Study on Phenanthroline Carboxamide for Lanthanide Separation: Influence of Amide Substituents

Article abstract of DOI:10.1021/acs.inorgchem.1c01729

Phenanthroline carboxamide compounds are promising for lanthanide intra-series separation. This paper presents a study on the effect of structure modification of phenanthroline carboxamides on the extraction of the whole lanthanide series. The study consists of theoretical calculations, extraction experiments of the 14 stable lanthanides, and extended X-ray absorption fine structure (EXAFS) analyses of Nd and Dy complexes. Tridentate monocarboxamides and tetradentate dicarboxamides show different trends in series extraction, although both preferentially extract the light lanthanides. The amide substituents, although not directly coordinating the metal ions, were also found to impact the distribution ratio, most probably due to a modification in the internal polarity of the molecules. This latter effect, if extrapolated to other nitrogen-based ligands such as pyridines or triazines, can be used to further fine-tune extractants for a process improvement.

Full text of DOI:10.1021/acs.inorgchem.1c01729

pubs.acs.org/IC
Article
Study on Phenanthroline Carboxamide for Lanthanide Separation:
Influence of Amide Substituents
Marie Simonnet,* Tohru Kobayashi, Kojiro Shimojo, Keiichi Yokoyama, and Tsuyoshi Yaita
Cite This: https://doi.org/10.1021/acs.inorgchem.1c01729
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Phenanthroline carboxamide compounds are prom-
ising for lanthanide intra-series separation. This paper presents a
study on the effect of structure modification of phenanthroline
carboxamides on the extraction of the whole lanthanide series. The
study consists of theoretical calculations, extraction experiments of
the 14 stable lanthanides, and extended X-ray absorption fine
structure (EXAFS) analyses of Nd and Dy complexes. Tridentate
monocarboxamides and tetradentate dicarboxamides show differ-
ent trends in series extraction, although both preferentially extract
the light lanthanides. The amide substituents, although not directly
coordinating the metal ions, were also found to impact the
distribution ratio, most probably due to a modification in the
internal polarity of the molecules. This latter effect, if extrapolated to other nitrogen-based ligands such as pyridines or triazines, can
be used to further fine-tune extractants for a process improvement.
Table 1. Examples of Phenanthroline Carboxamides Already
Studied in the Literature: R1, R2, and R3 as Described in
Figure 1
INTRODUCTION
■
Because of their use in electronics device, the demand for rare-
earth elements keeps increasing. They are usually recovered
altogether from mining.1 However, their similar chemical
properties make them difficult to separate from each other.
Currently used processes need many stages to get the required
metal purity due to the poor separation factors of the acid
phosphorus extractants.2,3 Research on alternative extractants
is thus necessary to improve these processes.
name
Et-Ph-DAPhen
R1/R2
R3
refs
ethyl/phenyl
CA
4
Et-Tol-DAPhen
Et(pEtPh)Phen
ethyl/tolyl
ethyl/p-ethylphenyl
CA 5, 6
CA 7, 8
N,N′-Et-N,N′-p-hexylphenyl-4,7- ethyl/p-hexylphenyl
CA
4
9
dibutoxy-DAPhen
butoxy onto the
phenanthroline
Thanks to extensive research on the separation of actinides
from lanthanides for nuclear reprocessing, much data on
lanthanide extraction can already be found in the literature. In
this study, we chose to focus on phenanthroline carboxamide-
based ligands because of their good extraction behavior as
reported in the literature. Table 1 lists some phenanthroline
carboxamides that have already been studied.4−16 The
dicarboxamide ligands have mainly been studied for the
separation of actinides from the lanthanides and not as a direct
application for adjacent lanthanides separation, although some
data on some or the whole Ln series are often provided. In all
of these studies, either a rather flat trend all along the series for
N-dialkyl ligands or a preference towards light Ln for N-alkyl,
N-aryl was obtained, and separation factors between minor
actinides and lanthanides could reach up to 2 orders of
magnitude.
TBuPhen
butyl/butyl
CA 7,
10−12
CA 13
14−16
TDdPTdA
OTPTA
dodecyl/dodecyl
octyl/tolyl
H
line monocarboxamide ligands and PTdA refers to phenan-
throline dicarboxamide ligands.
To increase the hydrophobicity of the ligand, the alkyl chain
for extraction experiment has been set to eight carbons (octyl,
named “O”); however, to lighten the calculations, the chain
was shortened to methyl (“M”). Previous study showed little
Received: June 8, 2021
In this study, both mono and dicarboxamide ligands have
been studied and compared. The different amide substituents
that were used in this study are described in Figure 1 and
Table 2. In the whole manuscript, PTA refers to phenanthro-
https://doi.org/10.1021/acs.inorgchem.1c01729
© XXXX American Chemical Society
Inorg. Chem. XXXX, XXX, XXX−XXX
A

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 1. Phenanthroline carboxamides used in this study; CA is the carboxamide and is present symmetrically in PTdA.
