission products from spent nuclear fuel,
the interface. The speed of extraction only depends on the mostly with 2-thenoyltrifluoroacetone (Htta) as a ligand.58–62
kinetics of the chemical reactions and extraction equilibrium
is reached faster. Upon cooling, phase separation occurs
and the metal complex is extracted to the organic phase
Experimental
(settling stage). HLLE was first described by Murata et al. for
Products
the extraction of iron(III) thenoyltrifluoroacetonate in propy-
lene carbonate.27 In the following decades, HLLE has been
Neodymium(III) chloride hexahydrate (99.9%), neodymium(III)
nitrate hexahydrate (99.9%) and neodymium(III) oxide (99.99%)
were purchased from Alfa Aesar (Karlsruhe, Germany). Lithium
bis(trifluoromethylsulfonyl)imide (99%) was obtained from
IoLiTec (Heilbronn, Germany). Hexafluoroacetylacetone (98%)
was obtained from ABCR GmbH (Karlsruhe, Germany). Choline
chloride (99%) was purchased from Acros Organics (Geel,
Belgium). Choline hydroxide (46 wt% in water), sodium chloride
(>98%), hydrogen bromide (48 wt%) and silver nitrate (>99.8%)
were obtained from Sigma-Aldrich (Diegem, Belgium). HNO3
(65%) was obtained from Chem-Lab (Zedelgem, Belgium) and
absolute ethanol from VWR (Leuven, Belgium). A 1000 mg L−1
dysprosium ICP standard was purchased from Merck (Overijse,
Belgium). The silicone solution in isopropanol was obtained
from SERVA Electrophoresis GmbH (Heidelberg, Germany).
mainly applied to the extraction and pre-concentration of metals
for analytical purposes, extraction of organic compounds and
extraction of biological material.28–36 Examples of hydrometal-
lurgical applications, however, are scarce and make use of mole-
cular solvents only.37–41 To the best of our knowledge, HLLE has
not yet been applied to hydrometallurgical applications with
ionic liquids, besides the previous work by our group.19,20
In this paper, the potential of the ionic liquid choline bis-
(trifluoromethylsulfonyl)imide, [Chol][Tf2N], for use in HLLE
of neodymium(III) is described (Fig. 1).26,42,43 Neodymium is
one of the most critical rare earths and is highly demanded for
use in permanent magnets, so that efficient recovery methods
for this element are being developed by several research
groups.44–55 [Chol][Tf2N] has been shown to exhibit thermo-
morphic behavior in water, with an UCST of 72 °C.26 Since no
specific chelating functions are present in the structure of
Equipment
[Chol][Tf2N], an extractant was added. Only polar extractants 1H and 13C NMR spectra were recorded in D2O on a Bruker
could be used in combination with [Chol][Tf2N], since this Avance 300 spectrometer, operating at 300 MHz for 1H and
ionic liquid is quite polar and non-polar extractants have a 75 MHz for 13C. 19F NMR spectra were recorded on a Bruker
poor solubility in this ionic liquid. Moreover, the complex Avance II+ 600 MHz spectrometer with a SCF F-H-D-05 probe
formed between Nd(III) and the extractant upon extraction (Bruker Biospin) at a frequency of 565 MHz. The temperature
should have a higher affinity for the ionic liquid phase than was controlled using a Bruker BCU 05 temperature controller,
with a thermocouple type T. A 5 mL pycnometer was used to
determine densities. FTIR spectra were recorded on a Bruker
Vertex 70 spectrometer (Bruker Optics). Both solid and liquid
samples were examined directly without further sample prepa-
ration using a Platinum ATR single reflection diamond attenu-
ated total reflection (ATR) accessory. Analysis was performed
with OPUS software. Carbon, hydrogen and nitrogen content
were determined using a CE Instruments EA-1110 element
Fig. 1 Chemical structure of (a) [Chol][Tf2N] and (b) [Chol][hfac].
analyzer. Water content was measured with a Mettler–Toledo
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 11566–11578 | 11567

View Article Online
Paper
Dalton Transactions
DL39 coulometric Karl Fischer titrator. Extraction experiments extraction mixture. Because of its hygroscopic behavior, a
were performed in a TMS-200 thermoshaker (Nemus Life). A [Chol][Tf2N] solution containing a certain amount of water
Bruker Picofox S2 total reflection X-ray fluorescence (TXRF) (8 wt% in this case) was used instead of pure dry ionic liquid,
spectrometer was used to determine the metal concentrations in order to prevent further uncontrolled uptake of water. More-
in both aqueous and organic (ionic liquid) phase and the over, water-containing [Chol][Tf2N] is liquid, thus easier to dis-
chloride content of the synthesized [Chol][Tf2N]. Extraction pense than solid [Chol][Tf2N].
mixtures were centrifuged using a Heraeus Labofuge 200. X-ray
intensity data were collected at 100 K on an Agilent Supernova
diffractometer, equipped with an Atlas CCD detector, using
Extraction experiments
At a temperature of 20 °C, 12 wt% of [Chol][Tf2N] dissolves in
Mo Kα radiation (λ = 0.71073 Å). The images were interpreted
and integrated with the CrysAlisPro software from Agilent
water and 10 wt% of water dissolves in [Chol][Tf2N].26 During
a typical extraction experiment, this mutual solubility was
taken into account and the ionic liquid phase and water were
always mixed in such proportions that two phases were
obtained in a 1 : 1 mass ratio in the final extraction mixture
after phase equilibration. A 1 : 1 mass ratio simplifies calcu-
lation of the distribution ratio D and percentage extraction
%E (eqn (1) and (2)). In order to obtain a 1 : 1 ratio, 1109 mg
Technologies. Using Olex2, the structure was solved with the
SHELXS structure solution program using Direct Methods and
refined with the SHELXL refinement package using full-matrix
least squares minimization on F2.63–65
Synthesis of choline bis(trifluoromethylsulfonyl)imide
Choline bis(trifluoromethylsulfonyl)imide ([Chol][Tf2N]) was
prepared following a literature procedure.25 Choline chloride
(60.085 g, 0.430 mol) was mixed with an equimolar amount of
of ionic liquid phase, containing
a certain amount of
[Chol][Tf2N] extractant (50 mmol kg−1, unless otherwise speci-
fied) was mixed with 891 mg of metal feed solution (nitrate,
chloride or bromide medium). The metal concentration in the
aqueous feed was 15 mmol kg−1 unless otherwise specified.
The initial Nd(III) concentration in the aqueous phase was
recalculated according to the phase equilibrium to 12 mmol
kg−1 (unless otherwise specified). The extraction mixture was
heated and gently shaken in a thermoshaker at 700 rpm and
80 °C for 3 min to obtain one homogeneous phase. The
mixture was cooled to room temperature to induce settling
and two phases were again obtained. After waiting for 1 h to
make sure settling was completed, a small sample (100 mg)
was taken from the aqueous phase to determine the metal con-
centration in the aqueous phase with a TXRF spectrometer. A
small amount (100 mg) of dysprosium nitrate ICP standard
solution (1000 mg L−1) was added to the sample as the refer-
ence for quantification and further diluted with 1 mL of
MilliQ® water. A dysprosium standard was used because the
detected dysprosium L-lines are close to those of neodymium,
but there is no overlap. Thus, secondary absorption by the
ionic liquid matrix on the carrier has a similar influence on
the radiation of dysprosium and neodymium. The prepared
sample solution was homogenized and a small droplet (5 μL)
was dispensed on a quartz sample carrier. Finally, the carrier
containing the sample was dried at 60 °C for 15 min. Prior to
dispensing the sample droplet, the quartz carrier was pre-
treated with a SERVA® silicone solution in isopropanol (10 µL)
in order to make the surface more hydrophobic and keep the
sample droplet in the middle of the carrier.
