Thermodynamic studies of U(vi) complexation with glutardiamidoxime for sequestration of uranium from seawater

Article abstract of DOI:10.1039/c3dt32940b

Glutardiamidoxime (H₂B), a diamidoxime ligand that has implications in sequestering uranium from seawater, forms strong complexes with UO₂²⁺. Five U(vi) complexes were identified in 3% NaCl solution. The stability constants and the enthalpies of complexation were measured by potentiometry and microcalorimetry. The competition between glutardiamidoxime and carbonate for complexing U(vi) in 3% NaCl was also studied in comparison with the cyclic glutarimidedioxime ligand (H₂A) previously studied.

Full text of DOI:10.1039/c3dt32940b

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Thermodynamic studies of U(VI) complexation with
glutardiamidoxime for sequestration of uranium from
seawater†
Cite this: Dalton Trans., 2013, 42, 5690
a
b
a
Guoxin Tian, Simon J. Teat and Linfeng Rao*
Glutardiamidoxime (H
2
B), a diamidoxime ligand that has implications in sequestering uranium from sea-
Received 7th December 2012,
Accepted 1st February 2013
2+
water, forms strong complexes with UO
2
. Five U(VI) complexes were identified in 3% NaCl solution. The
stability constants and the enthalpies of complexation were measured by potentiometry and microcalori-
metry. The competition between glutardiamidoxime and carbonate for complexing U(VI) in 3% NaCl was
DOI: 10.1039/c3dt32940b
www.rsc.org/dalton
2
also studied in comparison with the cyclic glutarimidedioxime ligand (H A) previously studied.
sorbent materials were used in the marine tests in 1999–2001,
Introduction
6
–8
and about 1 kg uranium was collected.
Though the amidoxime-fuctionalized materials have been
The ocean is an important source of uranium because it con-
4
–8
tains about 4.5 billion tons of uranium, a thousand times as shown to be capable of extracting uranium from seawater,
1
,2
much as the amount of uranium in terrestrial ores.
the configurations of the functional groups on the materials
However, it is extremely difficult to extract uranium from sea- were not identified and the coordination modes in the com-
water, because uranium compounds exist in extremely low con- plexes with uranium are not well understood. Based on the
2
centrations as very stable carbonate complexes, and because knowledge of the reaction between acrylonitrile and hydroxyl-
the ocean also contains many other metal ions (Na, K, Ca, Mg, amine, we have hypothesized that at least two configurations,
Al, and transition metals), some of which are in overwhel- a cyclic imide dioxime and an open-chain diamidoxime, could
mingly higher concentrations. To make the extraction process form if the conditions of the grafting process were not well-
economically competitive, the extracting agents must be highly controlled, and that the yields of the two configurations would
3
efficient and selective, robust and inexpensive.
affect the efficiency of the sorbents because the cyclic imide
Screening studies in 1980s showed that, among more than dioxime and the open-chain diamidoxime may have different
2
00 functionalized adsorbents, the materials with the amidox- binding strengths towards uranium. Quantification and com-
2
ime group, RC(NH )(NOH), were highly selective towards parison of the binding strengths of the cyclic imide dioxime
4
,5
uranium. Based on these studies, a radiation-induced graft- and the open-chain diamidoxime towards uranium could help
ing process was developed in Japan to functionalize a non- to optimize the grafting process and improve the efficiency of
woven polyethylene fabric with amidoxime groups. In this the extraction of uranium.
