Complexation of N4-tetradentate ligands with Nd(III) and Am(III)

Article abstract of DOI:10.1007/s10953-012-9928-y

To improve understanding of aza-complexants in trivalent actinide-lanthanide separations, a series of tetradentate N-donor ligands have been synthesized and their complexation of americium(III) and neodymium(III) investigated by UV-visible spectrophotometry in methanolic solutions. The six pyridine/alkyl amine/imine ligands are N,N′-bis(2-methylpyridyl)-1,2- diaminoethane, N,N′-bis(2-methylpyridyl)-1,3-diaminopropane, trans-N,N-bis(2-pyridylmethyl)-1,2-diaminocyclohexane (BPMDAC), N,N'-bis(2-pyridylmethyl)piperazine, N,N'-bis-[pyridin-2-ylmethylene]ethane-1,2- diamine, and trans-N,N-bis-([pyridin-2-ylmethylene]-cyclohexane-1,2-diamine. Each ligand has two pyridine groups and two aliphatic amine/imine N-donor atoms arranged with different degrees of preorganization and structural backbone rigidity. Conditional stability constants for the complexes of Am(III) and Nd(III) by these ligands establish the selectivity patterns. The overall selectivity of Am(III) over Nd(III) is similar to that reported for the terdentate bis(dialkyltriazinyl)pyridine molecules. The cyclohexane amine derivative (BPMDAC) is the strongest complexant and shows the highest selectivity for Am(III) over Nd(III) while the imines appear to prefer a bridging arrangement between two cations. These results suggest that this series of ligands could be employed to develop an enhanced actinide(III)- lanthanide(III) separation system.

Full text of DOI:10.1007/s10953-012-9928-y

J Solution Chem (2012) 41:2138–2153
DOI 10.1007/s10953-012-9928-y
Complexation of N4-Tetradentate Ligands with Nd(III)
and Am(III)
•
Mark D. Ogden Serguei I. Sinkov G. Patrick Meier
•
•
•
Gregg J. Lumetta Kenneth L. Nash
Received: 10 November 2010 / Accepted: 23 February 2012 / Published online: 6 December 2012
Ó Springer Science+Business Media New York 2012
Abstract To improve understanding of aza-complexants in trivalent actinide–lanthanide
separations, a series of tetradentate N-donor ligands have been synthesized and their
complexation of americium(III) and neodymium(III) investigated by UV–visible spectro-
photometry in methanolic solutions. The six pyridine/alkyl amine/imine ligands are N,N0-
bis(2-methylpyridyl)-1,2-diaminoethane, N,N0-bis(2-methylpyridyl)-1,3-diaminopropane,
trans-N,N-bis(2-pyridylmethyl)-1,2-diaminocyclohexane (BPMDAC), N,N’-bis(2-pyridyl-
methyl)piperazine, N,N’-bis-[pyridin-2-ylmethylene]ethane-1,2-diamine, and trans-N,N-
bis-([pyridin-2-ylmethylene]-cyclohexane-1,2-diamine. Each ligand has two pyridine
groups and two aliphatic amine/imine N-donor atoms arranged with different degrees of
preorganization and structural backbone rigidity. Conditional stability constants for the
complexes of Am(III) and Nd(III) by these ligands establish the selectivity patterns. The
overall selectivity of Am(III) over Nd(III) is similar to that reported for the terdentate
bis(dialkyltriazinyl)pyridine molecules. The cyclohexane amine derivative (BPMDAC) is
the strongest complexant and shows the highest selectivity for Am(III) over Nd(III) while
the imines appear to prefer a bridging arrangement between two cations. These results
suggest that this series of ligands could be employed to develop an enhanced actinide(III)–
lanthanide(III) separation system.
Keywords Am(III) ꢀ Nd(III) ꢀ Complexation ꢀ Nitrogen donor ligands ꢀ
Spectrophotometry
M. D. Ogden (&) ꢀ G. P. Meier ꢀ K. L. Nash (&)
Department of Chemistry, Washington State University, P.O. Box 644630, Pullman,
WA 99164-4630, USA
e-mail: mark.ogden@ansto.gov.au
K. L. Nash
e-mail: knash@wsu.edu
Present Address:
M. D. Ogden
ANSTO Minerals, Locked Bag 2001, Kirrawee DC, NSW 2232, Australia
S. I. Sinkov ꢀ G. J. Lumetta
Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99354, USA
123

J Solution Chem (2012) 41:2138–2153
2139
1 Introduction
A serious limiting factor in the continued development of nuclear power is management of
irradiated nuclear fuel. One strategy for reducing the long-term radiotoxicity of spent
nuclear fuel is the transmutation of long-lived a-emitting actinide elements into shorter-
lived fission products by bombardment with fast neutrons. Unfortunately, the lanthanide
(Ln) fission products, which represent 40 % by mass and 1/3 of the atoms created in U/Pu
fission, complicate transmutation as they have high neutron absorption cross sections [1],
thus competing with actinides for neutrons. Therefore, the separation of lanthanides from
actinides, most importantly americium and curium, is critical to developing advanced
nuclear fuel management strategies involving transmutation. Owing to the chemical sim-
ilarities of Ln(III) and An(III), the mutual separation of the two groups is one of the most
challenging tasks in separation science. The trivalent ions of the 4f and 5f series are hard
acids in the Pearson classification of acids and bases [2], and thus preferentially interact
with hard donor ligands like O and F- more strongly than with N, S or P donor ligands.
Results in the previous literature have demonstrated that donor atoms softer than oxygen,
particularly N and S, are most effective for accomplishing the separation of the groups
using liquid–liquid extraction systems [3], despite the weaker bond strength. It is believed
that the slightly softer character of the actinide ions compared to the lanthanide ions
accounts for this observation.
The effectiveness of nitrogen groups for this purpose is illustrated by the complexation
of Sm3? and Am3? by hexadentate N,N,N0,N0-tetrakis(2-pyridylmethyl)-ethylenediamine
(TPEN) (Fig. 1), the 2-methylpyridyl structural analogue of ethylendiaminetetraacetic acid
(EDTA). Jensen and coworkers [4] report that the TPEN stability constant for Am3? is 100
times larger than that of the Sm3? complex; at the same time, it is 1011 times weaker than
the EDTA complex [4]. Those authors attribute the increased actinide selectivity of
nitrogen-donor ligands to a more exothermic energy of binding. In a systematic study of
the thermodynamics of complexation of Ln(III) and Am(III) with the planar tridentate
nitrogen ligand 2-amino-4,6-di-(pyridin-2-yl)-1,3,5-triazine (ADPTZ) (Fig. 1) at 25 °C in
75/25 vol% methanol/water, Miguirditchian reported that the enthalpy of binding to
Am(III) was more exothermic than that of Ln(III) species [5].
It also has been shown that more sterically constrained ligands that focus potential
donor groups in a favorable geometry have an increased overall binding affinity. The net
increase in interaction strength results from a reduced energetic penalty associated with
reorganizing ligand structures as the complex is formed, and under suitable circumstances
a greater interaction strength in the metal-donor atom bonding. The effect of preorientation
of ligand donor atoms on binding is exemplified in the comparison of lanthanide(III)
complexes with 1,10-phenanthroline-2,9-dicarboxylic acid (PDA) (Fig. 1) and N,N0-
ethylenediaminediacetic acid (EDDA). The log10 K101 values for La(III) and Gd(III) with
EDDA are 7.04 and 8.13, respectively [6]. The preorganization of PDA donor groups into a
potentially tetradentate planar arrangement allows coordination of ligand donor atoms to
the metal ion without the need to rearrange donor groups in the ligand. Hancock and
coworkers report log10 K101 values of 13.5 and 16.1 for La(III) and Gd(III) respectively,
which are 106–108 times stronger than EDDA [6].1
1
In this paper stability constant and DG values for complexation are given in MHL stoichiometric format.
That is, a log10 K value for the complexation of x metal ions, y protons and z ligands would be denoted as
log10 Kxyz, and the Gibbs energy as DGmhl.
123