lithium
bis(trifluoromethylsulfonyl)imide
(123.556
g,
0.430 mol) in 50 mL of distilled water. The mixture was stirred
for one hour at room temperature. The ionic liquid phase-
separated from the aqueous phase and was washed with small
amounts of ice-cold water in order to remove chloride impuri-
ties. The presence of chloride in the water layer after washing
was tested by the addition of a small aliquot of a concentrated
silver nitrate solution to the washing solution. Finally, the
ionic liquid was dried in vacuo on a rotary evaporator (60 °C). A
white solid product was formed. Yield: 74% (67.225 g,
1
0.176 mol) H NMR (300 MHz, [D6]DMSO, δ): 5.29 (t, 1H, OH,
J = 5.0 Hz); 3.85 (m, 2H, CH2); 3.41 (m, 2H, CH2); 3.12 (s, 9H,
3 × CH3). 13C NMR (75.47 MHz, [D6]DMSO, δ): 119.63 (q, 2 ×
CF3, J = 322 Hz); 67.23 (N–CH2); 55.29 (CH2); 53.34 (3 × CH3).
Elemental analysis: found (%): C, 21.49; H, 3.86; N, 7.20%;
calc. for C7H14N2O5F6S2: C, 21.88; H, 3.67; N, 7.29%. Water
content: 2200 mg kg−1. Chloride content: <10 mg kg−1
.
Density: 1.500 g cm−3 (60 °C). FTIR (cm−1) 1477 (CH2); 1348
(SO2); 1181 (CF3); 1138 (SO2); 1050 (SO2); 953; 867; 791; 741
(CF3); 611; 570 (CF3); 513 (CF3).
Preparation of the ionic liquid phase
As described in the literature, [Chol][hfac] is not stable at ele-
vated temperatures and decomposition occurs fast.56 Thus,
special drying techniques are required to obtain a pure and
dry product. However, since the extraction studies described in
this paper do not require dry [Chol][hfac], but [Chol][hfac] dis-
solved in the ionic liquid phase, [Chol][hfac] was prepared
directly in the ionic liquid [Chol][Tf2N]. Choline hydroxide
(46 wt% solution in water) and hexafluoroacetylacetone
(Hhfac) were added to dry [Chol][Tf2N] in equimolar quan-
tities. Water was added to this mixture to obtain a water
content of 8 wt%. A small amount of this [Chol][hfac]–[Chol]-
[Tf2N] mixture was then added to [Chol][Tf2N] (containing 8 wt%
of water) to obtain the organic (ionic liquid) phase of the
Extraction of Nd(III) is evaluated by calculating the distri-
bution ratio D and percentage extraction %E by using the
following equations:
½Mꢀorg
D ¼
ð1Þ
ð2Þ
½Mꢀaq
½Mꢀorg
%E ¼
100%
½Mꢀorg þ ½Mꢀaq
11568 | Dalton Trans., 2014, 43, 11566–11578
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
where [M]org and [M]aq are the metal ion concentration (in mol mixture to be studied, in a temperature-controlled water bath.
kg−1) in the organic phase and the aqueous phase, respect- The solvent mixture was heated and stirred until the mixture
ively. [M]aq was measured using TXRF. Since aqueous and turned homogeneous. Next, the heating was turned off and
organic phase were present in a 1 : 1 mass ratio, the metal ion the mixture was slowly cooled. The transparency of the
concentration in the organic phase was calculated as follows:
mixture was monitored visually and the temperature at which
the homogeneous mixture became cloudy was recorded. The
determination of this cloud point temperature was repeated
three times to get an accurate result. The mixtures all con-
tained 1109 mg of [Chol][Tf2N] solution (water content of 8 wt%
and a varying amount of [Chol][hfac]) and 891 mg of
water or diluted HNO3 (concentration varying between 0 and
2 mol L−1).
½Mꢀorg ¼ ½Mꢀaq;in ꢁ ½Mꢀaq
ð3Þ
where [M]aq,in and [M]aq are the initial and final (measured)
metal ion concentration in the aqueous phase, respectively.
Preparation of neodymium(III) bromide hydrate
Neodymium(III) bromide hydrate was prepared by adding
Nd2O3 (1 g, 3.0 mmol) to a HBr solution (48%, 3.006 g,
18 mmol), diluted in 10 mL of water. The mixture was stirred
for 1 h and NdBr3·xH2O was crystallized out of solution by
drying on a rotary evaporator.
Study of decomposition [Chol][hfac]
19F NMR spectroscopy was applied to monitor the decompo-
sition reaction of [Chol][hfac] with water. [Chol][hfac] was pre-
pared by adding choline hydroxide (46 wt% in water) to pure
hexafluoroacetylacetone. Immediately after mixing, a sample
was taken and diluted in deuterated methanol (CD3OD). The
sample was placed inside the NMR spectrometer and heated to
80 °C using the temperature controller of the NMR device.
Every 26 s, a 19F NMR spectrum was recorded. The sample was
not locked or shimmed before the start of the measurements
since the decomposition reaction of the pure [Chol][hfac] in
water occurred fast and locking and shimming would require
too much time. The spectrum reference (SR) was used because
standard 19F references evaporated at 80 °C and could not be
applied.
Study of temperature influence on extraction
The influence of temperature on extraction was studied using
four identical extraction mixtures (same procedure as above
for the extraction experiments). All mixtures were heated at
different temperatures (30 to 80 °C) for 2 min. The last 5 s of
heating, the mixtures were shaken (600 rpm). Next, the
mixtures were quenched in an ice bath and centrifuged
(5000 rpm, 1 min). The aqueous phase was isolated and its
neodymium content was determined by TXRF (same procedure
as above for the extraction experiments).
Stripping experiments
Stripping experiments were performed by mixing 1.0 g of Crystallography
[Chol][Tf2N] phase obtained after extraction of Nd(III), with
Single crystals of [Chol][Nd(hfac)4] suitable for X-ray diffrac-
1.0 g of diluted HNO3 solution (concentration ranging from
0 to 2 mol L−1). The mixtures were heated to 80 °C and shaken
for 3 min to obtain one homogeneous solution and again
cooled to room temperature to induce phase separation. After
1 h of settling, the metal content in the aqueous and organic
phase was measured with TXRF. The sample preparation of
the aqueous phase proceeded according to the method
described above. Samples of the organic phase were prepared
by adding a small amount of dysprosium standard solution
(1000 mg L−1) to 100 mg of the organic phase and diluting it
with 1 mL of pure ethanol. Approximately 2 μL of the prepared
sample solution was dispensed on the quartz glass carrier.
Stripping was evaluated using the stripping efficiency %S,
which is calculated as follows:
tion were obtained by crystallization upon cooling of an
extraction mixture (Nd concentration in feed was 0.1 mol
kg−1, extractant concentration in organic phase was 144 mmol
kg−1). The choline ligand was modeled in two positions with
an occupancy of 0.605(9) : 0.395(9). The second orientation
binds to an extra water molecule O52 (occupancy 0.395(9)).
The choline ligand and O52 were isotropically refined with
DFIX restraints on both choline orientations. Non-hydrogen
atoms of other parts were anisotropically refined and the
hydrogen atoms in the riding mode with isotropic tempera-
ture factors were fixed at 1.2 times Ueq of the parent atoms
(1.5 for methyl groups). The anisotropic displacement para-
meters of the fluorine atoms were restrained to be approxi-
mately isotropic and components of the displacement
parameters in the directions of the C–F bond were restrained
to be equal.