Ln3+ + 3NO−3 + mL ↔ LnLm(NO3)3
Table 2. List of Extractants Synthesized and/or Calculated
in This Study
where m represents the number of ligands coordinated to the
lanthanide ion. Considering that PTA, respectively PTdA, is
tridentate, respectively tetradentate, the possible stoichiometries are
m = [1,2,3], with nitrate ions involved in the missing coordinating
bonds. Later in the text, stoichiometries will be described as
lanthanide:ligand in the absence of nitrate ions and lanthanide:li-
gand:nitrates otherwise.
extraction/calculation
R1
R2
R3
DOPTA/DMPTA
OPPTA/MPPTA
OTPTA/MTPTA
EHTPTA
OTPTdA/MTPTdA
OdMAPTA/MdMAPTA
OdMAPTdA/MdMAPTdA
NA/MtMAPTA
octyl/methyl octyl/methyl
octyl/methyl phenyl
octyl/methyl p-Tolyl
2-ethyl-hexyl p-Tolyl
octyl/methyl p-Tolyl
octyl/methyl 3,5-dimethylphenyl
octyl/methyl 3,5-dimethylphenyl
NA/methyl
H
H
H
H
CA
H
CA
H
Extraction Experiments. The aqueous phase consisted of a
mixture of 7 ppm of all lanthanide nitrates in 0.1 M HNO3. NaNO3
was used to increase the concentration of nitrate ions. Due to the
poor solubility of PTA in aliphatic diluents, the extractants were
dissolved in chloroform to avoid any additional polarity effects arising
from the diluents.13 The extraction then consisted of mixing together
500 μL of each phase in a glass vial for 1 h at a thermostated
temperature of 25 °C, followed by 5 min of centrifugation. The back-
extraction was achieved using distilled water at an O/A ratio of 1:3.
We used inductively coupled plasma mass spectrometry (ICP-MS,
Agilent) for the sample analyses, with a calibration curve from 0.7 to
70 ppb. Because ICP-MS can only measure aqueous samples, 550 μL
of the aqueous phase was initially prepared, from which 50 μL was
diluted for ICP-MS analysis. The concentrations in the organic phase
were then obtained by the difference between the initial
concentrations and the concentrations in the aqueous phase after
extraction. In the few cases when the extraction was really low, this
method could however not be used due to higher uncertainties and
the organic concentration was considered equal to the back-extraction
concentration. In both cases, results were analyzed from the
lanthanide distribution ratios DLn calculated as the organic phase
concentration divided by the aqueous phase concentration:
3,4,5-
trimethylphenyl
effect of the chain length8,14 so their chemical behavior can be
considered identical.
EXPERIMENTAL SECTION
■
Synthesis. All ligands have been synthesized in lab from
phenanthroline mono- (respectively, di-) carboxylic acid purchased
from Intellim Corporation. The octyl-amines have been prepared by
mixing the proper aniline with bromooctane (TCI) at 70 °C for 20 h.
The yield of this step was lowered due to the formation of di-
octylamines. The product was then purified through column
chromatography and checked by NMR and electrospray ionization-
mass spectrometry (ESI-MS). The final product was then obtained by
mixing the synthesized aniline with the carboxylic acid in the presence
of DMT-MM and DIPEA (yields 70−85%, Figures 2), followed by
column chromatography purification and NMR and ESI-MS checking
and S1−S7, Tables S1 and S2).
[Ln]org
[Ln]aq
[Ln]init − [Ln]aq
[Ln]aq
DLn
=
=
Calculations. Geometric optimizations and frequency calculations
have been performed with Gaussian09 software on the JAEA
supercomputer HPE SGI8600. M06 method has been used along
with MWB ECPs for the lanthanide atoms and D95V** basis set for
CHON atoms. The aqueous lanthanide ions were considered as
Two samples were diluted, measured, and averaged for each
experiment to increase the reliability of the results. Mass balance was
systematically checked by comparing the initial concentration to the
sum of extracted and back-extracted aqueous phases, yielding a
difference equal to or lower than 5%. Experimental errors, e.g., initial
concentrations, pipetting, solvent vaporization, etc., are expected
around 5%, and ICP analysis errors, considered from both the
instrument RSD and the difference from the two replicates, led to
+
Ln(NO3)2 ·5H2O or Ln3+·9H2O species. No water molecules were
added onto the organic species. The ligand energies were calculated as
dimer to account for possible H-bonds.
Complex Formation. The formation of the complex can be
written as
Figure 2. PTA synthesis route.
B
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
errors between 0 and 10% depending on the samples, resulting in an
overall error of maximum 15%.
Extended X-ray Absorption Fine Structure (EXAFS) Experi-
ment. Aqueous solutions containing 50 mM (respectively 70 mM)
NdCl3 (respectively, DyCl3) were mixed with 0.1 M extractant in
chloroform. The organic phase was then withdrawn and irradiated at
the BL22XU in Spring8. BL22XU is an insertion device beamline to
use high-brightness X-rays, equipped with a monochromator that
enables fast scan to obtain high S/N ratio absorption spectra, making
it possible to measure XAFS spectra of samples at low concentrations.