2140
J Solution Chem (2012) 41:2138–2153
NH2
N
N
N
N
N
N
N
N
N
N
HO
O
N
N
N
O
HO
ADPTZ
TPEN
PDA
O
OH
O
O
N
N
N
R
N
N
N
N
R
R
N
N
N
N
R
N
N
N
R
HO
N
N
R
N
OH
R
R
HO
O
R-BTP
R-BTBP
DCTA
N
N
N
N
N
N
N
N
N
Terpy
TPTZ
Fig. 1 Structurally focused and soft nitrogen based complexants and extractants, discussed and used as
comparisons in this paper, are TPEN = N,N,N0,N0-tetrakis(2-pyridylmethyl)-ethylenediamine, ADPTZ =
2-amino-4,6-di-(pyridin-2-yl)-1,3,5-triazine, PDA = 1,10-phenanthroline-2,9-dicarboxylic acid, CDTA =
trans-1,2-cyclohexanedinitrilotetraacetic acid, R-BTBP = 6,60-bis-(5,6-di(R,R)-[1, 2, 4]-triazin-3-yl]-
[2,20]bipyridine, R-BTP = 2,6-bis-(5,6-di(R,R)-[1, 2, 4]-triazin-3-yl)pyridine, Terpy = 2,20,60,200-terpyri-
dine, TPTZ = 2,4,6-tri-(2-pyridyl)-1,3,5-triazine
Although this difference cannot be attributed fully to steric factors (due to the different
basicities of the aliphatic and aromatic nitrogen atoms), there is not doubt that the struc-
tural rigidity must contribute to the difference in complex stability.
A class of compounds that combine nitrogen heterocycles with a degree of structural
preorientation are the 2,6-pyridine bistriazine (R-BTP) and 2,9-bipyridine-bistriazine
(R-BTBP) complexants (Fig. 1) that have been extensively investigated in the European
advanced fuel cycle research program [7]. In these lipophilic complexants, three or four
soft nitrogen donor groups are found in a planar preorientation with a bonding pocket
properly sized for the coordination of trivalent lanthanide and actinide cations. Examples
of these reagents have been successfully tested for the separation of An(III) from Ln(III).
The alkyl groups on the triazines rings of these molecules increase the solubility of the
molecule in aliphatic diluents. The BTP extractants act as tridentate ligands and form metal
complexes with donor atoms including the nitrogen atom of the central pyridine ring and
the nitrogen atoms at the 2-position of the triazinyl rings [8]. These compounds have
demonstrated separation factors of up to 100 for Am3?/Eu3? [9]. Unfortunately, reagent
instability, slow phase transfer kinetics, and less than ideal compatibility with aliphatic
diluents have been issues [8].
The aim of this investigation is to explore the relationship between ligand structural
features, basicity of the ligand donor atoms and relative Ln(III)/An(III) binding affinity for
123

J Solution Chem (2012) 41:2138–2153
2141
NH
NH
NH NH
NH NH
N
N
N
N
N
N
BPMDAC
BPMDAE
BPMDAP
N
N
N
N
N
N
N
N
N
N
N
N
BPMPIP
BPMDIC
BPMDIE
Fig. 2 Sterically hindered 2-methylpyridyl ligands synthesized in this study, where N,N0-bis(2-methylpyr-
idyl)-1,2-diaminoethane (BPMDAE), N,N0-bis(2-methylpyridyl)-1,3-diaminopropane (BPMDAP), trans-
N,N-bis(2-pyridylmethyl)-1,2-diaminocyclohexane
(BPMDAC),
N,N’-bis(2-pyridylmethyl)piperazine
(BPMPIP), N,N’-bis-[pyridin-2-ylmethylene]ethane-1,2-diamine (BPMDIE), and trans-N,N-bis-([pyridin-
2-ylmethylene]-cyclohexane-1,2-diamine (BPMDIC)
a series of structurally related tetraaza-pyridine derivatives. The ultimate objective is to
advance knowledge of such bonding interactions to enable the preparation of even more
effective An(III)/Ln(III) separations reagents. The structures of the ligands synthesized,
their correct names and abbreviations are given in Fig. 2. Cation bonding affinity is
assessed through spectrophotometric complexation studies of neodymium(III) and ameri-
cium(III) in methanol at low ionic strength to determine stability constants, which allow
assessment of metal complex speciation. Methanol was selected as the medium to improve
ligand solubility and to reduce water activity in the system.2
2 Experimental
2.1 Synthesis of N4-Tetradentate Ligands
The ligands were synthesized following established procedures with small alterations for
optimization as discussed previously [10]. Established literature procedures are as follows
[11] for N,N0-bis(2-methylpyridyl)-1,2-diaminoethane (BPMDAE), N,N’-bis-[pyridin-2-
ylmethylene]ethane-1,2-diamine (BPMDIE), and N,N0-bis(2-methylpyridyl)-1,3-diamino-
propane (BPMDAP), [12] for trans-N,N-bis-([pyridin-2-ylmethylene]-cyclohexane-1,2-
diamine (BPMDIC), and [13] for trans-N,N-bis(2-pyridylmethyl)-1,2-diaminocyclohexane
(BPMDAC). The diimines were synthesized by condensation of a chosen pyridine
2
The interactions of f-element with water are strong enough to inhibit metal–amine bonding, and thus the
water content must be reduced. As the ultimate destination of these reagents (after suitable modification to
increase their lipophilicity) is the organic phase of a solvent extraction system, the methanol results are
considered an adequate indicator of the relative interaction strength of Ln(III)/An(III) with the subject
ligands. This technique has been employed in the BTP and similar previous investigations.
123