½Mꢀaq;str
%S ¼
100%
ð4Þ
½Mꢀaq;str þ ½Mꢀorg;str
Crystal data for [Chol][Nd(hfac)4]: C25H17F24NNdO9.395, M =
1081.96 g mol−1, monoclinic, P21/n (no. 14), a = 11.03826(19)
Å, b = 18.5663(5) Å, c = 17.8663(3) Å, β = 92.0479(17)°, V =
where [M]aq,str is the metal concentration in the aqueous
phase after stripping and [M]org,str the metal concentration in
the organic phase after stripping.
3659.18(13) Å3, T = 100 K, Z = 4, ρcalcd = 1.964 g cm−3
,
μ(Mo Kα) = 1.592 mm−1, F(000) = 2101.0, crystal size 0.27 ×
Cloud point determination
0.22 × 0.16 mm3, 26 960 reflections measured, 7466 unique
Cloud point determinations were performed by placing a (Rint = 0.0266) which were used in all calculations. The final
4 mL vial, closed with a cap and containing the solvent wR2 was 0.1290 (all data) and R1 was 0.0481 (I > 2σ(I)).
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 11566–11578 | 11569

View Article Online
Paper
Dalton Transactions
Results and discussion
Variation of extraction parameters
Homogeneous liquid–liquid extraction was carried out with
the [Chol][Tf2N]–[Chol][hfac] extraction system (Fig. 2). [Chol]-
[Tf2N] shows thermomorphic behavior with an UCST of
72.1 °C.26 Some additives to the aqueous phase of a [Chol]-
[Tf2N]–H2O mixture increase the UCST, such as inorganic salts
like Nd(NO3)3 in this case. Therefore, the extraction mixtures
were heated to 80 °C to assure that a homogeneous solution is
formed during the mixing stage.
For HLLE, it is in principle sufficient to just heat the extrac-
tion mixture above the UCST to obtain a homogeneous phase,
without applying any agitation. However, similarly to [Hbet]-
[Tf2N], diffusion of the [Chol][Tf2N] molecules into the
aqueous phase is too slow to obtain a homogeneous phase in
Fig. 3 Influence of temperature on %E of Nd(III) from a nitrate solution
with [Chol][hfac]–[Chol][Tf2N]. 50 mmol kg−1 [Chol][hfac] in organic
a short time scale.19,20 Thus, the mixture should be gently phase, 12 mmol kg−1 Nd(III) in aqueous phase.
shaken (5 s) to obtain a homogeneous phase. To show the
advantage of the formation of a homogeneous phase during
extraction, this procedure (heating the extraction mixture to the
performed on a shorter time scale compared to conventional
required temperature and shaking it for 5 s) was applied to the
extraction of Nd(III) from a nitrate solution at different tempera-
tures between 30 and 80 °C (Fig. 3). To obtain the required
temperature, the mixtures were heated for 2 min. Next, they
were shaken for 5 s and immediately quenched in an ice bath.
At low temperatures (30–45 °C), almost no Nd(III) was extracted.
At higher temperatures (60 °C), but still below the UCST, extrac-
tion of Nd(III) occurred to some extent, but still less than 50% of
Nd(III) was extracted. The extracted amount is a little higher at
60 °C compared to 30 °C, since an increase in temperature
lowers the viscosity of the ionic liquid phase, thus facilitating
diffusion of ions through the films near the interface. Extraction
at 80 °C (above the UCST) clearly showed the positive effect of
the formation of a homogeneous stage. After shaking for 5 s, a
homogeneous phase was obtained and immediately, the extract-
able complex was formed resulting in almost 100% extraction.
To obtain the same %E at temperatures below the UCST, the
extraction mixture had to be shaken for 2 min at 65 °C and up
to 10 min at 30 °C. So, it can be concluded that HLLE can be
liquid–liquid extraction and the extraction mixtures have to be
shaken for a much shorter time.
The influence of several extraction parameters on %E was
examined: the extractant concentration, pH and metal feed
concentration. The percentage extraction increased when the
amount of [Chol][hfac] extractant in the extraction mixture was
raised (Fig. 4). Almost 100% extraction was achieved with
60 mmol kg−1 [Chol][hfac] in the organic phase. The same
extraction behavior as a function of [Chol][hfac] concentration
was observed for extraction from chloride solution, meaning
that the counter ion in the aqueous metal solution (being
chloride or nitrate) has no influence on the extraction behav-
ior. Extraction from both types of feed solution proceeded
equally well. From Fig. 4, it can be seen that the pH increased
Fig. 2 Homogeneous liquid–liquid extraction of Nd(III) (purple color)
from a nitrate solution with [Chol][hfac] (yellow color) dissolved in
[Chol][Tf2N]. Left: before extraction (20 °C), middle: during homo-
geneous stage (80 °C), right: after settling (20 °C). The upper phase is
the aqueous phase, the lower phase is the ionic liquid phase.
■
Fig. 4 %E ( ) of Nd(III) from a nitrate solution with [Chol][hfac]–[Chol]-
[Tf2N] as a function of the [Chol][hfac] concentration in the organic
phase. 12 mmol kg−1 Nd(III) and 0.01 M HNO3 in aqueous phase, equili-
brium pH is indicated with
□
.
11570 | Dalton Trans., 2014, 43, 11566–11578
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
with increasing [Chol][hfac] concentration. This can be ration-
alized considering the following acid dissociation equilibrium:
Hhfac ⇄ hfacꢁ þ Hþ
ð5Þ
HNO3 was initially added to the extraction mixtures to keep
the pH low enough to prevent hydrolysis of Nd(III) ions. The
amount of protons (H+) present in the aqueous start solution
was the same in every extraction mixture (0.01 mmol). When
hfac− is added (e.g. as [Chol][hfac]), the acid dissociation equi-
librium of Hhfac shown in eqn (5) was shifted to the left and
the equilibrium proton concentration decreased, resulting in
an increasing pH value.
Hexafluoroacetylacetone (Hhfac) is an organic acid, with
hfac− as its conjugated base. The acid concentration in the
aqueous phase is expected to influence the extraction, since
part of the hfac− anions are protonated when acid is present.
It was shown that at pH 5.5 (which corresponds to an extrac-
tion mixture without extra HNO3 added to the aqueous phase),
nearly all Nd(III) ions were removed from the aqueous
phase (Fig. 5). No precipitation was observed, so the removal
of Nd(III) from solution was completely due to extraction to the
organic phase. Moreover, it can be concluded that no non-
extractable hydrolysis products were formed, since all metals
are extracted and a %E of almost 100% was obtained. So, it
is not necessary to add extra acid to the feed of the extraction
mixture to prevent hydrolysis. Additionally, from Fig. 5 it can
be seen that by lowering the equilibrium pH after addition of
HNO3 to the aqueous phase, the percentage extraction
decreased sharply between pH 3–5. At pH < 2, extraction of
Nd(III) is almost negligible since all [hfac]− anion are proto-
nated and the ionic liquid [Chol][hfac] is destroyed.
Fig. 6 %E of Nd(III) with [Chol][hfac]–[Chol][Tf2N] as a function of the
initial Nd(III) concentration in the aqueous phase. 50 mmol kg−1 [Chol]-
[hfac] in organic phase, 4–12 mmol kg−1 Nd(III) and 0.01 M HNO3 in
aqueous phase.
tration because the amount of extracted metal ions is always
approximately constant (Fig. 6).