process, the irradiated polyethylene fabric was grafted with
To quantify and compare the binding strengths of different
acrylonitrile (CH vCHCN) and then reacted with hydroxyl- functional groups towards uranium, two small molecular
2
amine (NH
2
OH), resulting in the sorbents functionalized with ligands, glutarimidedioxime (H
2
A) and glutardiamidoxime
amidoxime (or related moieties) that could form strong (H
2
B), were synthesized and used as the water-soluble surro-
complex(es) with uranium under seawater conditions. The gates of the cyclic imide dioxime and the open-chain diami-
doxime on the sorbent (Scheme 1). The results for the cyclic
ligand (H
from this group. The present study is focused on the com-
plexation of uranium with the diamidoxime ligand (H B) in
comparison with H A. Conditions for synthesizing the ligands
were optimized to obtain high yields of each. The binding
2
A) have been summarized in a recent publication
9
a
2
Chemical Sciences Division, Lawrence Berkeley National Laboratory, 1 Cyclotron
Rd., Berkeley, California 94720, USA. E-mail: lrao@lbl.gov; Fax: +01 5104865596;
Tel: +01 5104865427
2
b
Advanced Light Source, Lawrence Berkeley National Laboratory, 1 Cyclotron Rd.,
2+
strength with UO
2
and the enthalpy of complexation were
Berkeley, California 94720, USA
†
investigated by potentiometry, spectrophotometry and micro-
calorimetry. The most probable coordination modes in the
Electronic supplementary information (ESI) available: Crystal and refinement
data, and the CIF file for glutardiamidoxime. CCDC 913296. For ESI and crystal-
lographic data in CIF or other electronic format see DOI: 10.1039/c3dt32940b
uranyl/H
2
B
complexes were discussed based on the
5
690 | Dalton Trans., 2013, 42, 5690–5696
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
2
+
3 2 H
B , H for the total
B, HB , and B2−), and acidity (C
+
−
H
4
B , H
hydrogen ion, where –CH = COH). For determining the protona-
tion constant of the ligand, 20 mL of the ligand solution (C
B
=
0
.01 to 0.02 M; C
H
= (−0.02) to (−0.04) M) was titrated with
1.0 M HCl. 50–100 data points were collected in each titration.
For determining the stability constants of the U(VI) complexes,
20 mL of U(VI)/H
2 U H
B solutions (C = 0.20–0.50 mM; C =
2.0–4.0 mM: CB = 1.0–2.0 mM) were titrated with 0.100 M
NaOH. About 40–50 data points were collected for each titra-
2
Scheme 1 Preparation routes for glutarimidedioxime (H A, upper) and glutar-
1
5
tion. A nonlinear regression program Hyperquad 2008 was
used to calculate the protonation constants of the ligand and
the stability constants of U(VI) complexes.
2
diamidoxime (H B, lower).
thermodynamic data and the crystal structures of related com-
plexes. Besides, spectrophotometric experiments were con- Spectrophotometry
ducted to evaluate the ability of glutardiamidoxime to compete
with carbonate for binding UO2 under seawater conditions.
Two types of spectrophotometric titrations were performed
using a Cary 6000i spectrophotometer (Varian Inc.) from 500
to 200 nm with an interval of 1.0 nm: (1) addition of a
buffered ligand solution to the U(VI) solution; (2) addition of
HCl to the solution containing both U(VI) and the ligand.
Usually, 15–20 additions were made in each titration, generat-
ing a set of 16–21 spectra for calculation.
2
+
Experimental
Chemicals
All experiments were conducted at 25 °C and an ionic strength
of 0.5 M (NaCl), close to the seawater condition of 3% NaCl.
Hydroxylamine (50 wt% solution in water) and glutaronitrile
Microcalorimetry
An isothermal microcalorimeter (Model: ITC 4200, Calorimetry
Sciences Corp.) was used to determine the enthalpy of the reac-
tions. The calorimeter was calibrated with the same pro-
(
99%) from Sigma-Aldrich were used as received. Boiled/cooled
Milli-Q water was used in preparation of all solutions. The
stock solution of U(VI) was prepared by dissolving UO in HCl
and standardized by fluorimetry using standard solutions of
3
1
6
cedures in the literature. Multiple titrations with different
concentrations of U(VI), ligand and acidity were performed to
reduce the uncertainty of the results. For the protonation of
the ligand, 0.9 mL solution containing the ligand was placed
in the reaction cell and titrated with 0.1 M HCl. For the com-
plexation of U(VI) with the ligand, 0.9 mL solution containing
1
0
+
3 4
U(VI) in 1 M H PO . The concentration of free H in the U(VI)
stock solution was determined by the Gran titration.