The incident radiation was monochromatized with the liquid-
nitrogen-cooled Si(311) crystal. The spectra were recorded during
quick scanning of monochromator and fluorescence mode. To
measure the X-ray intensities, ionization chambers of 17 cm filled
with nitrogen gas and 19-element high-purity germanium (HPGe)
detector (Mirion Technologies (Canberra) KK) were used for
incident and fluorescence X-rays, respectively. The measurement
was performed on Nd (respectively, Dy)-K-edge. The obtained
EXAFS spectra were analyzed according to a standard method using
the program WinXAS32. Theoretical phase shifts and backscattering
amplitude functions were calculated by the program code FEFF9
based on Gaussian geometry optimization. Due to the limited
beamline time, only Nd and Dy complexations with OTPTA,
OTPTdA, and OdMAPTdA spectra were obtained. This latter
spectrum was obtained in a mixture of chloroform and octanoic acid.
Figure 4. Optimized geometry of Nd(MPPTdA)23+ (green = Nd, blue
= N, red = O, gray = C, white = H).
RESULTS
■
Table 3. Mean Distances of the Closest Atoms around Nd in
PTA Ligands Having 0−3 Methyl Groups Attached to the
Aniline, Calculated in the Gas Phase with Two Bidentate
Nitrate Ions in the Inner Sphere
Computational Results. Density functional theory (DFT)
calculations have been performed on PTA and PTdA
compounds for geometry optimization, followed by frequency
calculations. As an example, Figures 3 and 4 show the
+
MPPTA
MTPTA
MdMAPTA
MtMAPTA
optimized geometries of Nd(MTPTA)2(NO3)2 and Nd-
(MPPTdA)23+, respectively.
Nd−O
Nd−N
Nd−O (NO3)
2.52
2.72
2.58
2.49
2.72
2.56
2.48
2.72
2.56
2.51
2.74
2.55
increase seems to be compensated by a shortening of the Nd-
−
NO3 distance compared to the other ligands.
Frequency calculations led to a possible comparison of the
free Gibbs energies of the PTA compounds. Results are
provided in Table 4. As already mentioned in ref 4, the
Table 4. Free Gibbs Energies ΔG (kJ/mol) Calculated as
the Difference between the Complex Nd3+-L2 (Respectively,
+
Nd-L2-(NO3)2 ) Energy and Hydrated Nd (Respectively,
Hydrated Nd(NO3)2), L2, and H2O Energies
kJ/mol
number of methyl
without nitrates
with nitrates
MtMAPTA
MdMAPTA
MTPTA
MPPTA
DMPTA
3
2
1
0
NA
NA
−63.38
−62.50
−54.88
−52.93
NA
Figure 3. Optimized geometry of Nd(MTPTA)2(NO3)+ (green =
Nd, blue = N, red = O, gray = C, white = H).
−81.43
−75.86
−74.77
−67.91
For the tridentate PTA ligands, both 1:2 and 1:3
stoichiometries have been considered. Both stoichiometries
were found stable in the gas phase. Upon the addition of
nitrate ions in the inner sphere, only the 1:2 complex has been
studied, as previous studies reported slope analyses of 2 only in
nitrate media14 and a crystal structure with the 1:2:2
stoichiometry,15,16 the third nitrate ion being in the outer
sphere. Therefore, only the 1:2 results are reported in Table 3,
showing the average distance of the closest atoms around Nd
for different PTA ligands with two nitrate ions. The distances
of the phenanthroline atoms to Nd are very similar, especially
Nd−N. Nd−O slightly shortens from phenyl to dimethyl-
phenyl but then increases again for trimethylphenyl. This latter
presence of an aromatic ring greatly improves the complex
stability, as proved by the increase in the absolute value of free
Gibbs energy. This observation is also in good agreement with
the so-called anomalous aryl effect,17 observed for several other
extractants,18,19 which shows higher extraction ability with aryl
substituents. Additionally, the free Gibbs energies gradually
decrease with the number of methyl substitutions onto the
phenyl ring, with a higher difference between one and two
methyl substitutions. The methyl being far from the
coordination site, this result is especially interesting because
it stresses the importance of the internal polarity of the ligand
C
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
toward a higher stability of the metal-ligand complex.6,20 It is
difficult to evaluate whether the lower energies found with
nitrates compared to without nitrates are due to the method
for calculating the aqueous Nd, which considers the Nd(NO3)2
hydrated species without the nitrate-ion hydration sphere, or a
real effect due to the insertion of additional atoms within the
first coordination shells, which somewhat destabilizes the
complex.
The PTdA ligands have been studied as 1:2 complexes, with
a potential addition of one nitrate ion, and 1:1 complexes with
an addition of three nitrate ions, as previously reported in the
literature for similar compounds.4,6 Convergence issues arose
for the frequency calculations of the biggest molecules, and the
free Gibbs energies could not be compared similarly to the
PTA ligands. For Nd-MTPTdA, the most stable stoichiometry
was the 1:2 complex without nitrate ions with an energy of
−285 kJ/mol, much higher than that of MTPTA, −76 kJ/mol.
The 1:2:1 complex was also stable (−188 kJ/mol) but not the
1:1:3 species (50 kJ/mol). This can be linked to the increased
Ln−O and Ln−N distances of the phenanthroline carbox-
amide upon nitrate addition (Table 5). However, as stated
Table 7. Difference between the Ln−O Distances of Ln in
Water for Their Most Stable CN and the Ln−O,N Distances
in Ln(MTPTA)2(NO3)2 and Ln(MTPTdA)2NO3 (CN =
a
10)
a
Ln-OH2
MTPTA
difference
MTPTdA
difference
La
Pr
2.585
2.540
2.520
2.490
2.460
2.430
2.380
2.679
2.614
2.611
2.584
2.561
2.535
2.501
0.094
0.074
0.091
0.094
0.101
0.105
0.121
2.679
2.647
2.643
2.604
2.583
2.571
2.532
0.094
0.107
0.123
0.114
0.123
0.141
0.152
Nd
Sm
Gd
Dy
Tm
a
Values taken from ref 21.