The extraction capacity of the system was determined by
examining the maximum loading of Nd(III) in the organic
phase. It was shown earlier that 100% extraction could be
obtained for an extractant concentration of 60 mmol kg−1 or
higher (Fig. 4). This counts for a Nd(III) feed solution of
15 mmol kg−1. To increase the loading of the organic phase,
the metal feed concentration and the extractant concentration
were increased. When testing the extraction from a 100 mmol
kg−1 Nd(III) solution (excess), precipitation of the extractable
Nd(III) complex was observed for [Chol][hfac] concentrations of
more than 144 mmol kg−1. To determine how much Nd(III) can
maximally be extracted with 144 mmol kg−1 of extractant,
extraction was tested from different concentrations of Nd(III)
feed solution, between 10 and 100 mmol kg−1. As expected,
the percentage extraction %E decreased with increasing [Nd]aq,in
(Fig. 7), similarly as shown earlier in Fig. 6. The neodymium
concentration in the organic phase [Nd]org logically increases
with increasing Nd(III) feed concentration, up to 43 mmol
kg−1. Thus, it can be concluded that the maximum loading of
the [Chol][Tf2N] ionic liquid with the Nd(III)–[Chol][hfac]
complex is equal to 43 mmol kg−1. Beyond this point, the
complex is no longer completely soluble in the organic phase
and precipitates.
The influence of the metal ion concentration in the
aqueous feed on %E was also investigated, at constant pH and
at constant extractant concentration. As expected, the percen-
tage extraction decreased with increasing metal feed concen-
Extraction stoichiometry
To get a better understanding of the extraction mechanism, a
slope analysis was performed. The extraction equilibrium can
be described by eqn (6):
Nd3þ þ nhfacꢁ NdðhfacÞ
ð6Þ
3ꢁn
!
n
The bar indicates species in the organic phase. During the
homogeneous stage of the extraction, the dissociated hexa-
fluoroacetylacetonate anion coordinates to the Nd(III) ion.
Upon phase separation during the settling stage, the formed
Fig. 5 %E of Nd(III) from a nitrate solution with [Chol][hfac]–[Chol]-
[Tf2N] as a function of the equilibrium pH of the aqueous phase at room
temperature. 50 mmol kg−1 [Chol][hfac] in organic phase, 12 mmol kg−1
Nd(III) and 0–0.1 M HNO3 in aqueous phase.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 11566–11578 | 11571

View Article Online
Paper
Dalton Transactions
■
□
Fig. 7 Maximum loading of the organic phase: %E and Nd concen-
tration in the organic phase after extraction with [Chol][hfac]–[Chol]-
[Tf2N] as a function of the initial Nd(III) concentration in the aqueous
phase. 144 mmol kg−1 [Chol][hfac] in organic phase. Equilibrium pH was
constant at pH 6.
Fig. 8 Log D versus log [Chol][hfac] for the extraction of Nd(III) from a
nitrate solution with [Chol][hfac]–[Chol][Tf2N]. 30–65 mmol kg−1 [Chol]-
[hfac] in organic phase, 12 mmol kg−1 Nd(III) in aqueous phase initially.
Equilibrium pH was constant at pH 6. Dashed line indicates a slope of
4. The deviations at low and high extractant concentrations are dis-
cussed in the text.
complex diffuses to the bulk of the organic phase. The corres-
ponding equilibrium constant Kex is expressed as:
ruled out that the small deviation in slope is due to changes in
the phase behavior which could influence the calculated per-
centage extraction. As will be described in the section ‘Phase
behavior’ at the end this paper, it was observed that the pres-
ence of [Chol][hfac] in the ionic liquid phase had a lowering
effect on the cloud point. However, at an ionic strength of
1 mol L−1, the mutual solubility of ionic liquid and water, and
thus the mass of the phases, remains approximately constant
regardless of the amount of [Chol][hfac] present. Thus, the
increase in slope of the plot of log D versus log [Chol][hfac] in
Fig. 8 is most likely due to the approximation of [hfac−] by the
initial concentration of [Chol][hfac]. At low and intermediate
extractant concentrations compared to the initial Nd(III) con-
centration, the equilibrium extractant concentration [hfac−] is
proportional to the initial concentration of [Chol][hfac].
However, at higher concentrations (excess) of extractant com-
pared to Nd(III), this no longer holds. The initial extractant
concentration of [Chol][hfac] is no longer a good approxi-
mation of the equilibrium concentration [hfac−]. This explains
h
i
3ꢁn
NdðhfacÞn
Kex
¼
ð7Þ
ꢀ
ꢁꢀ
ꢁ
n
Nd3þ hfacꢁ
In this equation, the equilibrium concentration of Nd-
3−n
(hfac)n
is equal to the analytical concentration of neody-
mium in the organic phase [Nd]org and [Nd3+] is equal to the
analytical concentration of neodymium in the aqueous phase,
[Nd]aq. Thus, the distribution ratio D can be written as:
h
i
3ꢁn
NdðhfacÞn
½Ndꢀorg
½Ndꢀaq
ꢀ
ꢁ
D ¼
¼
ð8Þ
Nd3þ
After substitution of (8) in (7) and taking the logarithm of
both sides of eqn (7), log D is expressed as being linearly
related to n times the logarithm of [hfac−]:
log D ¼ log Kex þ n log½hfacꢁꢀ
ð9Þ
The slope of a plot of log D versus log [hfac−] is equal to n, the increase in slope in the plot of log D versus log [Chol][hfac]
which is the number of coordinated hfac− ligands per Nd(III) with higher [Chol][hfac] concentrations. Also, at very low
ion. Notice that [hfac−] is the equilibrium concentration of extractant concentrations, almost no Nd(III) is extracted and
free hexafluoroacetylacetonate anions, which is an unknown the initial [Chol][hfac] concentration is not a good estimation
parameter. It is approximated by the initial concentration of of the equilibrium [hfac−] concentration.
[Chol][hfac] (in mmol kg−1) in the ionic liquid phase.
When drawing a straight line with a slope of 4 in Fig. 8, it
Log D versus log [Chol][hfac] for the extraction of Nd(III) can be seen that this line fits well the middle part of the plot.
with [Chol][hfac]–[Chol][Tf2N] does not show a perfectly linear This implies that Nd(III) is extracted by [Chol][hfac] in a 1 : 4
relation (Fig. 8). When drawing the tangent at different parts stoichiometry and a tetrakis Nd(III) hexafluoroacetylacetonate
of the plot, it can be seen that the slope of the tangent varies complex is formed upon extraction. The formation of anionic
between a value of 4 at low extractant concentrations and a tetrakis β-diketonate complexes is often observed for the
value larger than 4 at high extractant concentrations. The extraction of lanthanide ions into ionic liquids.66 The avail-
same non-linear behavior was observed when extracting from ability of a counter ion and the possibility for ion exchange
an aqueous phase with a constant high ionic strength of 1 mol into a second phase enable the unexpected formation of a tetra-
L−1, by addition of NaCl to the aqueous phase. Thus, it can be kis complex in an aqueous system. However, the slope analysis
11572 | Dalton Trans., 2014, 43, 11566–11578
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
alone does provide valid proof for the extraction of a Nd(III) the first coordination sphere of the lanthanide ion.57,67–72
complex with a 1 : 4 stoichiometry.