1
1
1
2–14
The same procedure
that was used to prepare the cyclic
9
glutarimidedioxime (H
glutarimidedioxime (H
2
A) was used to prepare the open chain
2
B), except that the reactions were con-
+
U(VI), the ligand and H was titrated with a solution of NaOH.
ducted at different temperatures (Scheme 1). Using the same
starting materials at the molar ratio of 1 : 2 (glutaronitrile and
A titration usually contains 40–50 additions (0.005 mL each) of
the titrant to the reaction vessel. The observed total heat
(Qex,j, j = 1 to n, n = 40–50) values were corrected for the heats
of titrant dilution (Qdil,j) that were measured in a separate run.
hydroxylamine), the cyclic glutarimidedioxime (H
2
A) was
obtained at 80–90 °C while the open chain glutardiamidox-
ime (H B) was obtained at room temperature. To obtain H
9
2
2
B
The net reaction heat at the jth point (Q ) was obtained from
r,j
in high yields, 9.4 g glutaronitrile (99%) and 14.5 g hydroxyl-
the difference: Qr,j = Qex,j − Qdil,j. The value of Qr,j is a function
amine (50% in H O) were dissolved in 200 mL of 1/1 (V/V)
2
U B H
of the concentrations of the reactants (C , C , and C ), the
ethanol/water and reacted at room temperature with stirring
equilibrium constants, and the enthalpies of the reactions that
for 5 days, resulting in H
yield.
2
B as colorless crystals with >90%
1
7
occurred in the titration. The computer program HypDeltaH
was used to calculate the enthalpy of ligand protonation and
complexation with U(VI). In the calculation, the protonation
constants and the stability constants of U(VI) complexes
obtained by potentiometry were used.
1
Ligand H
CH –CH –CH
.51 ppm, 4H; –C(NOH)NH , 5.98 ppm, 4H; –CH –C(NH )-
2
B was characterized by H-NMR (pyridine-d
5
):
–
2
2
2 2 2 2
–, 2.19 ppm, 2H; –CH –CH –C(NOH)NH ,
2
2
2
2
NOH, 10.97 ppm, 2H. The purity was determined to be >99.5%
with potentiometry by titrating the H B solution with standard
NaOH. The crystal structure of H B·H O was also obtained
2
2
2
Results and discussion
(see ESI†).
Protonation of glutardiamidoxime
Potentiometry
A typical potentiometric titration for the protonation of glutar-
The procedures for potentiometric titrations have been diamidoxime (H B) is shown in Fig. 1. Differing from the pro-
2
9
described elsewhere. Multiple titrations were conducted with tonation titration curve for the ligand H
solutions of different concentrations of U(VI) (C as total steps of protonation are involved, the titration curve in Fig. 1
U(VI)]), ligand (C as the total ligand concentration including can be fitted with four steps of protonation, from B , through
2
A system where three
9
U
2
−
[
B
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 5690–5696 | 5691

View Article Online
Paper
Dalton Transactions
0
Fig. 1 Protonation titration of glutardiamidoxime in 3% NaCl. V = 20 mL,
0
0
Fig. 2 Potentiometric titration for the complexation of H
2
B with U(VI) in 3%
C
B
H
= 0.0181 M, C = −0.0127 M. Titrant: 1 M HCl.
0
0
0
0
NaCl. V = 20 mL, C
.106 M NaOH.
= 0.0553 mM, C = 0.535 mM, C = 0.011 mM. Titrant:
U B H
0
−
+
2+
HB , H
2
B, H
3
B , to H
4
B . The calculated protonation con-
stants are listed in Table 1. It is interesting to compare the
stepwise protonation constants of the two ligands (H B and
) is typical
and the second protonation
) is essentially the same as the first one, indi-
cating that the two oxime groups in H B are independent and
in Fig. 2. The data were best fitted with the model including
2
2+
the formation of five U(VI) complexes, UO
2
B, UO
2
H
2
B ,
1
2.13
H
2
A). For H
2
B, the first protonation constant (10
2−
−
2+
UO2B , UO HB and UO H B , shown by eqn (1):
1
8,19
2
2
2
2
4
2
for the oxime group (–NOH)
constant (10
1
2.06
UO22 þ mH þ nB ¼ UO2HmBnð2nꢀmꢀ2Þꢀ
þ
þ
2ꢀ
ð1Þ
2
where (m,n) = (0,1), (2,1), (0,2), (1,2) and (4,2). Table 1 shows
the calculated stability constants for the five complexes.
the protonation or deprotonation of one group has little effect
on the basicity of the other group. In contrast, the protonation
1
2.06
The complexation of U(VI) with glutardiamidoxime was also
studied with spectrophotometric experiments. The spectra of
the ligand under different conditions are compared in Fig. 3a.