As predicted by the Gaussian calculation, there is an
important increase in distribution ratio in the presence of an
aromatic cycle onto the nitrogen atom of the amide and a
further increase with a higher number of methyl groups grafted
onto the phenyl ring (Figures 5−7).
Figure 5 shows the evolution of the distribution ratio along
the lanthanide series of four PTA ligands at two nitric acid
concentrations. Without any aromatic ring, low and rather flat
lanthanide distribution ratios are obtained at 0.1 M HNO3,
with a slight increase in light lanthanide with the increase in
nitric acid concentration. In the presence of an aromatic ring,
there is a maximum around Pr−Sm, which is in agreement with
Table 7. This maximum, as reported by Nakase et al.,16 can be
explained by ion-size effects: the cation lies in a cavity induced
by the N−N−O atoms of the phenanthroline, whose size is
better suited for the bigger light Lns, hence the preference of
all PTA and PTdA ligands toward the light Lns. To adapt to
the different ionic radii, the amide oxygens need to slightly
rotate, but this rotation is limited in the presence of the aryl
ring, making the cavity less suitable for the biggest cations La/
Ce. For all three extractants, increasing the nitric acid
concentration decreases the distribution ratio of light
lanthanides but has less influence on the heavy lanthanides.
The branched octyl compound, EHTPTA, has a lower
distribution ratio than the linear OTPTA, probably because
of additional steric effects. The trend along the lanthanide
series is however similar.
Figures 6 and 7 show the lanthanide distribution ratios of
several PTA and PTdA at 1 M NaNO3, with a variation in
ligand concentration for Figure 7. Again, very low distribution
ratios are obtained without any aromatic ring (Figure 6).
Interestingly, very different patterns are obtained with mono-
and dicarboxamide: PTA-nitrate systems show a maximum
within the light Ln (Pr−Sm) followed by a slight decrease all
along the series, whereas PTdA-nitrate systems show a sharp
decrease from La to Sm followed by a quasi-plateau, this latter
trend being also observed by Ustynyuk et al.4 with their
dicarboxamide aryl ligands. This difference indicates a different
ligand coordination, which was not unexpected considering
that PTA ligands coordinate as tridentate (2 N, 1 O) and
PTdA ligands coordinate as tetradentate (2 N, 2 O), yielding,
respectively, coordination numbers of 6 and 8 for PTA and
PTdA, respectively, for the expected 1:2 complexes.
Table 5. Mean Distances of the Closest Atoms around Nd in
MTPTdA Ligands for the 1:2:0, 1:2:1, and 1:1:3
Stoichiometries, Calculated in the Gas Phase
(MTPTdA)2
(MTPTdA)2NO3
MTPTdA(NO3)3
Nd−O
Nd−N
Nd−O (NO3)
2.49
2.68
NA
2.55
2.75
2.61
2.59
2.75
2.56
above, the increase in free Gibbs energies upon addition of
nitrate ions could be due to the different aqueous species
considered, as the literature usually considers the 1:1:3 species
for calculations,4,6 even though slope analysis confirmed the
presence of both 1:1 and 1:2 species.4,13 Moreover, complex
stability was reported to depend on diluents,13 which was not
considered here.
Table 6 provides the distances Nd−N and Nd−O for three
PTdA ligands without nitrate ions, as it was shown to be the
Table 6. Mean Distances of the Closest Atoms around Nd in
PTdA Ligands Having 0−2 Methyl Groups Attached to the
Aniline, Calculated in the Gas Phase without Nitrate Ions
MPPTdA
MTPTdA
MdMAPTdA
Nd−O
Nd−N
2.47
2.68
2.49
2.69
2.44
2.64
most stable. The trend is somewhat different from that of PTA,
as MTPTdA presents slightly longer distances than MPPTdA.
MdMAPTdA still has the shortest distances.
Because of the lanthanide series contraction, the ionic radii
decrease with increasing atomic number. Therefore, the
distance between lanthanide ions and their hydration
molecules in aqueous media also decreases. A similar effect
is expected for extractants in organic phase, as shown by the
calculated distances in Table 7. Interestingly, this decrease is
more pronounced for the ligands than for the hydration sphere,
as the difference between the distances increases with atomic
number, except for Pr-MTPTA and Sm-MTPTdA.
The higher distribution ratio found for Gd with PTdA has
not been fully understood yet, as this exception was not
reported for similar compounds.4,13 The first hypothesis was a
higher content of Gd in the lanthanide mix, but the ICP-MS
analysis did not show any higher concentration of Gd than the
Extraction Results. Extraction experiments of the whole
Ln series have then been performed.
D
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 5. Lanthanide distribution ratios in 0.1 M DOPTA (no phenyl), OPPTA (phenyl, 0 methyl), EHTPTA (phenyl, 1 methyl), and OTPTA
(phenyl, 1 methyl) in chloroform from HNO3 0.1 M (left) and HNO3 3 M (right).
explained by a difference in the activity of the nitrate ions
depending on the cation size.
