Synthesis of organophosphorus ligands with a central oxygen atom and their applications in solvent extraction

Article abstract of DOI:10.1016/j.tet.2019.06.004

Various structurally related glycolamide compounds were synthesised and evaluated by solvent extraction experiments for the separation of rare earth elements (REEs) from aqueous acid solutions using dodecane as the diluent. We describe herein the full synthesis of a family of bifunctional ligands with a central oxygen and focus our investigation on the effects of structural modification on the extraction efficiencies. The combination of an amide, a P[dbnd]O donor site, and a central oxygen in such a glycolamide ligand showed interesting extraction properties for heavy rare earth elements from phosphoric acid solutions.

Full text of DOI:10.1016/j.tet.2019.06.004

Accepted Manuscript
Synthesis of organophosphorus ligands with a central oxygen atom and their
applications in solvent extraction
Fanny Mary, Guilhem Arrachart, Antoine Leydier, Stéphane Pellet-Rostaing
PII:
S0040-4020(19)30644-1
https://doi.org/10.1016/j.tet.2019.06.004
TET 30396
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 4 March 2019
Revised Date: 31 May 2019
Accepted Date: 3 June 2019
Please cite this article as: Mary F, Arrachart G, Leydier A, Pellet-Rostaing Sté, Synthesis of
organophosphorus ligands with a central oxygen atom and their applications in solvent extraction,
Tetrahedron (2019), doi: https://doi.org/10.1016/j.tet.2019.06.004.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical abstract for:
Synthesis of organophosphorus ligands with a central oxygen atom and their
applications in solvent extraction
Fanny Marya, Guilhem Arracharta,*, Antoine Leydiera, Stéphane Pellet-Rostainga, *
a ICSM, CEA, CNRS, ENSCM, Univ Montpellier, Marcoule, France.
Email: guilhem.arrachart@cea.fr

1
ACCEPTED MANUSCRIPT
Synthesis of organophosphorus ligands with a central oxygen atom and their
applications in solvent extraction
Fanny Marya, Guilhem Arracharta,*, Antoine Leydiera, Stéphane Pellet-Rostainga, *
a ICSM, CEA, CNRS, ENSCM, Univ Montpellier, Marcoule, France.
Email: guilhem.arrachart@cea.fr
Abstract
Various structurally related glycolamide compounds were synthesised and evaluated by solvent
extraction experiments for the separation of rare earth elements (REEs) from aqueous acid solutions
using dodecane as the diluent. We describe herein the full synthesis of a family of bifunctional
ligands with a central oxygen and focus our investigation on the effects of structural modification on
the extraction efficiencies. The combination of an amide, a P=O donor site, and a central oxygen in
such a glycolamide ligand showed interesting extraction properties for heavy rare earth elements
from phosphoric acid solutions.
Keywords
Organophosphorus ligands; REEs; Solvent extraction; Phosphoric acid.
1. Introduction
Rare earth elements (REEs) have an essential role in high-value technologies, many of which are in
the green energy sector, including magnets, optics, electronics, and ceramics, among others.[1,2]
With increasing demand for REEs in the international marketplace, the separation and purification of
these elements has gained considerable importance. Alternative methods that can be used to reduce
the stress on the rare earth supply include the use of substitutes, recycling, extracting tailings from
conventional mining, or the exploration of unconventional resources.[3]
The development of projects for the recovery of REEs from unconventional resources has gained
considerable importance. Phosphate rocks, which contain REEs as a minor component, are
considered a secondary source for REEs.[4] Because of the amount of phosphate rock processed
annually for the production of phosphoric acid, this resource is a suitable alternative source of REEs.
Indeed, the large volumes available offset the relatively low concentrations of rare earth elements
present in the ores and waste products created by the phosphoric acid industry. Also, rare earth
elements are often just by-products.5,6 Therefore, these issues have encouraged researchers to
innovate and propose efficient methodologies for recovering these metals from phosphate rocks.7
Several methods have been investigated to remove uranium from wet-process phosphoric acid
(WPA);[8,9] the possibility to recover rare earth elements and uranium from these phosphoric acid
leaching solutions has been performed in recent years based on solvent extraction or ion-exchange.10
Less attention has been given to the individual recovery of REEs from phosphoric acid. Different
techniques have been proposed to recover the rare earth elements from crude phosphoric acid.[11-15]
As a function of the phosphoric acid concentration, Reddy et al. showed that organophosphorus
solvents can manage the selectivity between light and heavy rare earth elements. Light rare earth
elements (LREEs) are preferentially recovered at low phosphoric acid concentrations, while higher

2
ACCEPTED MANUSCRIPT
phosphoric acid concentrations are more suitable for the extraction of heavy rare earth elements
(HREEs.)16,17
The development of procedures as well as new processes to increase the purity and quality of the
final product is encouraged. The design of new ligands can contribute to solving bottlenecks.
Within the context of nuclear waste management, different types of ligands have been developed for
the separation of lanthanides and actinides.[18]
Several research groups have investigated the extraction of lanthanides(III) and actinides with
diglycolamides, which have an ether oxygen between two amide groups, in nitric acid solutions.[19-21]
The extractability of Ln(III) by these ligands follows the sequence diglycolamide > diamide, suggesting
that the introduction of an ether oxygen is effective for the intragroup separation of Ln(III) cations.[22]
Recently, it has been reported that the extraction efficiency of f-block elements with two
diastereomers of dimethyl tetraoctyl diglycolamide (Me2-TODGA) is strongly affected by the
stereochemical orientation of the methyl group. 23
Organophosphorus compounds are also employed for the extraction of REEs from different media.[24]
For example, 2-ethylhexyl 2-ethylhexylphosphoric acid (PC-88),[25] di-(2-ethylhexyl)phosphoric acid
(HDEHP),[26] or cyanex compounds,[27] either alone or in a synergistic combination with another
organic ligand,[28,29] are well-known systems for the solvent extraction of lanthanides.
The development of multifunctional molecules has also been proposed, and studies on the solvent
extraction of uranium using multifunctional organophosphorus compounds from phosphoric acid
have reported that these compounds have high selectivity.[30,31]
Consequently, it would be interesting to develop such multifunctional molecules for the selective
extraction of REEs from phosphoric acid. Based on the abovementioned performances of
diglycolamide and organophosphorus compounds, our objective was to synthesise multifunctional
molecules that combine a structural backbone containing a combination of an amide and a P=O
donor group separated by a central oxygen or a combination of two P=O groups separated by a
central oxygen. Previously, these hybrid ligands were proposed by Iqbal et al. for the complexation of
Am(III)/Eu(III) from nitric acid media.[32]
In the present paper, different combinations are explored, and the synthesis and characterization of
seven multifunctional organophosphorus ligands is reported. We investigate the extraction
behaviour and separation of REEs in phosphoric acid, the dependence of the extraction distribution
ratio on the structure of the ligands, and the effects of the ligand and H3PO4 concentration.
2. Results and discussion
2.1.Synthesis
2.1.1. Synthesis of phosphonates and phosphine oxides with a central oxygen atom (I and II)
The first objective was to synthesise a phosphorous analogue with the well-known diglycolamide
TODGA (tetraoctyldiglycolamide) structure20 by replacing the two amide groups with two P=O donor
groups. The synthesis of multifunctional molecules which combine two P=O donor groups separated
by a central oxygen was guided by a disconnection approach through a retrosynthetic analysis which
proposes a chloromethyl- and a hydroxymethyl- phosphonate or -phosphine oxides as synthons.
For this purpose, various phosphoryl precursors have been synthesised (Scheme 1). The synthesis of
dioctylphosphine oxide 1a was adapted from a method found in the literature,33 where di-n-
butylphosphite was reacted with n-octylmagnesium bromide in diethylether. A similar procedure was