Coordination of a water molecule to neodymium(III) tetrakis-
hexafluoroacetylacetonate complexes has been reported.73,74
Also in some lanthanide tetrakis 3-cyanopentane-2,4-dionate
complexes additional coordination of water molecules occurs
to give a nine-coordinate complex.75 Recently, Nockemann and
coworkers reported a crystal structure of a europium(III) tetra-
kis hexafluoroacetylacetonate complex with coordination of
the oxygen atom of a imidazolium-functionalized diphenyl-
phosphinoxide.76 The structure of the [Chol][Nd(hfac)4]
complex is somewhat similar to the latter complex, since the
coordinated ligand also provides the counter cation for the
anionic tetrakis complex. The unusual structure of [Chol]-
[Nd(hfac)4] is most likely due to a combination of different factors:
the increased Lewis acidity of the neodymium(III) ion due to
the fluorinated ligands, sterically little demanding β-diketo-
nate ligands and a small cation with a Lewis basic group (the
choline cation with its primary alcohol function).
Crystal structure of [Chol][Nd(hfac)4]
The assumption of the formation of a tetrakis Nd(III)hexafluoro-
acetylacetonate complex as derived above in the previous
section, was confirmed with single-crystal X-ray diffraction.
Upon increasing the loading of extracted Nd(III) in the ionic
liquid phase, crystal formation occurred during the settling
(cooling) stage. Compound [Chol][Nd(hfac)4] crystallizes in the
centrosymmetric space group P21/n with one molecule in the
asymmetric unit (Fig. 9). The central Nd(III) ion is coordinated
by four hexafluoroacetylacetonate ligands and one choline
ligand to form a nine-coordinate complex with a capped
square antiprism as coordination polyhedron. The choline
ligand is disordered over two positions with an occupancy of
0.605(9) : 0.395(9). The Nd–O distances between the Nd and
the hfac− ligands range between 2.429(4) Å and 2.489(4) Å. The
Nd–O distances of the choline ligand are significantly larger.
The distances are respectively 2.550(8) Å for the first orien-
tation and 2.618(16) Å for the second part. The second posi-
tion of the choline ligand leaves room for a water solvent
molecule, which interacts by hydrogen bonding with a nearby
fluorine atom (O52⋯F24i: 2.651(7) Å, symmetry code: (i) [−X,
−Y, −Z]) and the choline oxygen (O52⋯O51: 2.625(8) Å). In the
crystal packing three intermolecular non-classical hydrogen
bonds are noticeable. Their donor acceptor distances are
respectively: C3⋯F11 3.352(7) Å, C44⋯F22 3.075(11) Å and
C44⋯F3 3.172(9) Å. The crystal structure is remarkable,
because it is one of the rare examples of a nine-coordinate
lanthanide tetrakis β-diketonate complex.58 In general, this
type of complexes has an eight-coordinate lanthanide center,
with the four bidentate β-diketonate ligands saturating
Ion exchange
To preserve electrical neutrality upon extraction of the tetrakis-
Nd(III) complex, ion exchange has to occur between the organic
and the aqueous phase. Per extracted Nd(III) ion, one of the
four dissociated Chol+ cations neutralizes the extracted [Chol]-
[Nd(hfac)4] complex. The other three Chol+ cations either stay
in the organic phase or diffuse to the aqueous phase. When
the Chol+ cations are staying in the organic phase, nitrate ions
should be co-extracted to neutralize the three Chol+ cations. To
determine which hypothesis is valid, extraction was carried out
from a bromide solution. As described in the beginning of the
paper, it was seen that the type of neodymium salt, being
nitrate or chloride, did not influence the extraction perform-
ance. Since measurement of nitrate or chloride content by
TXRF is difficult, extraction was performed from a bromide
solution. TXRF is a suitable technique to measure bromide
content in ionic liquids.77 It was observed that no bromide was
co-extracted with the metal ions, which implies that the three
Chol+ cations diffuse to the aqueous phase. So, in exchange
for one extracted Nd(III) ion, three Chol+ are transferred to the
aqueous phase. To confirm the ion-exchange mechanism, the
influence of the addition of choline chloride to the aqueous
phase was examined. It was observed that the %E decreased sig-
nificantly with increasing choline chloride content (Fig. 10).
Thus, addition of choline chloride to the aqueous phase pushes
back the extraction equilibrium, which confirms that Chol+
cations migrate to the aqueous phase upon Nd(III) extraction.
Ion exchange is a typical mechanism for extraction of metal
ions into ionic liquids.59,78,79 Depending on the type of
extracted complex, being anionic or cationic, anions or cations
of the ionic liquid diffuse to the aqueous phase in exchange
for the extracted metal complex. In some extraction systems
with ionic liquids, the occurrence of ion exchange enhances
metal extraction to the ionic liquid phase compared to extrac-
tion to the organic phase with traditional molecular diluents
and could be considered an advantage.59 Still, loss of ions of
the ionic liquid to the aqueous phase by the ion exchange
Fig. 9 Crystal structure of [Chol][Nd(hfac)4]. Only one part of the
choline ligand orientation is shown.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 11566–11578 | 11573

View Article Online
Paper
Dalton Transactions
Fig. 11 Stripping of a [Chol][Tf2N] phase, loaded with Nd(III) (12 mmol
Fig. 10 Influence of the addition of choline chloride to the aqueous
phase on the %E of Nd(III) from a nitrate solution with [Chol][hfac]–
[Chol][Tf2N]. 50 mmol kg−1 [Chol][hfac] in organic phase, 12 mmol kg−1
Nd(III) in aqueous phase. Equilibrium pH was constant.
kg−1
)
after extraction with [Chol][hfac], with varying HNO3
concentration.
Mixing of the loaded [Chol][Tf2N] with a diluted HNO3 solu-
tion with a concentration of 0.1 M resulted in 98% stripping
efficiency of Nd(III) from the ionic liquid phase. All metal ions
were removed by stripping with a 1 M HNO3 solution (Fig. 11).
It is preferable to perform stripping in one step. Each time the
ionic liquid phase is brought in contact with an aqueous solu-
tion, part of the [Chol][Tf2N] diluent dissolves in the aqueous
phase, which leads to significant losses of valuable ionic
liquid. Determination of the loss of [Chol][Tf2N] is not
straightforward, since these losses depend on the composition
of the extraction mixture (acid concentration, metal concen-
tration, extractant concentration). After stripping, the ionic
liquid phase was available for a new extraction step. Some
HNO3 was dissolved in the ionic liquid phase, but this was
easily neutralized with a base, such as NaOH.
mechanism is a major limitation for the practical application
of ionic liquid solvent extraction systems. Part of the ionic
liquid is lost and the aqueous phase is contaminated. There-
fore, ion exchange should be avoided if possible.80 A possi-
bility to prevent the loss of Chol+ to the aqueous phase when
extracting Nd(III) with [Chol][hfac] is the replacement of three
of the [Chol][hfac] ionic liquid molecules by the molecular
acidic form of the extractant, namely hexafluoroacetylacetone
(Hhfac). Upon extraction of one Nd(III) ion, Hhfac exchanges
its protons to coordinate to the Nd(III) center. The fourth hfac−
ligand is provided, together with the counter ion, by the [Chol]-
[hfac] ionic liquid moiety. However, it was observed that about
50% less Nd(III) was extracted compared to extraction by [Chol]-
[hfac] only. In order to compare the obtained results with the
results of extraction with only [Chol][hfac], a small amount of
NaOH (35 μL, 1 M) was added to the extraction mixture to
obtain the same equilibrium pH as in previous experiments.