EXAFS. Some chosen organic samples were irradiated at the
Spring8 synchrotron to compare the coordination structures of
different ligands.
As shown in Figure 9, the signal-to-noise ratio becomes very
poor for k higher than 8 due to the limited Ln concentration
obtained from extraction. Consequently, the spectra have been
deglitched and smoothed as appropriate; an example is
provided in Supporting Information Figure S10. Moreover,
the fitting has been considered only for the first two
coordination spheres, so the weight of k was kept to the first
power and the Fourier transformation of the EXAFS
oscillations was performed in the range 2−3 to 8−10 Å−1.
Figure 6. Lanthanide distribution ratios in 0.1 M DOPTA in
chloroform, 1 M NaNO3, and 0.1 M HNO3.
Figures 10 and 11 show the EXAFS vibrations versus the
radial structural functions (R-space) of Nd(NO3)3 in the
OTPTA and OTPTdA systems, and their curve fitting results
are superimposed on the real data. The curve fitting is done for
the R-space, and the fitted values in k-space are theoretical
oscillations based on the fitted values for R-space.
other Ln. There was also no significant difference in mass
balance compared to the other lanthanides, and results for
OdMAPTdA and OTPTdA were analyzed by two different
ICP-MS. It still seems suspicious, but we could not find a
plausible error, which could lead to this surprising result.
For both mono- and dicarboxamide ligands, the increase
from one to two methyl groups onto the phenyl ring shows an
increase in the distribution ratio of about a factor of 2.
Moreover, separation factors of adjacent lanthanides also
slightly increase with OdMAPTdA compared to OTPTdA,
except for La/Ce (Table 8). This observation contrasts with
the rather small increase obtained from OPPTA to OTPTA
(Figure 5), as predicted by the higher increase in free Gibbs
energy from one to two methyl groups (about 8 kJ/mol
compared to 2 kJ/mol for 0 to 1 methyl group).
Although the adjacent light lanthanide separation factors
remain similar with increasing concentration, they drastically
increase for La/Lu. Ustynyuk et al.4 reported different orders
of reaction of the ligand along the lanthanide series, with values
around 2 for the light Lns and around 1 for the heavy Lns. This
difference could explain the increasing separation factors.
Contrary to PTA ligands, which are sensitive to the proton
concentration (Figure 5), PTdA ligands show a rather similar
behavior with NaNO3 and HNO3 (Figure 8). The slightly
higher distribution ratios actually found for HNO3 are
Table 9 lists the structural parameters obtained from the
EXAFS fitting for Nd with OTPTA and OTPTdA and for Dy
with OdMAPTdA, OTPTdA, and OTPTA. The corresponding
spectra are provided in the Supporting Information (Figure
S8). The first sphere consists of the two nitrogen atoms of the
phenanthroline (per ligand in the complex) and in several
oxygen atoms from both nitrate ions and the ligand. The
number of oxygen atoms varies depending on the ligand
(mono- or dicarboxamide) and on the number (1 or 2) and
coordination mode (mono- or bidentate) of the nitrate ions.
The most probable stoichiometries, as previously mentioned,
are 1:2:2 for OTPTA and 1:2:0, 1:2:1, and 1:1:3 for OTPTdA.
Except for the 1:2:0, all stoichiometries correspond to a
coordination number of 10, which was thus fixed for the fitting.
As seen from the DFT geometry optimization and
crystallographic data,15 the carbonyl oxygen atoms are actually
nearer than the phenanthroline nitrogen atoms. The fitting was
consequently performed by relating Ln−O and Ln−N
distances so that ΔR = 0.15−0.3 Å. The evaluation of ΔR
was performed by finding the minimal residue percentage with
repeated fittings by increments of 0.01 Å. The second sphere
consists mainly of the closest carbon atoms of the phenanthro-
line and, in addition, nitrogen atoms of the nitrate ions. An
E
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 7. Lanthanide distribution ratio versus atomic number with octyl-tolyl and dimethyl-aniline extractants in chloroform, 0.1 M HNO3, 1 M
NaNO3, and various ligand concentrations.
that previously reported,15 with O slightly farther and N
Table 8. Separation Factors of Adjacent Light Ln and of
slightly closer, but the averaged Dy-O,N are very similar. The
observed difference between the two studies might be due to
the different choice of diluent, which could induce some
variation in coordination. The surprisingly higher distances
obtained for Dy-OdMAPTdA could also be linked to a diluent
change (mixture of chloroform and octanoic acid), in addition
to the spectra being particularly noisy due to the very low
concentration obtained from extraction. Estimating from the
calculations and the Nd results, the Dy−OTPTdA complex
was actually expected to be the farthest, but the EXAFS results
were not as expected. Due to the small coordination space, it is
possible that the interatomic distances between oxygen atoms
are extremely close due to the large disorder in this structure.
In any case, from the extraction efficiency, it is also highly
possible that this complex is not stable in complexation with
heavy lanthanides. Further studies, such as crystallographical
analysis, will provide more detailed information on the
structure, which will be useful for designing compounds with
size-recognition ability in heavy lanthanides as well as light
lanthanides.