3
ACCEPTED MANUSCRIPT
used for the synthesis of octyl ethoxy phosphinate (1b). The reaction of compound 1a or 1b with
paraformaldehyde using sodium as a base in ethanol gave the corresponding hydroxymethyl-
phosphine oxide (2a or 2b) in high yield. A similar approach was used for the synthesis of various
hydroxymethyl-phosphonates (4) using triethylamine as a base instead of sodium leading to
phosphonates 4a and 4b respectively with a butyl (Bu) or an ethylhexyl (EtHex) group.
To investigate the effect of the steric hindrance of the methyl or octyl groups introduced on the
methylene bridge, we synthesised two molecules: compound 5a, which contains a methyl group
substituent, and 5b, which has an octyl group substituent; these were synthesised from dibutyl
phosphonate with methanal or octanal, respectively, following the procedure described for
compound 4.
The alcohol group of dioctyl (hydroxymethyl)-phosphine oxide (2a) was then chlorinated with PCl5,
yielding the desired dioctyl (chloromethyl)-phosphine oxide, 3.
Scheme 1. Synthesis of organophosphorus building blocks
Finally, both 2a and 4a were reacted with dioctyl chloromethyl-phosphine oxide (3) in the presence
of NaH to yield the desired phosphine oxides and phosphonates with central oxygen I and II ligands
in 82% and 36% yields, respectively (Scheme 2). The treatment in basic conditions can lead to the
monosaponification and/or saponification of the phosphonate.30,31 After purification, the mono-
phosphonate II was isolated.
Scheme 2. Synthesis of organophosphorus ligands with a central oxygen atom combining phosphine oxides or
phosphine oxide phosphonate functional groups.

4
ACCEPTED MANUSCRIPT
2.1.2. Synthesis of combined amides and phosphonates or phosphine oxides with a central
oxygen (III–IX)
Further modifications were investigated involving the replacement of only one of TODGA amide
group with P=O donors to give various phosphonate-O-acetamide and phosphinoxide-O-acetamide
ligands. This approach was proposed by Iqbal et al. using sodium hydride for the coupling of an
hydroxyl phosphorous derivative with an halogenated compound.[32] For this purpose, 2-chloro or 2-
bromo-N,N-dialkylacetamide derivatives (6) were synthesised, following the procedure illustrated in
Scheme 3.30,31
The extraction efficiency of the diglycolamides is related to the steric hindrance around the binding
site and from electro-inductive effects. The presence of branched alkyl groups or long, linear alkyl
groups on the amide of DGA act as electron donating groups, thus providing a more basic character.
Based on this and to improve the selectivity, solubility, extraction efficiency, and the behaviour of the
molecule during the extraction process, the substituent on the amide group has been chosen to be
an ethylhexyl (EtHex) or an octyl group (Oct).
Scheme 3. Synthesis of amido building blocks
Then, the hydroxymethyl phosphine oxide or hydroxymethyl phosphonate were combined with the
2-chloro or 2-bromo-N,N-dialkylacetamide (6) in the presence of a mixture of NaH and KI to give
seven ligands (Scheme 4), each containing one amide and one P=O donor group separated by a
central oxygen. Interestingly, in comparison to previous studies32 no dialkylphosphonate-O-
acetamides were obtained after the treatment. In fact, monosaponification occurred systematically
during the reaction, resulting in monoalkylphosphonate-O-acetamides, which are isolated after the
purification step.
It appears that the methyl or octyl groups introduced on the methylene bridge of the phosphorous
compounds (respectively 5a and 5b) significantly decreased the reactivity of the compound and
therefore the yield of the obtained corresponding products (VIII and IX).

5
ACCEPTED MANUSCRIPT
Scheme 4. Synthesis of organophosphorus ligands with a central oxygen atom combining amido, phosphine
oxide or phosphonate functional groups.
2.2.Solvent extraction results
In preliminary solvent extraction experiments, the extraction of a mixture of three heavy rare earth
elements (HREEs), such as gadolinium, dysprosium, and ytterbium, and a mixture of two light rare
earth elements (LREEs), such as lanthanum and neodymium, from phosphoric acid solutions was
investigated. Iron(III) was also studied as a competitive metallic cation. The aim of these experiments
was to investigate the extraction behaviour and separation possibilities for HREEs and LREEs and the
separation of HREEs as a group from LREEs in the aqueous acid phase. In addition, we hoped to
evaluate possible relationships between the structures of the ligands and their extraction efficiency.
The suitability of the ligand is associated with the distribution ratio, DM (DM = [M]org/[M]aq) and the
percentage extraction of metal E(%), where E(%) = ([M]org/[M]aq ini) × 100). The separation factor
SFM1/M2, which is the ratio of the distribution ratio DM1 and DM2 (SFM1/M2 = DM1/DM2) describes the
selectivity for one metal (M1) over another (M2). The parameters studied include the concentrations
of phosphoric acid and the ligand.
Figure 1 illustrates the extraction efficiency of REEs by ligands I and II. Ligand II differs from ligand I
by the presence of a phosphonic acid group instead of a phosphine oxide group; this has the effect of
improving the D values.