2142
J Solution Chem (2012) 41:2138–2153
˚
carboxaldehyde with the appropriate diamine in methanol over 5 A molecular sieves. The
tetradentate diamines were then prepared by reduction of the diimine with sodium boro-
hydride (NaBH4) in ice cold MeOH. After purification by solvent extraction, the diamines
were precipitated as their HCl salts [10]. Ligand stock solutions of compounds N,N’-bis
(2-pyridylmethyl)piperazine (BPMPIP), BPMDIE, and BPMDIC were prepared by dis-
solving the appropriate amount of the solid ligand in MeOH. Ligand stock solutions of
BPMDAE, BPMDAP, and BPMDAC were made up by neutralization of the HCl salt with
0.1 molꢀL-1 NaOH and solvent extraction into methylene chloride (CH2Cl2). The CH2Cl2
solution was dried using anhydrous sodium carbonate filtered and evaporated to give the
free amine compound as an oil. The oil was then weighed and dissolved in anhydrous
MeOH. For all ligands, their purity was determined by 13C and H NMR and comparison
1
with literature values [11–13].
2.2 Stock Solutions
The Nd(NO3)3 stock was made up from a NIST standard (25 mL) Nd(NO3)3
(0.06933 molꢀL-1) solution in 5 % HNO3; the solution was fumed a number of times to
remove excess HNO3. The stock was then reduced in volume to produce Nd(NO3)3
crystals. The crystals were collected and dissolved in anhydrous MeOH (99.8 %). The
methanol was evaporated 5 times to remove excess water from the stock solution. An
aliquot of this solution was taken and a working solution made. The concentration of Nd3?
{4.62 (2) 9 10-3 molꢀL-1} was verified spectrophotometrically (kmax = 579.95 nm,
emax = 11.3 Lꢀmol-1ꢀcm-1) [10].
The 243Am stock was made up by reconstitution of a dry sample using 2000 lL of
deionized water and 100 lL of 15.8 molꢀL-1 HNO3. A working stock was made by taking
250 lL of the mother solution in 14000 lL of deionized water. The concentration of Am3?
(2.379 9 10-3 molꢀL-1
)
was verified by spectrophotometry (kmax = 815.9 nm,
emax = 64.4 Lꢀmol-1ꢀcm-1) [14]. The working solution was then made up in MeOH
following the same approach described above for Nd3?. The Am3? working solution used
in the media effect study was evaporated eight times rather than five.
2.3 Spectrophotometric Titrations
All titrations were carried out at 0.03 molꢀL-1 ionic strength (by addition of NaNO3). To
an aliquot of ligand working solution, in a stoppered quartz cuvette suitable for magnetic
stirring (1 cm path length), incremental additions were made of a ligand solution in MeOH.
Spectra were recorded using an Ocean Optics S2000 spectrometer calibrated for the
wavelength range 348–1000 nm. The spectral region used to fit the stability constants was
720–770 nm for Nd(III) and 495–525 nm for Am(III). It was assumed that pH and ionic
strength remained constant during the titration. Stability constants were fit using the
SQUAD program [15] taking metal ion dilution into account. SQUAD is designed to
calculate the best values for stability constants from absorbance data by employing a
nonlinear least-squares approach. A satisfactory model of the complexation occurring in
solution is determined by an iterative process reducing the standard deviation in calculated
SQUAD parameters, the absorption data, the refined stability constant, and the standard
deviation of each spectrum. The validity of the species model can also be indicated by the
overall quality and shape of the calculated molar absorptivities. Non-negative least-squares
algorithm (NNLS) was used to remove negative molar absorptivity values after all
123

J Solution Chem (2012) 41:2138–2153
2143
plausible models had been tested. Considered in the model were M2L, ML and ML2
species in solution and best fit were determined by an iterative process.
2.4 Effect of Changing Media
To an aliquot of either Nd3? or Am3? working stock at 0.03 molꢀL-1 ionic strength (by
addition of NaNO3) in a stoppered quartz cuvette (1 cm path length), incremental additions
(10 lL) were made, initially up to seven additions then increasing for 20 (two additions),
25 (two additions), 50 (four additions), and a final addition of 100 lL of a pH = 3 aqueous
HNO3 solution were added. Spectra were recorded as noted above.
3 Results
3.1 The Effect of Changing Media
In Fig. 3a, b the effect of incremental additions of aqueous solution upon the methanolic
spectra of Nd3? and Am3? is shown. The spectra of Nd3? (Fig. 3a) after addition of 8
molar equivalents (8H2O:1 Nd3?) of water, the spectra resembles that of an aqueous
Nd(NO3)3 (reference) solution at 0.000272 molꢀL-1 (data not shown). The effect of
changing the medium is more clearly seen in the spectra of Am3? following the changes in
7
5
the F0 ? L6 absorption (503 nm) (Fig. 3b). A number of distinct species (2–3) can be
seen in the spectral pattern upon changing from methanolic media to aqueous media. The
shape of the band is markedly different from the aqueous peak and may indicate the
presence of multiple species in the methanol solution. In the purely methanolic spectra of
Am3? there are two distinct peaks at 505 and 511 nm, with an observable shoulder at
514 nm (Fig. 3b). By changing the Am3? solution from five to eight evaporations the
Fig. 3 The effect of changing media on M(III) nitrate in MeOH: a 0.00462 molꢀL–1 Nd3?, the black trace
lines are for the metal ion in pure methanol media and the grey lines are for mixed aqueous media, 73/27 %
v/v MeOH/H2O; b 0.00238 molꢀL-1 Am3?, the black trace lines are for the metal ion in pure methanol
media and grey lines are for mixed aqueous media, 82/18 % v/v MeOH/H2O. For both spectra incremental
aqueous additions are shown in light grey
123