Light Ln/Heavy Ln with PTdA Ligands at Different
Concentrations, 0.1 M HNO3, and 1 M NaNO3
La/Ce
Ce/Pr
Pr/Nd
Nd/Sm
La/Lu
OTPTdA
0.005 M
0.01 M
0.05 M
2.4
2.4
2.3
3.1
3.1
2.4
3.5
4.0
4.2
3.7
6.5
15.2
101
485
1335
OdMAPTdA
0.005 M
0.01 M
1.6
1.4
3.3
3.8
3.4
4.1
6.7
7.4
950
1502
overview of the considered atoms is provided in the SIF with
Figure S9.
As observed with the calculations, the interatomic distances
of both Nd and Dy seem slightly longer in OTPTdA than in
OTPTA. This could be due to the tridentate OTPTA being
less bulky than the tetradentate OTPTdA, and, with only one
arm, being more flexible to adapt to the metals. The distances
obtained by EXAFS for Dy-OTPTA are slightly different from
F
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 8. Lanthanide distribution ratios in OTPTdA 0.01 M in chloroform, 0.1 M HNO3 with additional 1 M NaNO3, 1 M HNO3, or 3 M HNO3.
Figure 9. Nd K-edge EXAFS spectra in OTPTdA and OTPTA, k space 2−10.
Figure 10. EXAFS fitting of Nd-OTPTdA and the corresponding neighboring atoms.
The difference in the distance between oxygen and nitrogen
in the first peak was smaller for Dy than for Nd (0.15 versus
0.2 Å). The obtained interatomic distances with oxygen were
2.40−2.48 Å. Considering the comparison with the Ln−O
distances in water for each coordination number,21 the
interatomic distances for Nd were generally in the range of
9−10 coordination, whereas those of Dy were closer to 8−9.
Based on the peak intensity of the second shell, it is however
more reasonable to consider the formation of a 1:2 complex for
both Dy and Nd. This results in light and heavy lanthanides
having the same stoichiometry but heavy lanthanides having
fewer coordinating bonds from nitrate ions.
The extraction properties of the ligands analyzed in this
study show that both tridentate PTA and tetradentate PTdA
have a preference for light lanthanides. It also indicates that the
bulky ligands have an optimal geometry in forming 1:2
complexes with lighter lanthanides. On the other hand, the
interatomic distances of the Dy complexes are appropriate for
the 8−9 coordination, which seems to form a rather tight
complex for these bulky ligands. It can be seen that the
separation properties of these ligands are closely related to the
G
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 11. EXAFS fitting of Nd-OTPTA and the corresponding neighboring atoms.
Table 9. Average Distances of the Atoms Nearest to Nd (Respectively Dy) as Found from EXAFS Fitting with the FEFF Files
2
Calculated from DFT Geometry Optimizations; S0 = 0.9 for All Calculations
OTPTdA/OTPTA
CN
r
σ2
Nd−O (+ O NO3)
Nd−N (+ O NO3)
Nd−C (+ N NO3)
6/6
4/4
14/9
2.55/2.54
2.75/2.74
4.10/4.00
R (Å)
0.01/0.01
0.006/0.01
0.006/0.01
σ2
OdMAPTdA/OTPTdA/OTPTA
CN
Dy−O (+ O NO3)
Dy−N (+ O NO3)
Dy−C (+ N NO3)
6/6/6
4/4/4
14/14/9
2.48/2.43/2.42
2.63/2.59/2.56
3.93/3.88/3.92
0.001/0.002/0.0011
0.001/0.001/0.001
0.001/0.001/0.001
size of the ion, the surface charge density, and the structure of
the molecule. In the comparison between OTPTdA and
OdMAPTdA complexes, contradictory results were obtained
from calculation and EXAFS fitting. Despite showing larger
interatomic distances, OdMAPTdA provides higher extraction
ability, thus proving a better complex stability. This result
shows that the correlation between the properties of the
electron donor atoms, i.e., oxygen and nitrogen atoms, and the
overall structure of the molecule has a large effect on the
stability of the complex and thus the separation properties. In
addition, in comparing the PTA and PTdA systems, the
inability of the PTdA ligand to adjust to heavier lanthanides
indicates that the flexibility of the molecular structure is an
essential parameter for the complexation. From a different
point of view, the separation factors in light lanthanides were
larger in the PTdA systems than that in the PTA systems,
indicating a more acute recognition of size differences. On the
other hand, despite having a smaller dependence on atomic
number due to a better flexibility, the PTA structure has its
best size fit in the vicinity of Nd, resulting in a maximum of
extraction, which was not observed for the PTdA ligands.
between adjacent light lanthanides and then a quasi-plateau.
On the contrary, monocarboxamide ligands show a maximum
of extraction around Nd/Sm and then a decrease all along.
With monocarboxamide ligands, separation factors from
adjacent lanthanides are quite low.
EXAFS studies confirmed the presence of nitrate ions within
the first coordination shell, and differences in Ln−O,N
distances could be linked to the extraction results. The main
difference between light and heavy lanthanides lies in the
number of nitrate ions within the first coordination sphere.
Results also stressed the importance of the extractant overall
structure for the efficiency and selectivity of the extraction.