6
ACCEPTED MANUSCRIPT
The difference in performance can be attributed to the fact that in the case of ligand I, the functional
groups are only phosphine oxide, considered as neutral / solvating agents. Ligand II presents a
phosphonic acid group at the same time, which can act as a cationic exchanger, and a phosphine
oxide moiety that can be considered donor-solvating.
(a)
(b)
10
8
10
8
Fe
La
Nd
Gd
Dy
Yb
Fe
La
Nd
Gd
Dy
Yb
6
6
4
4
2
2
0
0
0.5M 1M 3M 5M
H3PO4
0.5M 1M 3M 5M
H3PO4
[
]
[
]
Figure 1. Distribution ratio for a series of five lanthanides (1 mmol/L) and iron(III) (50 mmol/L) at constant total
ligand concentration (10 mmol/L) in dodecane after contact with H PO (0.5 - 5 M): a) ligand I and b) ligand II.
3
4
The distribution ratio and extraction of REEs decreases with increasing phosphoric acid concentration
which is in accordance with the results obtained by Krea using the DOPPA.34 Therefore, relatively low
concentrations of phosphoric acid are required for efficient REE extraction. This behaviour suggests
that the mechanism of metal transfer to the organic phase follows an ion exchange type mechanism.
Further, the results suggest large differences in extraction efficiencies of some metals over others at
certain acidities, thereby showing separation possibilities. A selectivity / affinity towards Yb over light
rare earth elements and iron has been observed with SFYb/M > 10 (Figure SI-1 see supporting
information). However, these values have to be taken with caution, given that the distribution ratio
values are very low, and the associated uncertainty is high.
The extraction from 0.5 M H3PO4 is suitable for the selective separation of Yb from the other metals,
the extraction of Fe(III) was also observed but was low compared to that of Yb(III). These properties
indicate that such a ligand is effective for Yb extraction, despite the lower efficiency observed in
concentrated acidic solutions.
Surprisingly, the corresponding monoalkylphosphonate-O-acetamides differ from the above-studied
ligands. In particular, for ligand III, the extraction performance was high at all acidity values of the
aqueous phase (Figure 2). This behaviour has been observed for neutral extractants such as TODGA

7
ACCEPTED MANUSCRIPT
for the extraction of REEs from acidic media such as nitric or sulfuric acid.35,36 Additionally, it is
possible to carry out mutual extraction of HREEs from highly concentrated H
3
PO . The extractability
4
of Nd was quite low compared to Yb, Dy, and Gd; furthermore, La was not extracted, and Fe(III) was
poorly extracted, which resulted in high separation factors for HREEs over Fe at high phosphoric acid
concentrations (Figure SI-2). The mutual separation of REEs over Fe with a pronounced affinity for
HREEs such as Yb and Dy can be achieved when the concentration of the ligand is increased.
From the strong D
M
dependence on the H
3
PO concentration, it is assumed that the ligands act as
4
solvating ligands. To study the influence of ligand concentration on the REE distribution ratio,
experiments were performed at a constant acidity (5 M H3PO4) with several ligand III concentrations
from 10 to 50 mM in dodecane. The log D values are plotted as a function of the log of free
extractant concentration (Figure 2b). A linear relationship with a slope of ~1.5 obtained by regression
reflects that the extraction mechanism involves a mixture of 1:1 and 1:2 ratios of complex:REE. It is
also important to note that the linear nature of Figure 2b indicates that there is no change in the
extraction mechanism over the concentration range studied (from 10 to 50 mM).
The extraction efficiency of the ligands is related to the steric hindrance around the binding site.
Indeed, the presence of branched alkyl groups (ethylhexyl - EtHex for ligand VI) instead of linear alkyl
groups (octyl – Oct for ligand III) on the amide part led to a decrease in extraction performance (see
supporting information Figure SI-3).
(b)
(a)
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
80
60
40
20
0
Yb
Dy
Fe
La
Gd
Nd
Gd
Dy
Yb
Nd
La
-2.0
-1.8
-1.6
-1.4
0.5M 1M 3M 5M
[H PO
Log (LIII)
]
4 ꢀ
3
Figure 2. a) Distribution ratio for a series of five lanthanides (1 mmol/L) and iron(III) (50 mmol/L) at constant
total ligand III concentration (10 mmol/L) in dodecane after contact with H
3
PO (0.5 - 5 M); b) distribution ratio
4
versus ligand III concentration. log(D La) = f(log[ligand III]): y = 1.411(±0.057)x + 2.032(±0.097), R2 = 0.993; log(D
Nd) = f(log[ligand III]): y = 1.471(±0.061)x + 2.998(±0.103), R2 = 0.993; log(D Gd) = f(log[ligand III]): y =
1.521(±0.023)x + 3.950(±0.039), R2 = 0.999; log(D Dy) = f(log[ligand III]): y = 1.473(±0.045)x + 4.273(±0.076), R2
= 0.996; log(D Yb) = f(log[ligand III]): y = 1.509(±0.048)x + 5.336(±0.084), R2 = 0.997.

8
ACCEPTED MANUSCRIPT
In addition, the extraction of the REEs is affected by the branching of an alkyl group on the
methylene close to the central oxygen (Figure 3). Indeed, with the introduction of an alkyl group
between the phosphorous and the central oxygen in compounds VIII and IX, the extraction
performance decreased for most of the REEs, except for Yd. In comparison, when an alkyl group was
introduced between the amide and the central oxygen (compound VII), the extractability of REEs
decreased significantly. The efficiency is also affected by acidity. The mechanism of metal transfer to
the organic phase seems to follow an ion exchange type mechanism, as the efficiency of the
extraction decreased as the acid concentration increased. Generally, acid and neutral organo-
phosphorus extractants were used with formula of (RO)3PO or R3PO and the extraction efficiency of
acid organo-phosphorus extractants decreased with the RO branched chain.11
8
6
4
2
0
8
6
4
2
8
6
4
2
(a)
(c)
(b)
Fe
La
Nd
Gd
Dy
Yb
Fe
La
Nd
Gd
Dy
Yb
Fe
La
Nd
Gd
Dy
Yb
0.5M 1M 3M 5M
0.5M 1M 3M 5M
0.5M 1M 3M 5M
[
H PO
]ꢀ
[
H PO
]ꢀ
[H PO ]ꢀ
3
4
3
4
3
4
Figure 3. Distribution ratios for a series of five lanthanides (1 mmol/L) and iron(III) (50 mmol/L) at constant
total ligand concentration (10 mmol/L) in dodecane after contact with H
3
PO (0.5 - 5 M): a) ligand VII; b) ligand
4
VIII; c) ligand IX.
Therefore, from these results, it appears that the separation of individual/groups of REEs from others
and from iron(III) and the separation of HREEs from LREEs is made possible by using a specific
organophosphorus. Furthermore, the presence of a branching alkyl group on the methylene closest
to the central oxygen does not affect the separation efficiency but impacts the performance with
lower D values. The steric hindrance on the methylene has a strong impact on the extraction
efficiency.
Additionally, the extraction performance was affected by modifying the alkyl substituent of the
phosphonate moiety. Although the trend was similar for compounds III and IV, a decrease in
efficiency was observed when a branching group, such as ethylhexyl chain (ligand IV), was used
instead of a linear butyl chain (ligand III). Additionally, the amount of iron extracted is higher when
an ethylhexyl chain is introduced on the phosphonate group. This outcome was even more
pronounced when the phosphonate was changed for a phosphine oxide group (V) (see supporting
information Figure SI-4).