2144
J Solution Chem (2012) 41:2138–2153
difference seen in the shape of the Am3? absorption band around 511 nm indicates
changes in the Am3? speciation.
3.2 Determination of Stability Constants
The results of the individual spectrophotometric titration can be seen for all six of the
N4-tetradentate ligands in Fig. 4 for Nd3? and Fig. 5 for Am3?. The stability constants
calculated from the spectra using the non-negative least squares (NNLS) method of fitting
in SQUAD are given in Table 1. It is assumed in the fitting that the concentration of free
H? is negligible in the methanolic solution and more importantly that it does not change
during the titration. Modeling of the species present in solution were completed using an
iterative process by minimization of the error data in the SQUAD output files fitting of the
absorbance data. The simplest model that fit the data was accepted as the most probable
model. The species considered were ML, ML2 and M2L. ML was taken first as the most
probable complex, with other species added as appropriate to minimize the overall
residuals of the fit. Species attaining a value greater than 0.05 in the sum of squares of the
standard deviation in spectral fitting were removed from the model. Using the formation of
the M2L species, in aqueous solution as an example, the stability constant presented is an
overall stability constant for the reaction designated by the symbol b:
b201
3þ
2M þ LðaqÞ ꢀM2L6ðaþqÞ
ðaqÞ
It can be seen readily in Fig. 4 that the spectral patterns for BPMDAE and BPMDAC are
similar in shape and intensity, possibly hinting at comparable binding symmetry around the
Nd3? metal center. For BPMPIP the spectra undergo minimal changes, which is evidence
for a weaker interaction when compared to all the other ligands. In the spectra for BPMDIE
and BPMDIC with Nd3?, similar shape and intensity is seen between the spectra of the two
coordination compounds with k max shifting toward shorter wavelengths relative to that of
the solvated cation. The kmax for Nd–BPMDAP also demonstrates a distinctive shift, but
toward longer wavelengths. For the Am3? complexes with BPMDAC and BPMDAP
(Fig. 5). evidence can be seen for the formation of two spectroscopically distinct species in
solution. The band with maxima at 512 and 520 nm on the shoulder of the predominant
510 nm band is seen at a 1:3 mol ratio for BPMDAP and 1:1.5 for BPMDAC. In the
terminal complex (at 510 nm in each case) the 512/520 nm bands are suppressed. How-
ever, for all the data we initially see the formation of a single spectroscopic envelope in
solution which has been assigned with the simplest assumption that only a distinct 1:1 ML
complex is formed in solution for BPMDAP while for BPMDAC a 1:2 ML complex was
added to the fit. This added complexity is also seen in the Am3?–BPMDAE spectra, though
the experiment was terminated at a 1:3 mol ratio.
With the BPMDIE and BPMDIC ligands the formation of a 2:1 (M2L) complex with the
ligand bridging two metal centers is proposed as the most probable structure. The 2:1 M:L
complex was the iterative best fit of the UV–Vis spectra using SQUAD. Precipitation of the
Am3? complex with BPMDIE and BPMDIC, at a ligand to metal ratio of 1:1, that ended
the titration, could indicate the formation of a polymeric 1:1 species in solution. For the
fitting of stability constants for BPMPIP in Figs. 4 and 5 for Nd3? and Am3?, respectively,
dilution was taken into account and the weak stability constants calculated from the subtle
changes in the 735 and 745 nm spectral regions for Nd3? and the 506–515 nm region for
Am3?
.
123

J Solution Chem (2012) 41:2138–2153
2145
Fig. 4 Nd(III) spectrophotometric titrations with incremental additions of N4-tetradentate diamine based
ligands, using a 1 cm cuvette. Black lines represent Nd3? (4.62 9 10-3 molꢀL-1) in methanolic media,
dotted grey lines represent incremental additions and dark grey line represents the end point: a BPMDAE
(5.00 9 10-2 molꢀL-1), the end point results in a M:L ratio of 1:3.0; b BPMDAP (5.02 9 10-2 molꢀL-1),
the end point results in a M:L ratio of 1:4.1; c BPMDAC (5.00 9 10-2 molꢀL-1), the end point results in a
M:L ratio of 1:1.5; d Nd3? (4.99 9 10-3 molꢀL-1) with BPMPIP (5.18 9 10-2 molꢀL-1), the end point
results in a M:L ratio of 1:10.4; e BPMDIE (5.151 9 10-2 molꢀL-1), the end point results in a M:L ratio of
1:1.3; f BPMDIC (5.12 9 10-2 molꢀL-1), the end point results in a M:L ratio of 1:1.6
123

2146
J Solution Chem (2012) 41:2138–2153
Fig. 5 Am(III) (2.20 9 10-3 molꢀL-1) spectrophotometric titrations with incremental additions of N4-
tetradentate diamine based ligands, using a 1 cm cuvette. The black lines represent Am3? in menthanolic
media and the dotted grey lines represent incremental additions: a BPMDAE (5.00 9 10-2 molꢀL-1), the
thick dark grey line is the end point resulting in a M:L ratio of 1:3.0; b BPMDAP (5.02 9 10-2 molꢀL-1),
the thick light grey line is the primary species formed in solution at a M:L ratio of 1:3.0 and the thick dark
grey line is the end point resulting in a M:L ratio of 1:9.9; c BPMDAC (5.00 9 10-2 molꢀL-1), the thick
light grey line is the primary species formed in solution at a M:L ratio of 1:1.5 and the thick dark grey line is
the end point resulting in a M:L ratio of 1:7.9; d BPMPIP (5.05 9 10-2 molꢀL-1), the thick dark grey line is
the end point resulting in a M:L ratio of 1:10.4; e BPMDIE (5.14 9 10-2 molꢀL-1), the end point resulting
in a M:L ratio of 1:1.0; f BPMDIC (5.12 9 10-2 molꢀL-1), the end point resulting in a M:L ratio of 1:1.0
123