Even if the coordination site remains the most important
parameter, a modification on the outer sphere of the extractant
can also impact the extraction. Here, a significant change in the
extraction efficiency arose by modifying the amide aromatic
substituent. Besides the aryl effect, the number of methyl
substitutions was also found to be an important parameter,
most probably due to a change in the ligand polarity. This
effect could be extrapolated to other N-based carboxamide
ligands, such as pyridines, to improve their extraction
efficiency. The high affinity of the dicarboxamide ligands for
the light lanthanides can also make them good candidates for
nuclear reprocessing, as previous studies also showed high
affinity for Am.
CONCLUSIONS
■
In this study, we have performed the extraction of the
lanthanide series from nitrate media by several phenanthroline
carboxamide ligands in chloroform. A different extraction
pattern is observed between mono- and dicarboxamide ligands.
Dicarboxamide ligands show decreasing distribution ratios
along the whole series, with reasonably high separation factors
H
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
computational and experimental studies,″ Chemical Communications.
Chem Commun. 2015, 51, 7466−7469.
(5) Xiao, C.-L.; Wang, C.-Z.; Yuan, L.-Y.; Li, B.; He, H.; Wang, S.;
Zhao, Y.-L.; Chai, Z.-F.; Shi, W.-Q. Excellent Selectivity for Actinides
with a Tetradentate 2,9-diamide-1,10-phenanthroline Ligand in
Highly Acidic Solution: a Hard-soft donor Combined Strategy.
Inorg. Chem. 2014, 53, 1712−1720.
(6) Wang, C.; Wu, Q.-Y.; Wang, C.-Z.; Lan, J.-H.; Nie, C.-M.; Chai,
Z.-F.; Shi, W.-Q. Theoretical insights into selective separation of
trivalent actinide and lanthanide by ester and amide ligands based on
phenanthroline skeleton. Dalton Trans. 2020, 49, 4093−4099.
(7) Alyapyshev, M.; Ashina, J.; Dar’in, D.; Kenf, E.; Kirsanov, D.;
Tkachenko, L.; Legin, A.; Starova, G.; Babain, V. 1,10-phenanthroline-
2,9-carboxamides as ligands for separation and sensing of hazardous
metals. RSC Adv. 2016, 6, 68642−68652.
(8) Alyapyshev, M.; Babain, V.; Tkachenko, L.; Kenf, E.; Voronaev,
I.; Dar’in, D.; Matveev, P.; Petrov, V.; Kalmykov, S.; Ustynyuk, Y.
Extraction of actinides with heterocyclic dicarboxamides. J. Radioanal.
Nucl. Chem. 2018, 316, 419−428.
(9) Zarubin, D. N.; Bushkov, N. S.; Lavrov, H. V.; Dolgushin, F. M.;
Ustynyuk, N. A.; Ustynyuk, Y. A. 4,7-di-n-butoxy-1,10-phenanthro-
line-2,9-dicarboxamide: a tetradentate ligand featuring excellent
solubility in nonpolar media. INEOS Open 2019, 2, 130−133.
(10) Makrlík, E.; Vanura, P.; Selucky, P.; Babain, V.; Dar’in, D.;
Alyapyshev, M. N,N,N′,N′-tetrabutyl-1,10-phenanthroline-2,9-dicar-
boxamide as very effective extraction agent for trivalent europium and
americium. Acta Chim. Slov. 2017, 64, 582−589.
(11) Manna, D.; Soumyaditya, M.; Arunasis, B.; Subrata, C.; Tapan,
K. G. Actinide selectivity of 1,10-phenanthroline-2,9-dicraboxamide
and its derivatives: at theoretical prediction followed by experimental
validation. Dalton Trans. 2015, 44, 1332−1340.
(12) Lemport, P. S.; Matveev, P. I.; Yatsenko, A. V.; Evsiunina, M.
V.; Petrov, V. S.; Tarasevich, B. N.; Roznyatovsky, V. A.;
Dorovatovskii, P. V.; Khrustalev, V. N.; Zhokhov, S. S.; Solov’ev, V.
P.; Aslanov, L. A.; Petrov, V. G.; Kalmykov, S. N.; Nenanjdenko, V.
G.; Ustyniuk, Y. A. The impact of alycyclic substituents on the
extraction ability of new family of 1,10-phenanthroline-2,9-diamides.
RSC Adv. 2020, 10, 26022.
(13) Tsutsui, N.; Ban, Y.; Suzuki, H.; Nakase, M.; Ito, S.; Inaba, Y.;
Matsumura, T.; Takeshita, K. Effects of diluents on the Separation of
Minor Actinides from Lanthanides with tetradodecyl-1,10-phenan-
throline-2,9-diamide from Nitric Acid Medium. Anal. Sci. 2020, 36,
241−246.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.inorgchem.1c01729.