9
ACCEPTED MANUSCRIPT
In most of the systems studied, the order of the extractability is as follows: Yb > Dy > Gd > Nd >> La.
Yb(III) is therefore selectively extracted over the other REEs. The tendency of the prepared ligands to
preferentially extract Yb can be linked to lanthanide contraction. The observed differences in the
extraction efficiency are due to their ionic radii. The percentage extraction decreases with increasing
ionic radii. The results are in good agreement with reported literature.11 According to the hard-soft
acid-base (HSAB) theory, lanthanides are hard acids while the abovementioned O donor ligands are
hard bases. With increasing atomic number, the hardness of the lanthanides also increases, resulting
in an increase in extraction abilities. Cations with higher positive charge densities are strongly bound,
resulting in the heavier Ln(III)-ligand complexes being more stable than the lighter Ln(III) ones. In
contrast, the Ln(III) ions bind readily to water molecules and the energy required to dehydrate the
Ln(III) ions is enhanced with increasing atomic number/decreasing ionic radius.
The extraction efficiency for Yb is greater than that for LREEs, suggesting that, in all systems, the
coordination ability of the ligand to Ln(III) is greater than the dehydration energy of Ln(III). The
extractability of Fe(III) is quite low compared to Yb; this can also be connected to the hard acid
properties of Fe(III), which are lower than those of REE(III). In addition to these factors, we show that
the geometry of the coordination site in the ligand also has an effect on the extraction efficiency and
the selective separation of Yb (HREEs) and LREEs or separation of Yb (HREEs) from LREEs.
The literature review on solvent extraction and separation of rare earth elements shows that
many systems have been investigated for mineral acid media but not many from phosphoric acid
11,15,16
media.
Separation possibilities among REEs from phosphoric acid using organo-phosphorus
reagents has been reported using TOPS 99, an equivalent to di-2-ethylhexyl phosphoric acid. 37
The researchers show that the separation of a mixture of REEs was suitable from acid and extractant
effects.
Comparable performances were reported for different organo-phosphorus reagents under
reasonably similar conditions with the increase of extraction efficiency following ionic radii.
For example, Wang et al. showed that with 0.5 M of di-2-ethylhexyl phosphoric acid (HDHEP), less
than 20% of LREEs were extracted (D < 0.25) from a 10% P2O5 solution (2 M H3PO4) containing 1 g/L
of rare earth oxides while efficiencies higher than 70% were obtained for HREEs (D ≥ 2).11 The
extraction efficiencies can be managed by changing the HDEHP concentration. In a similar way, the
use of 0.7 M of TOPS 99 allowed preferential extraction of HREEs (extraction ≥ 90%; D ≥ 8) in
comparison to LREEs (40% <extraction > 75%; 0.7 < D > 3) from an 1 M H3PO4 with 25 ppm of
elements.37
Therefore, compared to the literature, this work obtained competitive extraction efficiencies. The
structure of the ligand can be adapted depending on the objective: selective or mutual extraction as
a function of the acidity of the medium. Considering efficiencies are related to acidity, it is possible to
consider the stripping of the REEs from the loaded organic phases by adjusting the acidity of the
stripping solution.
3. Conclusion
The synthesis of multifunctional molecules that combine two P=O donor groups or one P=O donor
group and one amide group separated by a central oxygen was performed. The extraction efficiency
of ligands towards REEs from H3PO4 medium was investigated. Depending on the target, the
structure of the ligand can be adapted. The possibility of the separation of one or groups of metals
from HREEs and LREEs has been highlighted. The mutual separation of REEs can be achieved with a
pronounced affinity for Yb when using monoalkylphosphonate-O-acetamide without alkyl groups on

10
ACCEPTED MANUSCRIPT
the methylene close to the central oxygen. However, the selective separation of Yb from the other
metals is effective with a mono-phosphonate-O-phosphine oxide ligand or with the introduction of
an alkyl group on the central methylene of the alkylphosphonate-O-acetamide.
4. Experimental
4.1. Synthetic chemistry
4.1.1. Chemicals and identification techniques.
Chemicals (analytically pure >99%) were purchased from Sigma Aldrich or Alfa Aesar and were used
without further purification. Anhydrous solvents were purchased from Acros (AcroSeal®). Reactions
were monitored by Thin Layer Chromatography (Merck TLC Silica Gel 60 F254) using different
revelatory systems: phosphomolybdic acid (25 g/500 mL ethanol), sulfuric acid (10% in ethanol), or
ninhydrin (1% in ethanol) following by heating and potassium permanganate (10 g K2CO3, 1.5 g
KMnO4, 150 mL H2O, 1.5 mL NaOH 10%). Flash chromatography purifications were performed on a
Combiflash Agilent Intelliflash 971-FP. NMR analyses were performed on Bruker 400 Ultrashield VS
1
spectrometer (400 MHz for H, 100 MHz for 13C, 162 MHz for 31P). Displacements are reported in
parts per million (ppm). Considering the presence of similar alkyl chains in the ligand structures,
some overlapping of carbon signals can be observed. The purity of the various compounds has been
determined thanks to the NMR analysis and be assumed to be ≥95%. HRMS were realized using a
Synapt G2-S (Waters) spectrometer.
4.1.2. Procedure and characterizations of compounds
4.1.2.1. Phosphoryl precursors
- Compound 1a : dioctylphosphine oxide. The synthesis was adapted from a literature method.33,38
- Compound 1b : ethyl-octylphosphinate.
To a mixture of octyl magnesium bromide (2 M, 20 mL) in dry Et2O was added dropwise triethyl
phosphite (10 ml, 57.5 mmol) under argon over 1 h at 0°C. Subsequently, HCL 1 M (10 mL) was added
until complete dissolution of the precipitated salts (addition of Et2O may be required). The resulting
mixture was stirred at room temperature and then heated at reflux for 4 h. After cooling to room
temperature, the mixture was extracted with Et2O (100 mL). The combined organic layers were
washed with HCl 1 M (100 mL), water (100 mL), brine (100 mL), then dried over MgSO4 and
evaporated to dryness. After flash chromatography on silica gel (Cyclohexane-ethyl acetate 50-50),
the ethyl-octylphosphinate 1b (72 %) was obtained as a colorless oil.