J Solution Chem (2012) 41:2138–2153
2147
Table 1 The stability constants (log10 K) for the binding N4-tetradentate ligands with Am3? and Nd3? in
MeOH determined at 21 °C (value for the complexation of x metal ions, y protons and z ligands is denoted
as log10 Kxyz
)
Ligand
Nd3? log10 K201
Nd3? log10 K101
Nd3? log10 K102
BPMDAE
BPMDAP
BPMDAC
BPMPIP
4.6 0.1
4.49 0.08
5.3 0.1
–
2.01 0.06
2.15 0.06
3.10 0.08
0.72 0.08
–
–
–
–
–
–
BPMDIE
BPMDIC
4.2 0.1
4.84 0.05
Am3? log10 K201
Am3? log10 K101
Am3? log10 b102
BPMDAE
BPMDAP
BPMDAC
BPMPIP
4.81 0.07
4.75 0.05
5.40 0.03
–
3.01 0.08
3.04 0.05
4.14 0.09
1.20 0.08
–
–
5.87 0.07
5.81 0.2
–
–
–
BPMDIE
BPMDIC
5.29 0.08
5.91 0.06
–
4 Discussion
4.1 The Effect of Changing Media
The methanolic spectrum of Nd3? is markedly different from that seen in aqueous solution
(Fig. 3). This probably indicates that anions (i.e., nitrate), which do not usually coordinate
in aqueous solution, are able to enter the inner coordination sphere of the metal in
anhydrous methanol. In the work of Tanaka et al. [16], the solvent composition in the
primary solvation sphere of Eu3? was determined using luminescence. In the presence of
Cl- and ClOꢁ4 in the MeOH/H2O binary solvent mixtures it was found that even at bulk
phase molar fractions of water as low as 0.1, the metal ion is preferentially solvated by
water with approximately one methanol entering the inner coordination sphere at 90 %
MeOH/H2O. This work is representative of weakly coordinating anions such as Cl-, ClO4ꢁ,
and CF3SOꢁ3 , but perhaps not of the planar NOꢁ3 anion.
In contrast, Colette et al. [17] studied the complexation behavior of selected BTPs to
Eu3? using electrospray ionization mass spectrometry (ESI–MS) in 50 % v/v aqueous/
methanol solutions, and interpreted the results as indicating significant solvation by
methanol. The predominant species formed in the absence of BTP were [M(NO3)x
(MeOH)y(H2O)z]3-x where x = (1–2), y = (0–3) and z = (0–1). This is further exemplified
by the relative abundances of these mixed nitrate/methanol species formed at pH = 2.8 with
various lanthanides [18] at an approximate ionic strength of 0.016 molꢀL-1 in 50 % methanol
solution. The most abundant species formed for Nd(III), Sm(III) and Eu(III) were
[Ln(NO3)2MeOH]? [ [Ln(NO3)2(MeOH)2]? [ [Ln(NO3)2]? [ [Ln(NO3)2(MeOH)3]?.
It is unclear why the hypersensitive behavior of Nd3? is different in methanolic solution
from what is usually seen in aqueous media, with the peak between 720 and 770 nm
displaying the greater sensitivity to changes in the inner coordination sphere (4I9/2 ? F7/2
4
,
4
2
4S3/2), rather than the hypersensitive peak at 570–600 nm (4I9/2 ? G5/2, G7/2) which is
usually seen to respond in aqueous systems.
123

2148
J Solution Chem (2012) 41:2138–2153
This changing medium effect is more dramatically seen in the spectra of Am3?
(Fig. 3b). A number of distinct species (2–3) can be seen in the spectral pattern upon
changing from methanolic media to aqueous media as nitrates and (most probably)
methanol are coordinated in the inner sphere. The shape of the band is markedly different
from the aqueous peak and indicates that multiple Am3? species are present in the
methanol solutions. From these observations it is not possible to identify the several
species, though the coordination sphere clearly must be filled with some combination of
H2O, MeOH and NOꢁ3 . By changing the Am3? preparation from five to eight evaporations
with dry methanol (Fig. 3b and starting solution spectra in Fig. 5) the difference in the
shape of the Am3? absorption band around 511 nm most likely indicates an increasing
number of species with MeOH (or NOꢁ3 ) in the inner coordination sphere (Fig. 5). Based
on considerations of ion–ligand interaction strengths, it is reasonable to project that
methanol is the weakest donor. The formation of outer sphere Am3?–NO3ꢁ complexes in
aqueous media suggests that in methanol H2O and NOꢁ3 may well be of comparable donor
strength. These results do not allow any more profound conclusions to be drawn, though
the changes are clearly of interest.
4.2 Determination of Stability Constants
The spectrophotometric titration results were fitted using SQUAD (Table 1) with the initial
assumption that a 1:1 M:L complex was predominant in each system. For the ligands
BPMDAC and BPMDAP with Am3?, the formation of two spectroscopically distinct
species in solution is indicated in the spectrophotometric titration, one appearing at M:L up
to 1:1.5–1:3, the other dominating at higher concentrations of the ligand. At the low metal-
to-ligand ratios involved with the BPMDIE and BPMDIC complexation titration, the
spectrophotometric data was best fit using SQUAD by proposing the domination of a 2:1
(M2L) complex in solution with the ligand bridging two metal centers. For all the 2:1 M:L
species fitted, higher selectivity is observable in the diimine based complexant when
compared with the diamine. This could indicate an enhanced selectivity for Am3? with
aromatic/hybridized N-donors when compared to aliphatic N-donors. BPMDAE, BPM-
DAP and BPMDAC show a slight increase in actinide binding in the M2L complex when
compared with the Nd3? values although the degree of difference is lost within the error.
It can be seen in the log10 K101 values (Table 1) that there is an inherent selectivity for
the nitrogen ligands for Am3? over Nd3?. For all the N-donors we see an increased
selectivity for Am3? over Nd3? by a factor of approximately 3–11, which is comparable to
selectivities seen for N-donors in the literature [8]. As the bridging diamine changes from
1,2-diaminoethane to 1,2-diaminocyclohexane (BPMDAE versus BPMDAC), the stability
constant for Am3? increases by approximately a factor of 30. It is reasonable to attribute
the increase to the preorientation of the donor groups in BPMDAC, although the corre-
sponding increase in the Nd3? stability constants are not as pronounced. The implicit extra
enhancement for Am3? binding may indicate better ‘‘overlap’’ of orbitals and more cov-
alency in the Am3? complex.
The stability constants for BPMPIP are notably low; this may indicate that the piper-
azine backbone of BPMPIP arranges the ligand donor atoms in a configuration that is
poorly oriented for multidentate coordination to Ln3?/An3?. This result probably indicates
that the piperazine backbone is preferentially arranged in the chair conformer, while the
boat conformation would be required for utilization of the full dentation of the ligand
N-donor atoms. It may be assumed that the net energy gain that would be associated with
123