Spectra 1, 2 and 3 represent three solutions of 0.1 mM glutar-
diamidoxime in the absence of U(VI) at different acidities.
constants of the two oxime groups in H
2
A are 10
and
1
0.7
1
0
, respectively, suggesting that the protonation of one
oxime group reduces the basicity of the other oxime group. In
addition, the third and fourth stepwise protonation constants
of H B (10 and 10 for the amide nitrogen) are significantly
2
higher than that for the imide nitrogen of H A (10 ).
9
5
.8
4.8
2
2
.1
9
Based on the protonation constants of H B (Table 1), the
2
2
+
2−
4 2
major species in the three solutions are H B , H B, and B ,
Obviously, the basicity of the amide nitrogen in H B is much
2
respectively. Obviously, the absorption spectra of the ligand
species with various degrees of protonation differ from each
other, but all their absorption bands are in the wavelengths
below 300 nm. In the presence of 0.05 mM U(VI) (spectra 4 in
higher than that of the imide nitrogen in H A.
2
Complexation of U(VI) with glutardiamidoxime
Stability constants. A representative potentiometric titration Fig. 3a), two new bands associated with the complexation with
of the complexation of U(VI) with glutardiamidoxime is shown U(VI) appear at about 300 nm and 350 nm.
Table 1 Thermodynamic parameters of the protonation and complexation of glutardiamidoxime at 25 °C and 0.5 M ionic strength (NaCl)
ΔH (kJ mol−
1
)
ΔS (J K−1 mol−1
)
Ligand
Reaction
log β
+
2−
Glutardiamidoxime (H
2
B)
H + B = HB−
12.13 ± 0.12
24.19 ± 0.07
29.98 ± 0.07
34.77 ± 0.07
17.3 ± 0.3
29.2 ± 0.3
26.1 ± 0.3
36.4 ± 0.3
56.3 ± 1.0
12.06 ± 0.23
22.76 ± 0.31
24.88 ± 0.35
17.8 ± 1.1
−52 ± 2
−103 ± 3
−124 ± 6
−151 ± 8
−49 ± 6
−102 ± 6
−123 ± 7
−133 ± 8
−207 ± 16
−36.1 ± 0.5
−69.7 ± 0.9
−77 ± 6
−59 ± 8
−71 ± 6
−101 ± 10
−118 ± 6
−154 ± 25
58 ± 7
+
2−
2
3
4
H + B = H
H + B = H
H + B = H
UO
2
3
4
B
B
B
117 ± 10
158 ± 20
159 ± 26
167 ± 19
217 ± 21
88 ± 23
251 ± 27
384 ± 51
110 ± 2
+
2−
+
+
2−
2+
2+
+ B2− = UO
2
2
B
+
2+
2−
2+
2
H + UO
2
+ B = UO
2
(H
2
B)
2
+
2−
2−
UO
2
+ 2B = UO
2
2
B
2
+
2+
2−
(HB)B−
H + UO
+ 2B = UO
2
+
2+
2
2−
2+
4
H + UO
+ 2B = UO
2
(H
2
B)
2
a
+
2−
−
Glutarimidedioxime (H
2
A)
H + A = HA
+
2−
2
3
H + A = H
H + A = H
2
A
A
202 ± 3
+
2−
+
3
218 ± 14
142 ± 19
197 ± 14
188 ± 24
309 ± 14
307 ± 59
2
+
2−
UO
2
+ A = UO
2
+ 2A = UO
2
2
A
+
2+
2−
+
H + UO
+ A = UO (HA)
2
22.7 ± 1.3
27.5 ± 2.3
36.8 ± 2.1
43.0 ± 1.1
2
+
2−
2−
UO
2
2
A
2
+
2+
2−
(HA)A−
H + UO
+ 2A = UO
+ 2A = UO (HA)
2
+
2+
2
2−
2
H + UO
2
2
a
Data from ref. 9.