11
ACCEPTED MANUSCRIPT
31P NMR (CDCl3, 162 MHz) δ(ppm): 39
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 7.04 (d, JP-H = 728 Hz, 1H, P-H), 4.22 - 4.05 (m, 2H, O-CH2-
CH3), 1.81-1.73 (m, 2H, P-CH2), 1.65 – 1.54 (m, 2H, P-CH2-CH2), 1.42-1.24 (m, 13H, -CH2- & O-CH2-CH3),
0.89 (t, J = 6.8 Hz, 3H, CH2-CH2-CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 62.5 (P-O-CH2-CH3), 31.9, 30.7, 30.5, 29.4, 29.2, 28.5, 22.8,
20.9 (CH2), 16.5, 14.2 (CH3)
HRMS-ESI [M+H]+ Calc. for C10H24O2P : 205.1514; found: 205.1510
- Compound 2a : dioctyl-hydroxylmethyl-phosphine oxide. The synthesis was adapted from a
literature method.39
- Compound 2b : ethyl-(hydroxymethyl)(octyl)phosphinate.
The title compound (38%) was prepared following the same procedure used for the synthesis of
compound 2a.
31P NMR (CDCl3, 162 MHz) δ(ppm): 25.8
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 4.98 (s, 1H, -OH), 4.41-4.08 (m, 2H, CH2-O), 3.88-3.84 (m, 2H,
P-CH2-OH), 1.84-1.77 (m, 4H, CH2-P), 1.64-1.56 (m, 2H, CH2-CH2-P), 1.40-1.27 (m, 13H, -CH2- & O-CH2-
CH3), 0.88 (t, J = 6.8 Hz, 3H, CH2-CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 62.8 (P-O-CH2-CH3), 60.1 (d, JC-P= 75.4 Hz, P-CH2-OH), 31.9,
30.9, 30.5, 29.4, 29.2, 28.5, 22.8, 20.9 (CH2), 16.3, 14.1 (CH3)
HRMS-ESI [M+H]+ Calc. for C11H26O3P : 237.1620; found: 237.1608
- Compound 3 : dioctyl-chloromethyl-phosphine oxide. The synthesis was adapted from a literature
method.39
- Compound 4a : dibutyl-(hydroxymethyl)-phosphonate. The synthesis was adapted from a literature
method.40
- Compound 4b : diethylhexyl-(hydroxymethyl)-phosphonate
The title compound (78%) was prepared following the same procedure used for the synthesis of
compound 4a.
31P NMR (CDCl3, 162 MHz) δ(ppm): 24.2

12
ACCEPTED MANUSCRIPT
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 4.04-3.98 (m, 4H, CH2-O), 3.92 (d, 2H, P-CH2-OH), 1.60-1.54
(m, 2H, -CH-), 1.44-1.28 (m, 16H, CH2-), 0.94 – 0.83 (m, 12H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 67.7, 67.0 (P-O-CH2-), 60.0 (d, JC-P= 75.4 Hz, P-CH2-OH), 40.6
(P-O-CH2-CH), 31.0, 30.4, 28.5, 27.2, 22.9 (CH2), 14.3, 14.1, 10.5, 10.4 (CH3)
HRMS-ESI [M+H]+ Calc. for C17H38O4P : 337.2508; found: 337.2514
- Compound 5a : dibutyl-1-(hydroxyethyl)-phosphonate
The title compound (54%) was prepared following the same procedure used for the synthesis of
compound 4a starting from acetaldehyde instead of paraformaldehyde.
31P NMR (CDCl3, 162 MHz) δ(ppm): 25.8
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 4.87 (s, 1H, OH), 4.07-3.95 (m, 5H, CH & CH2-O), 1.62-1.54
(m, 4H, -CH2- CH2-O), 1.38-1.28 (m, 7H, -CH2- & CH-CH3), 0.86 (t, J = 6.8 Hz, 6H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 69.0 (d, JC-P= 77.3 Hz, P-CH(CH3)-OH), 65.8 (P-O-CH2-CH2),
60.1 (d, JC-P= 75.4 Hz, P-CH2-OH), 31.8, 20.6 (CH2), 13.9, 13.4 (CH3)
HRMS-ESI [M+H]+ Calc. for C10H24O4P : 239.1412; found: 239.1410
- Compound 5b : dibutyl-1-(hydroxynonyl)-phosphonate
The title compound (58%) was prepared following the same procedure used for the synthesis of
compound 5a starting from nonanal.
31P NMR (CDCl3, 162 MHz) δ(ppm): 25.4
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 4.13-4.04 (m, 4H, CH2-O), 3.85-3.80 (m, 1H, CH), 1.82-1.59
(m, 6H, -CH2-), 1.42-1.24 (m, 16H, -CH2-), 0.93-0.84 (m, 9H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 69.0 (d, JC-P= 77.3 Hz, P-CH(Oct)-OH), 65.8 (P-O-CH2-CH2),
60.1 (d, JC-P= 75.4 Hz, P-CH2-OH), 31.8, 31.5, 30.8, 29.7, 29.5, 29.3, 26.6, 22.5, 20.8 (CH2), 14.2, 13.8
(CH3)
HRMS-ESI [M+H]+ Calc. for C17H38O4P : 337.2508; found: 337.2510
4.1.2.2. Amide precursors
-
Compound 6a
:
2-chloro-N,N-dioctylacetamide, compound 6b
:
2-chloro-N,N-bis(2-
ethylhexyl)acetamide and compound 6c: 2-bromo-N,N-bis(2-octyl)propanamide were synthesized
from a literature method.30
4.1.2.3. Multifunctional organophosphorous ligands
General procedure for ligands synthesis

13
ACCEPTED MANUSCRIPT
To a solution of alcohol derivative (2 or 4) (5 mmol) in THF (10 mL) was added NaH (60% in oil) (0.8 g,
20.0 mmol) and KI (5 mmol) at 0°C. The mixture was stirred at 0°C for 1 h, then a solution of
halogenated compound (3 or 6) (6 mmol) in THF (10 mL) was added dropwise at this temperature.
The resulting mixture was stirred at room temperature at least for 6 hour. The resulting mixture was
then carefully acidified at 0°C with 3M HCl or NH4Cl saturated aqueous solution. Once the
effervescence had stopped, the solvent was evaporated. The residue was dissolved in ethyl acetate
washed twice with NH4Cl aqueous solution and water. The organic layer was dried over sodium
sulfate, filtered and concentrated under reduced pressure and the residue was purified by flash
chromatography on silica gel (Ethyl acetate-methanol 100-0 / 80-20)
- Compound I : (Oxybis(methylene))bis(di-octylphosphine oxide)
The title compound (82%) was prepared following the general procedure starting from 2a and 3. The
product was recrystallized from ethyl acetate to give the compound I as a white powder (m.p :
48±0.5°C).
31P NMR (CDCl3, 162 MHz) δ (ppm): 46.1
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 3.90 (m, 4H, CH2-O), 1.78-1.57 (m, 16H, -CH2-), 1.44-1.18 (m,
40H, -CH2-), 0.98-0.84 (m, 12H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 69.8 (P-CH2-O), 32.9, 32.0, 31.4, 31.3, 29.4, 22.8, 22.0 (CH2),
14.3 (CH3)
HRMS-ESI [M] Calc. for C34H72O3P2 : 590.4957; found: 590.4955
- Compound II : Butyl-((di-octanoylphosphoryl)-2-oxoethoxy)methyl)-phosphonate
The title compound (36%) was prepared following the general procedure starting from 4a and 3. The
product was purified by flash chromatography on silica gel (Ethyl acetate-methanol 100-0 / 80-20) to
give the compound II as an oil.
31P NMR (CDCl3, 162 MHz) δ (ppm): 51.9, 19.8
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 10.9 (s, 1H, OH), 4.35 (s, 2H, CO-CH2-O), 4.15-4.1 (m, 2H, O-
CH2-), 3.95 (m, 4H, O-CH2-P), 3.32 (t, J = 7.5 Hz, 2H, CH2-N), 3.13 (t, J = 7.5 Hz, 2H, CH2-N), 1.72-1.63
(m, 2H, -CH2-), 1.61-1.49 (m, 4H, -CH2-), 1.49-1.37 (m, 2H, -CH2-), 1.37-1.20 (m, 20H, -CH2-), 0.94-0.86
(m, 9H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 69.9, 69.7 (P-CH2-O), 67.3 (P-O-CH2-CH2), 32.9, 32.0, 31.4,
31.3, 29.4, 22.8, 22.0, 19.1 (CH2), 14.3, 14.0 (CH3)