J Solution Chem (2012) 41:2138–2153
2149
the formation of the tetradentate BPMPIP complex is not adequate to overcome the
energetic barrier associated with transforming from chair-to-boat conformation of the
piperazine ring. The energy required to flip the piperazine from the boat to the chair
conformer in the tetradentate nitrogen donor ligand 1,4-bis(2-aminoethyl)-1,4-diazacy-
clohexane (BAE-P) has been calculated as approximately 29.3 kJꢀmol-1, which would
equate to a value in log10 K101 of 5.20 at 21 8C [19]. The BPMDAP stability constants are
lower than those for BPMDAC and comparable to those for BPMDAE, indicating that the
increased flexibility of the ligand backbone with the extra methylene in the bridge has
comparatively little impact on the complex strength of these large cations.
The apparent tendency of BPMDIE and BPMDIC for the M2L bonding arrangement is
by comparison with the amino-bridged ligands somewhat surprising. In the imines the
rigidity of the backbone appears to have very little effect on the stability of the 2:1 species
formed in solution. The apparent preference for M2L bonding is contrary to that reported
for lanthanide complexes with both BPMDIE and BPMDIC in the solid state, wherein
tetradentate coordination of the ligand with a metal center is seen [20–22]. It is likely that
the contrast between the solid state and solution phase arrangement of metals and ligands
arises from the collective effects of the general absence of metal-centered directed-valence
effects in the lanthanides, low degree of covalency in lanthanide-N bonding, and the
notably weak interaction between the f-elements and neutral N-donor atoms. As a
result of these influences, ligand exchange rates for lanthanide and actinide are typi-
cally rapid (for example, rate constants for water exchange have been reported in the
*108–107 molꢀL-1ꢀs-1 or faster) [23].
4.3 Ligand Selectivity
The log10 K101 values describing the complexes of the N4-tetradentate ligand are compared
with other heterocyclic N-donors from the prior literature in Table 2 (Structures given in
Figs. 1 and 2). To separate one metal from another in one contact with 99 % efficiency
requires a separation factor of 100, e.g. for Am3? D = 10 and for Eu3? D = 0.1 and hence
D1/D2 = 100 = SFAm/Eu. The energy difference required for a separation factor of 100 is
11.4 kJꢀmol-1 [4]. Assuming that the extraction mechanism is the same for both metal ion
species then the Gibbs energy change for such a separation can be approximated by the
difference in log10 K101 using Eq. 1:
À
Á
Am
101
Eu
101
ꢁRT log10
K
=K
ꢂ DðDGSFÞ
ð1Þ
The observed trend is that virtually all ligands that contain N-heterocyclic donor groups
show an inherent selectivity for Am3? over Ln3?. The better affinity of the soft nitrogen
ligands such as ADPTZ and BTPs for An(III) versus Ln(III) ions is commonly attributed to
the weakly basic nature of these ligands [8]. This results in the formation of stronger
actinide–nitrogen bonds with some covalent character. The higher binding affinity of
ADPTZ and BTP for Nd3? when compared with the N4-tetradentate ligands is directly due
to the effect of ligand preorientation with log10 K101 values of 4.62 (4) and 2.90 (5),
respectively [8]. Both of these ligands, even though they are tridentate, are coplanar and do
not undergo any great degree of rearrangement upon binding to a metal center which is
manifested as a favorable entropic contribution to DG for cation binding. Considering the
log10 K101 values in for Am3? in Table 2 it is seen that ADPTZ has the highest binding
strength. For separations applications, selectivity is often a more important feature of a
ligand than bonding strength. The selectivity of the ligand for An3? over Ln3? can best be
123

2150
J Solution Chem (2012) 41:2138–2153
Table 2 The stability constants for the binding N4-tetradentate ligands with Am3?, Eu3?, Sm3? and Nd3?
in MeOH
Ligand
Nd3? log10 K101
Am3? log10 K101
D log10 K101
ref
ADPTZa
Me-BTP
Terpy
4.62 0.04
2.90 0.05
5.8 0.1
1.2
[5]
[8]
[8]
[8]
[4]
3.4
TPTZ
3.5
6.73 0.06c
TPEN
Eu3? log10 K101
Sm3? log10 K101
TPTZ
Me-BTPb
Terpy
BBImP
TPEN
a
3.5
[8]
[18]
[8]
[8]
[4]
2.9 0.2
2.4
1.70 0.05
4.70 0.02
Determined in 75/25 % methanol/water solution at 25 °C at low ionic strength
b
c
Determined in 50/50 % methanol/water using ESI-MS
Value and error is the average and standard deviation of potentiometric and spectrophotometrc deter-
mined stability constants
judged by examining the difference between the stability constants for Nd3? and Am3?
(log10 K101). A general trend can be seen for the overall selectivity of the ligands (Table 2)
as follows: TPEN [ Phen [ ADPTZ [ BPMDIC [ BPMDAC = Terpy = BPMDAE [
BPMDIE = PTMP [ BPMDAP [[ BPMPIP [ TPTZ.
Choppin and Rizkalla [24] have reported previously on the correlation of the com-
plexation strength of Am3? and Nd3? complexes with acetate and aminopolycarboxylate
ligands. They observed an 8 % enhancement in the strength of bonding of Am in this
correlation and attributed the difference to the greater covalency in Am bonding. In Fig. 6
a similar correlation between stability constants of Am3? complexes and selected lan-
thanide ions with carboxylate, aminopolycarboxylate and pure N-donor ligands has been
developed. The equilibrium constants plotted have been assembled from the recent liter-
ature on N-donor ligands, most (but not all) obtained in alcoholic media, and the NIST
Stability Constants database [25]. The slope of log10 K101 (Am3?) relative to log10 K101
(Ln3?) gives a linear correlation in each case. The respective slopes are 0.98 ( 0.02) for
the hard base carboxylate species, 1.07 ( 0.03) for the (mixed donor) aminopolycarb-
oxylate system and 1.32 ( 0.08) for the nitrogen-only donor ligand systems. The
increasing slopes indicate greater intrinsic selectivity. The cost of this enhanced selectivity
in the pure-N-donor systems is the substantially weaker complexes that are formed, as
indicated by the shorter range of log10 K101 values in the nitrogen donor system. That the
effect is seen only in the aminopolycarboxylate and N-donor-only systems, but not in the
carboxylate systems, indicates that such soft-donor covalency effects are not important in
oxygen donor systems, but are observable for N-donors.
The calculated theoretical separation factors for all the ligands reported here are
comparable to experimentally determined group separation factors of Eu(III) and Am(III)
by BTP and hemi-BTP molecules from the prior literature [26]. The diimine ligands
BPMDIE and BPMDIC in their K201 values show similar selectivities to the K101 values for
123