5
692 | Dalton Trans., 2013, 42, 5690–5696
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 3c shows a spectrophotometric titration in which a solu-
tion of U(VI) was titrated with Na B. The molar ratio of the
2
B U
ligand to uranium (C /C ) was increased from 0 to 2.6 in the
titration. The spectra changes can be discussed in two
different phases: (1) In the early part of the titration, as shown
B U
by the first 6 spectra where C /C = 0–2, the two absorption
bands at 300 nm and 350 nm appeared and became more
intense, indicating the formation of the U(VI)/H B complexes.
2
In the meantime, the absorption bands at shorter wavelengths
were intensified because the concentration of the uncom-
plexed ligand was also increased along the titration. (2) In the
B U
late part of the titration where C /C > 2, as more ligands were
added, the intensities of the 300 and 350 nm bands increased
only slightly while the intensities of the bands at shorter wave-
lengths continued to increase significantly. The observation
suggested that, under the experimental conditions, the
1
: 2 metal/ligand complexes were the limiting species of the
U(VI)/H B complexes and the complexes were considerably
strong. At C /C = 2, most of U(VI) was complexed with the
ligand. Further additions of Na B only slightly increased the
2
B
U
2
2
amounts of U(VI)/H B complexes, but significantly elevated the
concentration of the free ligand. As a result, the spectra in
the late part of the titration were nearly unchanged at
300–350 nm, but significantly intensified at shorter
wavelengths.
Efforts were made to calculate the stability constants of
U(VI)/H B complexes or validate the stability constants obtained
2
by potentiometry with the spectrophotometric data, but were
unsuccessful. The reason for the failure probably lies in the
fact that we were monitoring the absorption spectra of the
ligand. The absorption spectra of the ligand in different U(VI)
2
+
2+
−
complexes, including UO H B , UO H B , UO B, UO HB
2
2
4
2
2
2
2
2
2
−
and UO
2
B
2
, could be too similar to be distinguished from
each other.
Enthalpy of complexation. A representative calorimetric
titration of U(VI) complexation with glutardiamidoxime is
shown in Fig. 4. The enthalpies, calculated from the calori-
metric data, and the entropies of complexation calculated
accordingly are summarized in Table 1.
2
Fig. 3 Spectrophotometry for the complexation of U(VI) with H B in 3% NaCl.
(
a) 1 - 0.1 mM Na
2
B/0.01 M HCl; 2 - 0.1 mM H
2
B; 3 - 0.1 mM Na
2
B/0.01 M
0
0
0
NaOH; 4 - 0.1 mM Na
2
B/0.05 mM U(VI). (b) V = 2 mL, C
U
= 0.05 mM, C
H
=
=
0
0
0
0
0
mM, C
B
= 1 mM; titrant: 0.01 M HCl (Vadded = 1.505 mL). (c) V = 2 mL, C
U
0
.025 mM, C
H
= 0.027 mM; titrant: 1 mM Na
2
B (Vadded = 0.13 mL). Spectra in
(b) and (c) are normalized to the initial volume.
Coordination modes in the complexes. Attempts to prepare
crystals of the U(VI)/H B complexes of good quality have not
2
been successful. The following discussions on the probable
Fig. 3b shows a spectrophotometric titration in which a solu- coordination modes are based on the reported crystal struc-
tion of U(VI) and neutralized H B was titrated with HCl. tures of U(VI) complexes with other amidoxime-related ligands
Initially, the two absorption bands at 300 nm and 350 nm were and the thermodynamic parameters of the complexation
observed, showing the presence of U(VI)/H B complexes in the (Table 1). The thermodynamic parameters help to form
2
2
beginning of the titration. As the acid was added, the intensity hypothesis on where the protonation occurs in the complexes
of the two bands decreased, indicating the dissociation of the and what the most probable coordination modes are.