14
ACCEPTED MANUSCRIPT
HRMS-ESI [M+H]+ Calc. for C22H49O5P2 : 455.3055; found: 455.3055
- Compound III : Butyl-((N,N-dioctylcarbamoyl)-2-oxoethoxy)methyl)-phosphonate
The title compound (72%) was prepared following the general procedure starting from 4a and 6a.
The product was purified by flash chromatography (SiO2, ethyl acetate – methanol (methanol 0 -20%)
to give the compound III as an oil.
31P NMR (CDCl3, 162 MHz) δ(ppm): 21.5
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 10.9 (s, 1H, OH), 4.35 (s, 2H, CO-CH2-O), 4.15-4.1 (m, 2H, O-
CH2-), 3.95 (m, 4H, O-CH2-P), 3.32 (t, J = 7.5 Hz, 2H, CH2-N), 3.13 (t, J = 7.5 Hz, 2H, CH2-N), 1.72-1.63
(m, 2H, -CH2-), 1.61-1.49 (m, 4H, -CH2-), 1.49-1.37 (m, 2H, -CH2-), 1.37-1.20 (m, 20H, -CH2-), 0.94-0.86
(m, 9H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 166.1 (C=O), 69.8 (P-CH2-O), 67.2 (P-O-CH2-CH2), 65.6 (C(O)-
CH2-O), 48.3, 46.3 (CH2-N), 31.8, 31.7, 29.4, 29.3, 29.2, 27.4, 26.8, 26.6, 22.9, 22.4, 22.0, (CH2), 14.3,
14.1, 13.8 (CH3)
HRMS-ESI [M+H]+ Calc. for C23H49NO5P : 450.3348 ; found: 450.3348
- Compound IV : Ethylhexyl-((N,N-dioctylcarbamoyl)-2-oxoethoxy)methyl)-phosphonate
The title compound (40%) was prepared following the general procedure starting from 4b and 6a.
The product was purified by flash chromatography (SiO2, ethyl acetate – methanol (methanol 0 -20%)
to give the compound IV as an oil.
31P NMR (CDCl3, 162 MHz) δ(ppm): 21.2
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 10.4 (s, 1H, OH), 4.37 (s, 2H, CO-CH2-O), 4.20-4.04 (m, 2H, O-
CH2-), 3.96 (m, 2H, O-CH2-P), 3.32 (m, 2H, CH2-N), 3.13 (m, 2H, CH2-N), 1.77-1.49 (m, 5H, -CH2-CH2-N
& -CH-), 1.48-1.12 (m, 28H, -CH2-), 0.98-0.80 (m, 12H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 166.0 (C=O), 69.7 (P-CH2-O), 67.2 (P-O-CH2-), 65.7 (C(O)-CH2-
O), 48.4, 46.2 (CH2-N), 36.8, 38.5 (CH), 31.8, 31.6, 29.4, 29.3, 29.2, 29.1, 27.4, 26.9, 26.8, 22.8, 22.6,
22.0 (CH2), 14.6, 14.3, 14.1, 10.9, 10.5, 10.4 (CH3)
HRMS-ESI [M+H]+ Calc. for C27H57NO5P : 506.3974; found: 506.3974
- Compound V : Octyl-((N,N-dioctylcarbamoyl)-2-oxoethoxy)methyl)-phosphinate

15
ACCEPTED MANUSCRIPT
The title compound (48%) was prepared following the general procedure starting from 2b and 6a.
The product was purified by flash chromatography (SiO2, ethyl acetate – methanol (methanol 0 -20%)
to give the compound V as an oil.
31P NMR (CDCl3, 162 MHz) δ(ppm): 48.6
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 4.35 (s, 2H, CO-CH2-O), 3.93 (m, 2H, O-CH2-P), 3.29 (m, 2H,
CH2-N), 3.20 (t, 2H, CH2-N), 1.85-1.70 (m, 2H, P-CH2-), 1.66-1.50 (m, 2H, P-CH2- CH2), 1.61-1.49 (m, 4H,
-CH2-CH2-N),1.44-1.18 (m, 30H, -CH2-), 0.93-0.84 (m, 9H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 166.1 (C=O), 69.8 (P-CH2-O), 65.8 (C(O)-CH2-O),48.3, 46.2
(CH2-N), 31.8, 31.6, 29.4, 29.3, 29.2, 29.1, 27.4, 26.9, 26.8, 22.8, 22.6, 22.0 (CH2), 14.3, 14.1, 13.9(CH3)
HRMS-ESI [M+H]+ Calc. for C27H57NO4P : 490.4025; found: 490.4024
- Compound VI : Butyl-((N,N-diethylhexylcarbamoyl)-2-oxoethoxy)methyl)-phosphonate
The title compound (42%) was prepared following the general procedure starting from 4a and 6b.
The product was purified by flash chromatography (SiO2, ethyl acetate – methanol (methanol 0 -20%)
to give the compound VI as an oil.
31P NMR (CDCl3, 162 MHz) δ(ppm): 20.6
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 7.96 (s, 1H, OH), 4.37 (s, 2H, CO-CH2-O), 4.17-4.07 (m, 2H, O-
CH2-), 3.92 (m, 4H, O-CH2-P), 3.40-3.20 (m, 2H, CH2-N), 3.06-3.00 (m, 2H, CH2-N), 1.72-1.56 (m, 4H, O-
CH2-CH2- & -CH-), 1.36-1.18 (m, 18H, -CH2-), 0.96-0.86 (m, 15H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 166.7 (C=O), 69.7 (P-CH2-O), 67.3 (P-O-CH2-CH2), 65.6 (C(O)-
CH2-O), 51.4, 48.6 (CH2-N), 36.8, 38.5 (CH), 30.6, 30.4, 28.9, 28.7, 24.0, 23.9, 23.1 (CH2), 14.3, 14.1,
10.9, 10.5, 10.4 (CH3)
HRMS-ESI [M+H]+ Calc. for C23H49NO5P : 450.3348 ; found: 450.3330
- Compound VII : Butyl-((N,N-diethylhexylcarbamoyl)-1-oxopropan-2-yl)methyl)-phosphonate
The title compound (56%) was prepared following the general procedure starting from 4a and 6c.
The product was purified by flash chromatography (SiO2, ethyl acetate – methanol (methanol 0 -20%)
to give the compound VII as an oil.