The same amount of hfac− was used (50 mmol kg−1), but
75 mol% was added as Hhfac and 25 mol% as [Chol][hfac],
instead of 100 mol% as [Chol][hfac]. This indicates that the
diffusion of Chol+ ions to the aqueous phase significantly
enhances extraction, which is in line with observations
described in the literature.78
Decomposition of [Chol][hfac]
The extraction efficiency of the [Chol][hfac] extractant
decreased when heated for a longer time at elevated tempera-
ture. Moreover, the extraction performance of the ionic liquid
phase also decreased significantly with increasing storage time
(Fig. 12). After 14 days of storage at room temperature (20 °C),
%E for the extraction of Nd(III) was reduced to less than 50%
of the percentage extraction at day 1. The degradation of the
[Chol][hfac]–[Chol][Tf2N] mixture is slowed down by storing
the mixture at lower temperatures (for instance in a refriger-
ator at 6 °C), but still %E decreased with more than 10% after
14 days of storage at 6 °C. The decrease in extraction perform-
ance was due to reactions between hexafluoroacetylacetone
and water, forming decomposition products such as trifluoro-
acetic acid and trifluoroacetone (Fig. 13) or Hhfac–water
adducts, such as the gem-diol product.81,82
Stripping studies
The extracted Nd(III) ions were again stripped from the loaded
[Chol][Tf2N] phase by mixing with a diluted HNO3 solution.
When acid was added to the loaded [Chol][Tf2N], the neody-
mium complex dissociated into free metal ion and protonated
ligands:
It has also been described in the literature that dry [Chol]-
[hfac] has a low thermal stability.56 The decomposition temp-
ð10Þ erature determined with TGA is stated to be equal to 83 3 °C,
½Cholꢀ½NdðhfacÞ4ꢀ þ 3Hþ Nd3þ þ 3Hhfac þ ½Cholꢀ½hfacꢀ
!
11574 | Dalton Trans., 2014, 43, 11566–11578
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 12 Decrease in extraction performance of a [Chol][hfac]–[Chol]-
[Tf2N] mixture containing 8 wt% water, as a function of storage time and
storage temperature: 20 °C (solid line) and 6 °C (dashed line). Initial
amount of HNO3 in the aqueous phase is 0.01 mmol.
Fig. 14 Decomposition reaction of [Chol][hfac] with water at 80 °C
monitored by 19F NMR spectroscopy.
Fig. 13 Decomposition reaction of hexafluoroacetylacetone with water
to trifluoroacetone and trifluoroacetic acid.
the extraction mixture, the cloud point of [Chol][Tf2N]–H2O
mixtures containing varying amounts of HNO3 and [Chol]-
[hfac] was determined. The cloud point is the temperature at
which the mixture becomes cloudy upon cooling. At this point,
water and the ionic liquid are no longer completely miscible
and they will phase separate. An ionic liquid rich organic
phase is formed, together with an ionic liquid lean aqueous
phase. The presence of HNO3 at low concentrations (<0.5 M) in
the IL–H2O mixture increased the cloud point with approxi-
mately 1 °C. At higher HNO3 concentrations (>1 M) however,
the cloud point decreased several degrees Celsius (Fig. 15).
These observations are similar to the phase behavior of imida-
zolium ionic liquids described in the literature.83,85 The
change in water solubility of a solute upon addition of certain
ions is related to salting-in/salting-out effects. A decrease in
water solubility is known as salting-out, an increase in solubi-
lity is known as salting-in. The properties of different ions
with respect to the salting of ionic liquids follow the Hofmeis-
ter series.85 It must be noted that these salting phenomena are
complex and depend both on the type of solute and the concen-
tration of the salting ion. The nitrate ion is located at the end
of the Hofmeister series, i.e. it generally has a weak salting-out
or even a salting-in effect on the solubility of organic com-
pounds in water. Thus, the slight increase in cloud point at
low HNO3 concentrations, corresponding to a slight decrease
but it can be seen from the reported thermogram that a mass
loss already occurs at temperatures below 80 °C.
Decomposition of [Chol][hfac] was monitored by 19F NMR
measurements. A mixture of [Chol][hfac] and water was kept at
80 °C and a 19F NMR spectrum was recorded every 26 s. At the
start of the experiment, only one peak (a, at −77.89 ppm) was
visible in the 19F NMR spectrum, which was contributed to the
CF3 groups of hfac−. Nearly immediately after the start of the
experiment, it was seen that this peak slowly diminished. At
the same time, a second peak appeared near the peak of hfac−
(b, at −77.91 ppm) (Fig. 14), which is most likely caused by the
formation of trifluoroacetic acid. Moreover, a third peak, most
likely assigned to trifluoroacetone, and a fourth peak appeared
at −85.48 ppm (d) and −76.29 ppm (c), respectively. These
observations showed that decomposition is fast at 80 °C. Thus,
it is preferable to perform extractions on a short time scale
and to avoid long mixing times at elevated temperatures. This
conclusion confirms that it is preferable to use HLLE, since it
can be executed on a very short time scale. In this way, the
extraction process is fast and decomposition is avoided as
much as possible.
Phase behavior
The addition of extractants and the presence of acids and ions in water solubility of the ionic liquid and thus a small salting-
in the aqueous phase have an influence on the phase behavior out effect is also in line with observations described in the lit-
and the mutual solubility of the ionic liquid and water.83–85 To erature, where it was stated that the salting effect (in or out) of
examine the effect of the addition of [Chol][hfac] and HNO3 to a certain ion depends on the concentration of the corres-
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 11566–11578 | 11575

View Article Online
Paper
Dalton Transactions
tants could not be used. Thus, zwitterionic betaine was added
as an extractant. However, relatively high amounts of betaine
(200 g kg−1) were required to obtain good extraction efficien-
cies. The [Chol][Tf2N] ionic liquid described in this paper is
similar to [Hbet][Tf2N]. It has a similar chemical structure as
[Hbet][Tf2N], but with a hydroxyl function on the cation
instead of a carboxyl function, making it less acidic. They both
show a UCST-type of phase behavior with water. The critical
temperature of [Chol][Tf2N] (72 °C) is higher than [Hbet][Tf2N]
(55 °C), so higher temperatures are required during HLLE to
reach the homogeneous stage. Because of the lower acidity of
[Chol][Tf2N] in water, it is possible to use this ionic liquid as a
diluent in combination with (polar) acidic extractants, such as
β-diketones. In the extraction system described in this paper,
the [Chol][hfac] extractant is added to the organic phase in
stoichiometric amounts to obtain almost full extraction of
Nd(III), whereas an excess of zwitterionic betaine was required
to obtain high extraction efficiencies of Nd(III) in [Hbet][Tf2N].
Fig. 15 Influence of the concentration of HNO3 in the aqueous feed on
the cloud point of a [Chol][Tf2N]–H2O mixture. Standard deviation on
cloud point is 0.2 °C.
Conclusions
It was shown that the ionic liquid [Chol][Tf2N] is a suitable
diluent for the homogeneous liquid–liquid extraction of Nd(III)
from aqueous solution with [Chol][hfac] as extractant. Homo-
geneous liquid–liquid extraction is a novel energy-efficient
technique to perform ionic liquid solvent extraction, which is
based on the formation of a homogeneous phase during
mixing. [Chol][Tf2N] shows thermomorphic behavior in water,
with an UCST of 72 °C. The influence of several extraction
parameters was studied, including extractant concentration,
pH and Nd(III) feed concentration. At stoichiometric ratios of
Nd(III) and [Chol][hfac], nearly all metal ions were extracted
and loadings in the ionic liquid phase up to 43 mmol kg−1
were achieved. A speciation study was performed to determine
the extraction mechanism. Via slope analysis and crystal struc-
ture determination, it was shown that Nd(III) was extracted via
an anionic tetrakis(hexafluoroacetylacetonato)neodymate(III)
complex, with a choline cation as counter ion. Upon extrac-
tion, ion exchange occurred and three Chol+ ions were trans-
ferred to the aqueous phase. The crystal structure
determination of the extracted [Chol][Nd(hfac)4] complex
showed that the Chol+ counter ion was coordinated to the Nd(III)
center, forming a nine-coordinate complex. This is one of the
rare examples of coordination of the counter ion in tetrakis
lanthanide β-diketonate complexes to the lanthanide(III) ion.