J Solution Chem (2012) 41:2138–2153
2151
Fig. 6 Linear free energy relationship between actinide and lanthanide carboxylate (square box),
aminopolycarboxylate (circle), and N-donor ligands (diamond) (from data in Table 2). Values for
carboxylates and aminopolycarboxylates were taken at 0.1 molꢀL-1 ionic strength and 25 °C where possible
[25]
BPMDAC, BPMDAE and ADPTZ. Although these imine nitrogens are much more acidic
and softer than amine nitrogens, and thus show somewhat higher selectivity, they are prone
to hydrolysis. This decomposition is catalyzed by protons and hydroxide ions [27]. For this
reason, the imines are not useful as extracting agents.
5 Conclusion
In this study the stability constants for a series of N4 tetradentate ligands with Am3? and
Nd3? have been reported. For the ligands with bridging diamine groups, complex strength
increases in the order BPMDAC [ BPMDAE C BPMDAP [ BPMPIP. Constraining the
two amine groups into a conformation preoriented for binding (BPMDAC) results in an
increase in complex stability. Increasing the separation between the aliphatic amines
(BPMDAE, BPMDAP) has a comparatively small affect, while imposing a piperidine ring
between the pyridine donors significantly weakens both Nd3? and Am3? complexes. The
latter observation suggests that the piperidine ring retains the chair configuration and
the resulting complexes are at best bidentate. For the ligands with bridging diimine groups,
the complexes formed appear to preferentially involve simultaneous coordination of two
cations, presumably one each to an imine and pyridine nitrogen atom. For this model to be
correct, it is necessary that the bridging ethylene/cyclohexane groups orient the imine-
pyridine pairs in different directions from the eclipsed configuration implicit in Fig. 2. The
Am3?complexes with BPMDIC are stronger than the corresponding Am2BPMDIE6?
species. Within the accuracy of the fit parameters, the corresponding Nd3? complexes are
123

2152
J Solution Chem (2012) 41:2138–2153
of comparable stability. It appears that the cyclohexane bridge in the ligand backbone
increases the stability of the Am3? complexes with BPMDIC relative to BPMDIE. Of
future interest with the BPMDAE, BPMDAP and BPMDAC ligands is that the proton on
the amine can be substituted not just for aliphatic groups but with carboxylic acids, amides,
phosphonic acids, phosphinic acids and phosphine oxides to create a bank of compounds
for An(III) and Ln(III) complexation and separation.
Theoretical separation factors (comparative values of Nd and Am stability constants)
compare favorably with separation factors available from solvent extraction systems
available in the literature [8]. It is clear that ligands possessing favorably preoriented
ligand donor atoms offer increased complex stability, though it is not clear that they impact
Am/Nd selectivity. Figure 6 demonstrates that the greater is the number of nitrogen donors
engaged in cation binding, the higher is the separation potential. The indicated 32 %
preference for Am3? binding over Nd3? compares favorably with the 8 % enhancement
seen for aminopolycarboxylate ligands.
Acknowledgments This research was conducted at WSU and PNNL with funding provided by the U.S.
Department of Energy, Division of Nuclear Energy Science and Technology, Nuclear Energy Research
Initiative Consortium (NERI-C) program under projects number DE-FG07-07ID14896.
References
1. Rydberg, J., Choppin, G.R., Liljenzin, J.O.: Radiochemistry and Nuclear Chemistry, 3rd edn, pp. 1–687.
Butterworth-Heinemann, Massechusettes (2002)
2. Pearson, R.G.: Hard and soft acids and bases. J. Am. Chem. Soc. 85, 3533–3539 (1963)
3. Nash, K.L.: Separation chemistry for lanthanides and trivalent actinides. In: Gschneider Jr., K.A.,
Eyring, L., Choppin, G.R., Lander, G.H. (eds.) Handbook on the Physics and Chemistry of Rare Earths
Vol. 18–Lanthanides/Actinides Chemistry, pp. 197–238. Elsevier Science, Amsterdam (1994)
4. Jensen, M.P., Morss, L.R., Beitz, J.V., Ensor, D.D.: Aqueous complexation of trivalent lanthanide and
actinide cations by N,N,N’,N’-tetrakis(2-pyridylmethyl)ethylenediamine. J. Alloy. Compd. 303–304,
137–141 (2000)
5. Miguirditchian, M., Guillaneux, D., Guillaumont, D., Moisy, P., Madic, C., Jensen, M.P., Nash, K.L.:
Thermodynamic study of the complexation of trivalent actinide and lanthanide cations by ADPTZ, a
tridentate N-donor ligand. Inorg. Chem. 44, 1404–1412 (2005)
6. Melton, D.L., VanDerveer, D.G., Hancock, R.D.: Complexes of greatly enhanced thermodynamic
stability and metal ion size-based selectivity, formed by the highly preorganized non-macrocyclic ligand
1, 10-phenanthroline-2, 9-dicarboxylic acid. A thermodynamic and crystallographic study. Inorg. Chem.
45, 9306–9314 (2006)
7. Ekberg, C., Fermvik, A., Retegan, T., Skarnemark, G., Foreman, M.R.S., Hudson, M.J., Englund, S.,
Nilsson, M.: An overview and historical look back at the solvent extraction using nitrogen donor ligands
to extract and separate An(III) from Ln(III). Radiochim. Acta 96(4–5), 225–233 (2008)
8. Kolarik, Z.: Complexation and separation of lanthanides(III) and actinides(III) by heterocyclic N-donors
in solution. Chem. Rev. 108, 4208–4252 (2008)
9. Kolarik, Z., Mu¨llich, U., Gassner, F.: Extraction of Am(III) and Eu(III) nitrates by 2, 6-di(5, 6-dipropyl-
1, 2, 4-triazin-3-yl)pyridines. Solvent Extr. Ion Exch. 17, 1155–1170 (1999)
10. Ogden, M.D.: The effect of structural hindrance on the complexation of actinides and lanthanides by
ligands containing soft nitrogen donors. PhD Thesis, Washington State University (2009)
11. Newkome, G.R., Frere, Y.A., Fronczek, F.R., Gupta, V.K.: Chemistry of heterocyclic compounds. 104.
‘‘Obstacle effect’’ in palladium(II) complexes of tetraamine ligands with terminal pyridyl or picolyl
residues. Inorg. Chem. 24, 1001–1006 (1985)
12. Goodwin, T.J., Vagg, R.S., Williams, P.A.: Chiral metal complexes. 15. Alanine and proline complexes
of [N,N’-di(2-picolyl)-1R, 2R-diaminocyclohexane]cobalt(III). Proc. R. Soc. N. S. W 117, 1–6 (1984)
´
13. Schoumacker, S., Hamelin, O., Pecaut, J., Fontecave, M.: Catalytic asymmetric sulfoxidation by chiral
manganese complexes: Acetylacetonate anions as chirality switches. Inorg. Chem. 42, 8110–8116
(2003)
123