16
ACCEPTED MANUSCRIPT
31P NMR (CDCl3, 162 MHz) δ(ppm): 21.8
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 10.6 (s, 1H, OH), 4.47-4.20 (m, 1H, -CH-), 4.16-4.08 (m, 2H,
O-CH2-), 3.93 (m, 4H, O-CH2-P), 3.49-3.12 (m, 4H, CH2-N), 1.73-1.62 (m, 2H, O-CH2-CH2-), 1.61-1.48
(m, 4H, CH2-CH2-N), 1.47-1.37 (m, 5H, O-CH2-CH2-CH2- & -CH-CH3),1.34-1.24 (m, 20H, -CH2-), 0.96-0.82
(m, 9H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 166.3 (C=O), 78.7 (C(O)-CH-O), 69.7 (P-CH2-O), 67.3 (P-O-
CH2-CH2), 48.3, 46.3 (CH2-N), 31.8, 31.7, 29.4, 29.3, 29.2, 27.4, 26.8, 26.6, 22.9, 22.4, 22.0, (CH2), 16.8,
14.3, 14.1, 13.8 (CH3)
HRMS-ESI [M+H]+ Calc. for C24H51NO5P : 464.3505 ; found: 464.3503
- Compound VIII : Butyl-((N,N-dioctylcarbamoyl)-2-oxoethoxy)ethyl)-phosphonate
The title compound (10%) was prepared following the general procedure starting from 5a and 6a.
The product was purified by flash chromatography (SiO2, ethyl acetate – methanol (methanol 0 -20%)
to give the compound VIII as an oil.
31P NMR (CDCl3, 162 MHz) δ(ppm): 22.3
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 4.48-4.19 (m, 2H, CO-CH2-O), 4.21-4.01 (m, 2H, O-CH2-),
3.97-3.88 (m, 1H, O-CH-P), 3.30 (m, 2H, CH2-N), 3.13 (t, 2H, CH2-N), 1.73-1.61 (m, 2H, O-CH2-CH2-),
1.60-1.36 (m, 9H, -CH2-CH2-N-CH2-CH2- ; O-CH2-CH2-CH2- & -CH-CH3),1.37-1.21 (m, 20H, -CH2-), 0.93-
0.85 (m, 9H, CH3)
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 166.7 (C=O), 81.3 (P-CH-O), 69.7 (P-CH2-O), 65.6 (C(O)-CH2-
O), 48.3, 46.3 (CH2-N), 31.8, 31.7, 29.4, 29.3, 29.2, 27.4, 26.8, 26.6, 22.9, 22.4, 22.0, (CH2), 14.2, 14.1,
13.8, 10.6 (CH3)
HRMS-ESI [M+H]+ Calc. for C24H51NO5P : 464.3505 ; found: 464.3507
- Compound IX : Butyl-((N,N-dioctylcarbamoyl)-2-oxoethoxy)nonyl)-phosphonate
The title compound (22%) was prepared following the general procedure starting from 5b and 6a.
The product was purified by flash chromatography (SiO2, ethyl acetate – methanol (methanol 0 -20%)
to give the compound IX as an oil.
31P NMR (CDCl3, 162 MHz) δ(ppm): 22.1
1H NMR (CDCl3, 400 MHz, 25°C) δ(ppm): 9,85 (s, 1H, OH), 4.34-4.19 (m, 2H, CO-CH2-O), 4.16-4.0 (m,
2H, O-CH2-), 3.68-3.64 (m, 1H, O-CH-P), 3.29 (m, 2H, CH2-N), 3.21 (m, 2H, CH2-N), 1.90-1.84 (m, 2H, -
CH2-CH-P), 1.72-1.65 (m, 2H, O-CH2-CH2-), 1.59-1.47 (m, 4H, -CH2-CH2-N-CH2-CH2-), 1.47-1.35 (m, 2H,
O-CH2-CH2-CH2-), 1.34-1.18 (m, 32H, -CH2-), 0.93-0.81(m, 12H, CH3)

17
ACCEPTED MANUSCRIPT
13C NMR (CDCl3, 100 MHz, 25°C) δ(ppm): 166.7 (C=O), 81.3 (P-CH-O), 69.7 (P-CH2-O), 65.6 (C(O)-CH2-
O), 48.3, 46.3 (CH2-N), 31.8, 31.7, 29.4, 29.3, 29.2, 27.4, 26.8, 26.6, 24.7, 22.9, 22,6 22.4, 22.0,
21.6 (CH2), 14.2, 14.1, 13.9, 13.4 (CH3)
HRMS-ESI [M+H]+ Calc. for C31H65NO5P : 562.4600 ; found: 562.4604
4.2. Extraction experiments from H3PO4 solutions
To illustrate the entire rare earth elements series two light rare earth elements (LRREs): La and Nd
and three heavy rare earth elements (HREEs): Gd, Dy and Yb, have been chosen for the extraction
experiments. Rare earth elements stock solutions were prepared at the desired acidity from
lanthanum(III) nitrate hexahydrate (Sigma Aldrich, 99.99%), neodymium(III) nitrate hexahydrate (Alfa
Aesar, 99.99%), gadolinium(III) nitrate hexahydrate (Alfa Aesar, 99.9%), dysprosium(III) nitrate
hexahydrate (Alfa Aesar, 99.9%), ytterbium(III) nitrate hexahydrate (Alfa Aesar, 99.9%).
The desired concentrations (typically 5 mmol/L of Ln(NO3)3 in H3PO4, equimolar concentrations of
metals = 1 mmol/L) were prepared by dilution using ultrapure water (MilliQ, Millipore, 18 MΩ.cm−1)
and the acidity (typically H3PO4 0.5 to 5 M) was adjusted with phosphoric acid. The initial
concentrations of metals were measured by inductively coupled plasma/atomic emission
spectroscopy (ICP/AES Spectro Arcos).
The organic phase was prepared by dissolving ligands (from 5 to 100 mmol/L) in dodecane.
Organic phases were pre-equilibrated with an aqueous phase at the same acidity as the
extraction step without REE cations. The pre-equilibrated organic phases were then mixed with an
equal volume of an aqueous acidic stock solution of REEs in a thermostated shaker (Infor-ht®
Ecotron) at 25°C and 400 rpm for 1 h, (this time is sufficient to reach thermodynamic equilibrium).
Phases were separated after centrifugation at 4000 rpm for 30 minutes (sigma 3-16 PK). After
centrifugation, the separated aqueous phases were subjected to ICP-AES analysis in order to
determine i) the distribution ratio (DM), which is the ratio of the total concentration of metal M in the
organic phase to the total concentration of the same metal in the aqueous phase, i.e., DM
=
[Morg]/[Maq] and ii) the extraction efficiency E%=DM/(DM+(A/O)) with A/O the aqueous to organic
phase volumetric ratio (typically =1).
The experiments were carried out in duplicate measurements with precision of ± 5%.
Associated Content
Supporting Information: additional data on extraction behaviour and characterization of compounds
(Figure SI-1 to Figure SI-4,Table SI-1 and Table SI-2, NMR spectra).
Acknowledgments
The authors would like to acknowledge Béatrice BAUS-LAGARDE for her assistance with ICP-AES.