Finally, Nd(III) was successfully stripped from the loaded ionic
liquid phase with a 1 M HNO3 solution.
Fig. 16 Influence of the concentration of [Chol][hfac] in the organic
phase on the cloud point of a [Chol][Tf2N]–H2O mixture. Standard devi-
ation on cloud point is 0.2 °C.
ponding salt.83 The addition of the extractant [Chol][hfac] to
the [Chol][Tf2N]–H2O mixture had a similar effect on the cloud
point. At low [Chol][hfac] concentrations, the cloud point
increased 1 °C (Fig. 16). At higher concentrations (>50 mmol
kg−1), the cloud point decreased again below 72.2 °C, which is
the cloud point of a mixture without the presence of any
additives.
Comparison with the [Hbet][Tf2N] system
Acknowledgements
In previous work, the protic ionic liquid [Hbet][Tf2N] was pres-
ented for HLLE of rare-earth ions.19,20 Because pure [Hbet]- This research was financially supported by the FWO-Flanders
[Tf2N] only poorly extracts rare-earth ions, an extractant was (research project G.0900.13), and KU Leuven (projects GOA/13/
added. [Hbet][Tf2N] has a high acidity due to the presence of 008 and IOF-KP RARE3). The authors thank the Hercules Foun-
the carboxyl function on the cation and typical acidic extrac- dation for supporting the purchase of equipment through the
11576 | Dalton Trans., 2014, 43, 11566–11578
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
project AKUL/09/0035, Molybdenum high-energy X-ray source 24 Y. Kohno and H. Ohno, Phys. Chem. Chem. Phys., 2012, 14,
for in situ diffraction studies of advanced materials and single 5063–5070.
crystals. B.O. thanks Karel Duerinckx for assistance to the 19F 25 P. Nockemann, B. Thijs, S. Pittois, J. Thoen, C. Glorieux,
NMR measurements and Tom Vander Hoogerstraete for useful
scientific discussions. Support by IoLiTec (Heilbronn,
Germany) is also gratefully acknowledged.
K. Van Hecke, L. Van Meervelt, B. Kirchner and
K. Binnemans, J. Phys. Chem. B, 2006, 110, 20978–20992.
26 P. Nockemann, K. Binnemans, B. Thijs, T. N. Parac-Vogt,
K. Merz, A. V. Mudring, P. C. Menon, R. N. Rajesh,
G. Cordoyiannis, J. Thoen, J. Leys and C. Glorieux, J. Phys.
Chem. B, 2009, 113, 1429–1437.
Notes and references
27 K. Murata, S. Ikeda and Y. Yokoyama, Anal. Chem., 1972,
44, 805–810.
1 J. Rydberg, M. Cox, C. Musikas and G. R. Choppin, Solvent
Extraction: Principles and Practice, Marcel Dekker, Inc., 28 H. Ebrahimzadeh, Y. Yamini, F. Kamarei and S. Shariati,
New York, 2nd edn, 2004. Anal. Chim. Acta, 2007, 594, 93–100.
2 C. K. Gupta and N. Krishnamurthy, Int. Mater. Rev., 1992, 29 A. Schaadt and H. J. Bart, Chem. Eng. Technol., 2003, 26,
37, 197–248. 469–472.
3 C. K. Gupta and N. Krishnamurthy, Extractive Metallurgy of 30 J. D. Lamb and R. T. Peterson, Sep. Sci. Technol., 1995, 30,
Rare Earths, CRC Press, Boca Raton, 2005.
4 M. L. Dietz, Sep. Sci. Technol., 2006, 41, 2047–2063.
5 Y. Liu, J. Chen and D. Li, Sep. Sci. Technol., 2012, 47,
223–232.
6 H. Zhao, S. Xia and P. Ma, J. Chem. Technol. Biotechnol.,
2005, 80, 1089–1096.
3237–3244.
31 A. R. Ghiasvand, S. Shadabi, E. Mohagheghzadeh and
P. Hashemi, Talanta, 2005, 66, 912–916.
32 A. R. Ghiasvand and E. Mohagheghzadeh, Anal. Sci., 2004,
20, 917–919.
33 S. Gao, H. Jin, J. You, Y. Ding, N. Zhang, Y. Wang, R. Ren,
R. Zhang and H. Zhang, J. Chromatogr., A, 2011, 1218,
7254–7263.
7 T. Welton, Chem. Rev., 1999, 99, 2071–2083.
8 K. R. Seddon, J. Chem. Technol. Biotechnol., 1997, 68, 351–356.
9 P. Wasserscheid and T. Welton, in Ionic Liquids in Synthesis, 34 M. A. Farajzadeh, M. Bahram, S. Zorita and B. G. Mehr,
ed. P. Wasserscheid and T. Welton, Wiley-VCH, Weinheim,
2nd edn, 2007, ch. 3, pp. 41–118.
10 P. Wasserscheid and T. Welton, Ionic Liquids in Synthesis,
J. Hazard. Mater., 2009, 161, 1535–1543.
35 A. N. Anthemidis and K. I. Ioannou, Talanta, 2009, 80,
413–421.
ed. P. Wasserscheid and T. Welton, Wiley-VCH, Weinheim, 36 T. Golan, G. Dahan, Z. Ludmer, N. Brauner and
2002. A. Ullmann, Chem. Eng. J., 2014, 236, 47–58.
11 R. D. Rogers and K. R. Seddon, Science, 2003, 302, 792–793. 37 N. Alizadeh and K. Ashtari, Sep. Purif. Technol., 2005, 44,
12 N. V. Plechkova and K. R. Seddon, Chem. Soc. Rev., 2008,
37, 123–150.
13 M. Freemantle, An Introduction to Ionic Liquids, RSC
Publishing, Cambridge, 2009.
14 G. Huddleston and D. Rogers, Chem. Commun., 1998,
1765–1766.
79–84.
38 M. H. Hosseini and N. Alizadeh, Ind. Eng. Chem. Res., 2010,
49, 7068–7073.
39 Y. Elistratova, A. R. Mustafina, D. A. Tatarinov, V. F. Mironov,
V. A. Burilov, I. G. Tananaev and A. I. Konovalov, Russ.
Chem. Bull., 2011, 60, 790–796.
15 J. McFarlane, W. B. Ridenour, H. Luo, R. D. Hunt, 40 J. Z. Xu, N. Rajapakse and H. L. Finston, Radiochim. Acta,
D. W. DePaoli and R. X. Ren, Sep. Sci. Technol., 2005, 40,
1245–1265.
16 C. F. Poole and S. K. Poole, J. Chromatogr. A, 2010, 1217,
2268–2286.
17 G. Yu, D. Zhao, L. Wen, S. Yang and X. Chen, AIChE J.,
2012, 58, 2885–2899.
18 S. Wellens, R. Goovaerts, C. Möller, J. Luyten, B. Thijs and
K. Binnemans, Green Chem., 2013, 15, 3160–3164.
19 T. Vander Hoogerstraete, B. Onghena and K. Binnemans,
J. Phys. Chem. Lett., 2013, 4, 1659–1663.
1990, 49, 135–140.
41 A. Takahashi, Y. Ueki and S. Igarashi, Anal. Chim. Acta,
1999, 387, 71–75.
42 K. Sasaki, K. Takao, T. Suzuki, T. Mori, T. Arai and Y. Ikeda,
Dalton Trans., 2014, 43, 5648–5651.