J Solution Chem (2012) 41:2138–2153
2153
14. Sinkov, S.I., Rapko, B.M., Lumetta, G.J., Hay, B.P.: Bicyclic and acyclic diamides: Comparison of their
aqueous phase binding constants with Nd(III), Am(III), Pu(IV), Np(V), Pu(VI), and U(VI). Inorg.
Chem. 43, 8404–8413 (2004)
15. Leggett, D.J.: SQUAD: stability quotients from absorbance data. In: Leggett, D.J. (ed.) Computational
Methods for the Determination of Formation Constants, Chap. 6. Plenum Press, New York (1985)
16. Tanaka, F., Kawasaki, Y., Yamashita, S.: Preferential solvation of the europium(III) ion in water–
nonaqueous solvent mixtures: a luminescence lifetime study. J. Chem. Soc., Faraday Trans. I 84,
1083–1090 (1988)
17. Colette, S., Amekraz, B., Madic, C., Berthon, L., Cote, G., Moulin, C.: Use of electrospray mass
spectrometry (ESI-MS) for the study of europium(III) complexation with bis(dialkyltriazinyl)pyridines
and its implications in the design of new extracting agents. Inorg. Chem. 41, 7031–7041 (2002)
18. Colette, S., Amekraz, B., Madic, C., Berthon, L., Cote, G., Moulin, C.: Trivalent lanthanide interactions
with a terdentate bis(dialkyltriazinyl)pyridine ligand studied by electrospray ionization mass spec-
trometry. Inorg. Chem. 42, 2215–2226 (2003)
19. Hancock, R.D., Ngwenya, M.P., Evers, A., Wade, P.W., Boeyens, J.C.A., Dobson, S.M.: Open-chain
polyamine ligands with more rigid double connecting bridges. Study of their metal ion selectivities by
molecular mechanics calculation, crystallography, and thermodynamics. Inorg. Chem. 29, 264–270
(1990)
20. Drew, M.G.B., Foreman, M.R.St.J., Hudson, M.J., Kennedy, K.F.: Structural studies of lanthanide
complexes with tetradentate nitrogen ligands. Inorg. Chim. Acta 357, 4102–4112 (2004)
21. Kano, S., Nakano, H., Kojima, M., Baba, N., Nakajima, K.: An effect of the ionic radii of lanthanide(III)
ions on the structure and catalytic properties of chiral Schiff base-lanthanide(III) complexes. Inorg.
Chim. Acta 349, 6–16 (2003)
22. Tsiouri, M., Hadjiliadis, N., Arslan, T., Kariuki, B.M., Plakatouras, J.C.: Interactions of trivalent lan-
thanide cations with tetradentate Schiff bases: New lanthanide(III) complexes from (1S, 2S, N1 E, N2 E)-
trans-N,N’-bis(pyridin-2-ylmethylene)cyclohexane-1, 2-diamine. Inorg. Chem. Commun. 9, 429–432
(2006)
23. Huheey, J.E.: Inorganic Chemistry: Principles of Structure and Reactivity, 3rd edn, pp. 548–549. Harper
& Row, New York (1983)
24. Choppin, G.R., Rizkalla, E.N.: Solution chemistry of actinides and lanthanides. In: Gschneider Jr., K.A.,
Eyring, L., Choppin, G.R., Lander, G.H. (eds.) Handbook on the Physics and Chemistry of Rare Earths.
Vol. 18–Lanthanides/Actinides Chemistry, pp. 559–590. Elsevier Science, Amsterdam (1994)
25. Martell, A.E., Smith, R.S., Motekaitis, R.J.: NIST critically selected stability constants of metal
complexes. NIST Standard Reference Database 46 Version 8.0, Maryland (2004)
26. Drew, M.G.B., Hudson, M.J., Youngs, T.G.A.: QSAR studies of multidentate nitrogen ligands used in
lanthanide and actinide extraction processes. J. Alloy. Compd. 374, 408–415 (2004)
27. Reeves, R.L.: On the mechanism, substituent effects, and intramolecular catalysis in Schiff base
hydrolysis. J. Org. Chem. 30, 3129–3135 (1965)
123