ESI-MS main masses and their identification for OTPTA
and OTPTdA after chromatography; 1H NMR spectra
of each synthesized ligand in CDCl3; HMQC spectrum
of OTPTA in CDCl3; EXAFS fitting of Dy-OdMAPTdA
and Dy-OTPTA; atoms involved in the 1st and 2nd
coordination spheres; and signal treatment prior to
EXAFS fitting (PDF)
AUTHOR INFORMATION
■
Corresponding Author
Marie Simonnet − Materials Sciences Research Center, Japan
Atomic Energy Agency, 319-1195 Tokai-Mura, Japan;
orcid.org/0000-0001-8348-1177;
Email: simonnet.marie@jaea.go.jp
Authors
Tohru Kobayashi − Materials Sciences Research Center, Japan
Atomic Energy Agency, Sayo-gun, Hyogo 679-5148, Japan
Kojiro Shimojo − Materials Sciences Research Center, Japan
Atomic Energy Agency, Sayo-gun, Hyogo 679-5148, Japan;
orcid.org/0000-0001-8160-8373
Keiichi Yokoyama − Materials Sciences Research Center,
Japan Atomic Energy Agency, Sayo-gun, Hyogo 679-5148,
Japan
Tsuyoshi Yaita − Materials Sciences Research Center, Japan
Atomic Energy Agency, Sayo-gun, Hyogo 679-5148, Japan
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.inorgchem.1c01729
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Prof. Mori (Kobe University) for teaching
the synthesis method and Dr. Kathleen Schnaars (JAEA) for
the ESI-MS measurements. The synchrotron radiation experi-
ments were performed at BL22XU in SPring8 with the
approval of RIKEN (proposal no. 2020A3711). Authors would
also like to thank the New Energy and Industrial Technology
Development Organization (NEDO) for the comissionned
project this study belongs to.
(14) Simonnet, M.; Suzuki, S.; Miyazaki, Y.; Kobayashi, T.;
Yokoyama, K.; Yaita, T. Lanthanide intra-series separation by a
1,10-phenanthroline derivative: counterion effect. Solvent Extr. Ion
Exch. 2020, 38, 430−440.
(15) Kobayashi, T.; Suzuki, S.; Shiwaku, H.; Yaita, T. Lanthanides
complexation properties of O, N-hetero donor ligand PTA. Prog. Nucl.
Sci. Technol. 2018, 5, 74.
(16) Nakase, M.; Kobayashi, T.; Suzuki, S.; Sihwaku, H.; Grimes, T.
S.; Mincher, B. J.; Yaita, T. Relationship between structure and
coordination strength of N and N,O-hybrid donor ligands with
trivalent lanthanides. Solvent Extr. Ion Exch. 2018, 36, 633−646.
(17) Phifer, C. C.; McMillin, D. R. The basis of aryl substituent
effects on charge-transfer intensities. Inorg. Chem. 1986, 25, 1329.
(18) Ustynyuk, Y. A.; Gloriozov, I. P.; Kalmykov, S. N.; Mitrofanov,
A. A.; Babain, V. A.; Alyapyshev, M. Y.; Ustynyuk, N. A.
Pyridinedicarboxylic acid diamides selective ligands for extraction
and separation of trivalent lanthanides and actinides: DFT study.
Solvent Extr. Ion Exch. 2014, 32, 508−528.
(19) Myasoedov, B. F.; Chmutova, M. K. Main Regularities in
Extraction of Actinides by Bidentate Neutral Organophosphorus
Compounds. In Recent Progress in Actinides Separation Chemistry -
Proceedings of WASC;Zenko Yoshida, T. K. Y. M., Ed.; World
Scientific Publishing Co. Pte. Ltd., 1994; Vol. 94, pp 1−17.
(20) Baldwin, A. G.; Ivanov, A. S.; Williams, N. J.; Ellis, R. J.; Moyer,
B. A.; Bryantsev, V. S.; Shafer, J. C. Outer-sphere water clusters tune
REFERENCES
■
(1) Xie, F.; Zhang, T. A.; Dreisinger, D.; Doyle, F. A critical review
on solvent extraction of rare earths from aqueous solutions. Miner.
Eng. 2014, 56, 10−28.
(2) Li, D. Development course of separating rare earths with acid
phosphorus extractants: A critical review. J. Rare Earths 2019, 37,
468−486.
(3) Tunsu, C.; Menard, Y.; Eriksen, D. O.; Ekberg, C.; Petranikova,
M. ″Recovery of critical materials from mine tailings: A comparative
study of the solvent extraction of rare earths using acidic, solvating
and mixed extractant systems. J. Cleaner Prod. 2019, 218, 425−437.
(4) Ustynyuk, Y. A.; Borisova, N. E.; Babain, V. A.; Gloriozov, I.;
Manuilov, A. Y.; Kalmykov, S. N.; Alyapyshev, M. Y.; Tkachenko, L.
I.; Kenf, E. V.; Ustynyuk, N. A. ″N,N′-dialkyl-N,N′-diaryl-1,10-
phenanthroline-2,9-dicarboxamides as donor ligands for separation of
rare earth elements with a high and unusual selectivity. DFT
I
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
the lanthanide selectivity of diclycolamides. ACS Cent. Sci. 2018, 4,
739−747.
(21) D’Angelo, P.; Zitolo, A.; Migliorati, V.; Chillemi, G.; Duvail,
M.; Vitorge, P.; Abadie, S.; Spezia, R. Revised ionic radii of
lanthanoid(III) ions in aqueous solution. Inorg. Chem. 2011, 50,
4572−4579.
J
https://doi.org/10.1021/acs.inorgchem.1c01729
Inorg. Chem. XXXX, XXX, XXX−XXX