43 I. A. Shkrob, T. W. Marin and M. P. Jensen, Ind. Eng. Chem.
Res., 2014, 53, 3641–3653.
44 K. Binnemans, P. T. Jones, B. Blanpain, T. Van Gerven,
Y. Yang, A. Walton and M. Buchert, J. Cleaner Prod., 2013,
51, 1–22.
20 T. Vander Hoogerstraete, B. Onghena and K. Binnemans, 45 T. Vander Hoogerstraete, S. Wellens, K. Verachtert and
Int. J. Mol. Sci., 2013, 14. K. Binnemans, Green Chem., 2013, 15, 919–927.
21 Y. Fukaya, K. Sekikawa, K. Murata, N. Nakamura and 46 T. Itakura, R. Sasai and H. Itoh, J. Alloys Compd., 2006,
H. Ohno, Chem. Commun., 2007, 3089–3091. 408–412, 1382–1385.
22 Y. Fukaya and H. Ohno, Phys. Chem. Chem. Phys., 2013, 15, 47 M. Itoh, K. Miura and K. i. Machida, J. Alloys Compd., 2009,
4066–4072. 477, 484–487.
23 K. Fukumoto and H. Ohno, Angew. Chem., Int. Ed, 2007, 46, 48 O. Takeda, K. Nakano and Y. Sato, Mater. Trans., 2014, 55,
1852–1855.
334–341.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 11566–11578 | 11577

View Article Online
Paper
Dalton Transactions
49 H. Na, Y. Kim, H. Son, I. Jung and H. Choi, Curr. Nanosci., 68 P. Nockemann, E. Beurer, K. Driesen, R. Van Deun, K. Van
2014, 10, 128–130.
50 H. S. Yoon, C. J. Kim, K. Chung, S. J. Lee, A. R. Joe,
Hecke, L. Van Meervelt and K. Binnemans, Chem.
Commun., 2005, 4354–4356.
Y. H. Shin, S. I. Lee, S. J. Yoo and J. G. Kim, Korean J. Chem. 69 S. F. Tang and A. V. Mudring, Eur. J. Inorg. Chem., 2009,
Eng., 2014, 31, 706–711. 2009, 2769–2775.
51 C. H. Lee, Y. J. Chen, C. H. Liao, S. Popuri, S. L. Tsai and 70 K. Lunstroot, K. Driesen, P. Nockemann, K. Van Hecke,
C. E. Hung, Metall. Mater. Trans. A, 2013, 44, 5825–5833.
52 J. P. Rabatho, W. Tongamp, Y. Takasaki, K. Haga and
A. Shibayama, J. Mater. Cycles Waste Manage., 2013, 15,
171–178.
53 M. Nakamoto, K. Kubo, Y. Katayama, T. Tanaka and
T. Yamamoto, Metall. Mater. Trans. B, 2012, 43, 468–476.
L. Van Meervelt, C. Gorller-Walrand, K. Binnemans,
S. Bellayer, L. Viau, J. Le Bideau and A. Vioux, Dalton
Trans., 2009, 298–306.
71 K. Lunstroot, P. Nockemann, K. Van Hecke, L. Van Meer-
velt, C. Görller-Walrand, K. Binnemans and K. Driesen,
Inorg. Chem., 2009, 48, 3018–3026.
54 Y. Baba, F. Kubota, N. Kamiya and M. Goto, Solvent Extr. 72 S. Akerboom, M. S. Meijer, M. A. Siegler, W. T. Fu and
Res. Dev., Jpn., 2014, 15, 193–198. E. Bouwman, J. Lumin., 2014, 145, 278–282.
55 Y. Kikuchi, M. Matsumiya and S. Kawakami, Solvent Extr. 73 A. Mech, Polyhedron, 2008, 27, 393–405.
Res. Dev., Jpn., 2014, 21, 137–145.
74 F. Pointillart, O. Maury, Y. Le Gal, S. p. Golhen, O. Cador
56 S. Pandey, G. A. Baker, L. Sze, S. Pandey, G. Kamath,
H. Zhao and S. N. Baker, New J. Chem., 2013, 37, 909–919.
57 H. Mehdi, K. Binnemans, K. Van Hecke, L. Van Meervelt
and P. Nockemann, Chem. Commun., 2010, 46, 234–236.
58 K. Binnemans, in Handbook on the Physics and Chemistry of
Rare Earths, ed. K. A. Gschneidner Jr., J. C. G. Bünzli
and L. n. Ouahab, Inorg. Chem., 2009, 48, 7421–7429.
75 C. Merkens and U. Englert, Dalton Trans., 2012, 41,
4664–4673.
76 J. A. Vicente, A. Mlonka, H. Q. N. Gunaratne, M. Swadzba-
Kwasny and P. Nockemann, Chem. Commun., 2012, 48,
6115–6117.
and V. K. Pecharsky, Elsevier, Amsterdam, 2005, vol. 35, 77 T. Vander Hoogerstraete, S. Jamar, S. Wellens and
ch. 225, pp. 107–272. K. Binnemans, Anal. Chem., 2014, 86, 3931–3938.
59 I. Billard, A. Ouadi and C. Gaillard, Anal. Bioanal. Chem., 78 A. Stojanovic and B. K. Keppler, Sep. Sci. Technol., 2011, 47,
2011, 400, 1555–1566.
189–203.
60 N. Hirayama, Solvent Extr. Res. Dev., Jpn., 2011, 18, 1–14.
61 M. P. Jensen, M. Borkowski, I. Laszak, J. V. Beitz,
79 I. Billard, A. Ouadi and C. Gaillard, Dalton Trans., 2013, 42,
6203–6212.
P. G. Rickert and M. L. Dietz, Sep. Sci. Technol., 2012, 47, 80 X. Sun, H. Luo and S. Dai, Chem. Rev., 2011, 112, 2100–
233–243.
2128.
62 K. Binnemans, Chem. Rev., 2007, 107, 2592–2614.
63 G. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystallogr,
2008, 64, 112–122.
64 O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard
and H. Puschmann, J. Appl. Crystallogr., 2009, 42, 339–341.
65 CrysAlis PRO, Agilent Technologies UK Ltd, Yarnton,
Oxfordshire, England, 2012.
81 M. Ellinger, H. Duschner and K. Starke, J. Inorg. Nucl.
Chem., 1978, 40, 1063–1067.
82 D. A. Johnson, A. B. Waugh, T. W. Hambley and
J. C. Taylor, J. Fluorine Chem., 1985, 27, 371–378.
83 M. G. Freire, P. J. Carvalho, A. M. S. Silva, L. M. N. B. Santos,
L. P. N. Rebelo, I. M. Marrucho and J. A. P. Coutinho, J. Phys.
Chem. B, 2009, 113, 202–211.
66 M. P. Jensen, J. A. Neuefeind, J. V. Beitz, S. Skanthakumar 84 M. G. Freire, C. M. S. S. Neves, P. J. Carvalho, R. L. Gardas,
and L. Soderholm, J. Am. Chem. Soc., 2003, 125,
15466–15473.
A. M. Fernandes, I. M. Marrucho, L. M. N. B. Santos and
J. A. P. Coutinho, J. Phys. Chem. B, 2007, 111, 13082–13089.
67 R. Van Deun, P. Van Der Voort, I. Van Driessche and K. Van 85 L. I. N. Tomé, F. R. Varanda, M. G. Freire, I. M. Marrucho
Hecke, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2012,
68, m111–m112.
and J. A. P. Coutinho, J. Phys. Chem. B, 2009, 113,
2815–2825.
11578 | Dalton Trans., 2014, 43, 11566–11578
This journal is © The Royal Society of Chemistry 2014