18
ACCEPTED MANUSCRIPT
References and notes
1.
2.
Curtis, N. Rare earths : We can touch them everyday. ; Lynas Corporation: Sydney, 2011.
Binnemans, K.; Jones, P. T.; Blanpain, B.; Gerven, T. V.; Yang, Y.; A., W.; Buchert, M. Journal of
Cleaner Production 2013, 51, 1-22.
3.
4.
5.
Ebner, J. Europe’s Rare Earth Dependence on China: Future Perspectives, 2014.
Koopman, C.; Witkamp, G. J. Hydrometallurgy 2000, 58, 51-60.
Habashi, F. Journal of Chemical Technology and Biotechnology. Chemical Technology 1985,
35, 5-14.
6.
Cánovas, C. R.; Pérez-López, R.; Macías, F.; Chapron, S.; Nieto, J. M.; Pellet-Rostaing, S.
Journal of Cleaner Production 2017, 143, 497-505.
7.
8.
9.
Reddy, B. R.; Kumar, J. R. Solvent Extraction and Ion Exchange 2016, 34, 226-240.
Hurst, F.; Brown, K. B.; Crouse, D. J. Ind. Eng. Chem. Process Des. Dev. 1972, 11, 122-128.
Beltrami, D.; Cote, G.; Mokhtari, H.; Courtaud, B.; Moyer, B. A.; A., C. Chem. Rev. 2014, 114,
12002−12023.
10.
11.
12.
13.
Bunus, F. T. Mineral Processing and Extractive Metallurgy Review 2000, 21, 381-478.
Wang, L.; Long, Z.; Huang, X.; Yu, Y.; Cui, D.; Zhang, G. Hydrometallurgy 2010, 101, 41-47.
Preston, J. S.; Cole, P. M.; Craig, W. M.; Feather, A. M. Hydrometallurgy 1996, 41, 1-19.
Lokshin, E. P.; Tareeva, O. A.; Elizarova, I. R. Russian Journal of Applied Chemistry 2013, 86,
623-628.
14.
15.
16.
17.
Koopman, C.; Witkamp, G. J.; Van Rosmalen, G. M. Separation Science and Technology 1999,
34, 2997-3008.
Rey, J.; Atak, S.; Dourdain, S.; Arrachart, G.; Berthon, L.; Pellet-Rostaing, S. Solvent Extraction
and Ion Exchange 2017, 35, 321-331.
Radhika, S.; Kumar, B. N.; Kantam, M. L.; Reddy, B. R. Separation and Purification Technology
2010, 75, 295-302.
Reddy, B. R.; Radhika, S.; Kumar, B. N. Separation Science and Technology 2010, 45, 1426-
1432.
18.
19.
Dam, H. H.; Verboom, W.; Reinhoudt, D. N. Chem. Soc. Rev. 2007, 36, 367-377.
Sasaki, Y.; Sugo, Y.; Suzuki, S.; Tachimori, S. Solvent extraction and ion exchange 2001, 19, 91-
103.
20.
Ansari, S. A.; Pathak, P.; Mohapatra, P. K.; Manchanda, V. K. Chemical Reviews 2012, 112,
1751-1772.
21.
22.
23.
Leoncini, A.; Huskens, J.; Verboom, W. Chemical Society Reviews 2017, 46, 7229-7273.
Narita, H.; Yaita, T.; Tachimori, S. Solvent extraction and ion exchange 2004, 22, 135-145.
Wilden, A.; Kowalski, P. M.; Klaß, L.; Kraus, B.; Kreft, F.; Modolo, G.; Li, Y.; Rothe, J.;
Dardenne, K.; Geist, A.; Leoncini, A.; Huskens, J.; Verboom, W. Chemistry – A European
Journal 2019, 25, 5507-5513.
24.
25.
26.
27.
28.
29.
30.
Xu , L.; Xiao, Y.; Li, D. Q. J. Chem. Inf. Comput. Sci. 1992, 437-442.
Abdeltawab, A. A.; Nii, S.; Kawaizumi, F.; Takahashi, K. Sep. Purif. Technol. 2002, 26, 265-272.
Rabie, A. Hydrometallurgy 2007, 85, 81-86.
Li, W.; Wang, X.; Hui, Z.; Meng, S.; Li, D. Q. J. Chem. Technol. Biotechnol. 2007, 82, 376-381.
Wu, D. B.; Wei Li; Li, D. Q.; Yingxiong. Sep. Sci. Technol. 2007, 42, 847-864.
Huang, X.; Li, J.; Long, Z.; Zang, Y.; Xue, X.; Zhu, Z. J. Rare Earths 2008, 26, 410-413.
Turgis, R.; Leydier, A.; Arrachart, G.; Burdet, F.; Dourdain, S.; Bernier, G.; Miguirditchian, M.;
Pellet-Rostaing, S. Solvent Extr. Ion Exch. 2014, 32, 478-491.
31.
32.
33.
34.
Turgis, R.; Leydier, A.; Arrachart, G.; Burdet, F.; Dourdain, S.; Bernier, G.; Miguirditchian, M.;
Pellet-Rostaing, S. Solvent Extr. Ion Exch. 2014, 32, 685-702.
Iqbal, M.; Struijk, R. G.; Huskens, J.; Sypula, M.; Wilden, A.; Modolo, G.; Verboom, W. New J.
Chem. 2012, 36, 2048-2059.
Wang, F.; Tang, R.; Kao, J. L. F.; Dingman, S. D.; Buhro, W. E. Journal of the American Chemical
Society 2009, 131, 4983-4994.
Krea, M.; Khalaf, H. Hydrometallurgy 2000, 58, 215-225.

19
ACCEPTED MANUSCRIPT
35.
36.
37.
38.
39.
40.
Gergoric, M.; Ekberg, C.; Foreman, M. R. S. J.; Steenari, B.-M.; Retegan, T. Journal of
Sustainable Metallurgy 2017, 3, 638-645.
Yuan, H.; Hong, W.; Zhou, Y.; Pu, B.; Gong, A.; Xu, T.; Yang, Q.; Li, F.; Qiu, L.; Zhang, W.; Liu, Y.
Journal of Rare Earths 2018, 36, 642-647.
Radhika, S.; Nagaphani Kumar, B.; Lakshmi Kantam, M.; Ramachandra Reddy, B.
Hydrometallurgy 2011, 110, 50-55.
Williams, R. H.; Hamilton, L. A. Journal of the American Chemical Society 1952, 74, 5418-
5420.
Leydier, A.; Arrachart, G.; Turgis, R.; Bernier, G.; Marie, C.; Miguirditchian, M.; Pellet-
Rostaing, S. Hydrometallurgy 2017, 171 262–266.
Jeanmaire, T.; Hervaud, Y.; Boutevin, B. Phosphorus Sulfur and Silicon and the Related
Elements 2002, 177, 1137-1145.