+
9,20–22
21
U(VI)/H
2
B complexes due to the competition of H with U(VI)
Single-crystal structures
and computations
have
2
+
for the ligand. In addition, the intensity of the bands at shown that amidoxime-related ligands could bind UO2 with
shorter wavelengths (<300 nm) also decreases as the acidity three coordination modes: monodentate coordination with the
was increased, suggesting a higher degree of protonation of oxygen of the oxime group without the participation of the
the ligand along the titration. The last spectra in the titration amide group in bonding,
Fig. 3b) were very similar to spectra 1 shown in Fig. 3a, corres- bond,
2
0
2
η
coordination with the N–O
2
1,22
(
and multi-dentate coordination with the oxygen of
9
ponding to the solution of 0.1 mM Na
2
B/0.01 M HCl with the oxime group and the nitrogen of the imide group. The
first two coordination modes were observed in the complexes
2
+
H
4
B
as the major ligand species.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 5690–5696 | 5693

View Article Online
Paper
Dalton Transactions
2 2 2 2 2
Fig. 5 Hypothesized coordination modes. (a) UO B, (b) UO (H B) , (c) UO B
2−,
2
+
−
2+
(
2 2 2 4 2
d) UO HB , (e) UO H B .
2
+
that the probable coordination mode in UO
shown in Fig. 5b.
2 2
(H B) is as
Following the same analysis of the thermodynamic para-
meters, the equilibrium constants of the following protona-
9
.7
30.2
tion reactions could be estimated to be 10 and 10
eqn (3) and (4), respectively.
for
Fig. 4 Calorimetric titration of the complexation of U(VI) with H
2
B in 3% NaCl
þ
2ꢀ
¼ UO2HB2ꢀ
¼ UO2H4B22þ
H þ UO2B2
ð3Þ
ð4Þ
0
0
0
0
at 25 °C. V = 0.9 mL, C
U B H
= 0.246 mM, C = 0.97 mM, C = 0.257 mM; titrant:
0
.01 M HCl (5.0 μL per addition). (upper) thermogram; (lower) total heat (right
þ
2ꢀ
4
H þ UO2B2
y axis) and speciation of U(VI) (left y axis) vs. the titrant volume.
Again, these protonation constants are not unreasonable
for the protonation of the nitrogen of the oxime group. There-
−
fore, the probable coordination modes in UO
2
HB
2
and
that crystallized from non-aqueous solutions such as nitro-
2
+
2
0
21
UO H B could be hypothesized as shown in Fig. 5d and 5e.
2 4 2
Future efforts will be made to identify the structures so that
methane, mixtures of methanol/nitromethane or a hydro-
2
2
phobic ionic liquid.
The third coordination mode was
the above hypothesis can be validated or disputed.
observed in the complexes that crystallized from aqueous solu-
2
21,22
9
20
It should be pointed out that η coordination
could also
tions. In either the monodentate mode or the multi-
9
occur in these complexes, especially if the complexes form in
non-aqueous media. However, the structures obtained in non-
aqueous media, though highly informative, are less relevant to
the issue of U(VI) complexation in seawater.
dentate mode, an intra-molecular proton transfer occurs
within the oxime group (–NOH) from the oxygen to the nitro-
9
,20
gen in the absence of a strong base in the solution.
2
−
In the UO
2
B and UO
2
B
2
complexes, the ligand is fully
deprotonated. The most probable coordination modes for
these two complexes in aqueous solutions are shown in Fig. 5a
and 5c, where the deprotonation occurs on the two oxime
The ability of glutardiamidoxime to compete with carbonate
for the sequestration of U(VI) under seawater conditions
1
7.3
2−
groups. The stability constants of UO B (10 ) and UO2B2
The predominant species of uranium is the stable tricarbonato
2
2
6.1
4−
(
(
10 ) are very reasonable values for successive 1 : 1 and 1 : 2 U(VI) complex, UO (CO ) , under the seawater conditions
2 3 3
metal/ligand) complexes (Table 1).
The stability constant of the protonated UO H B complex sequestering agent must be able to compete with carbonate for
(pH ∼ 8.3, 2.3 mM total carbonate). As a result, an effective
2
+
2
2
29.2
is 10
(Table 1). Combining this with the stability constant complexing U(VI) at seawater pH. Using the stability constants
1
7.3
2
of UO B (10 ), we can estimate the equilibrium constant for of U(VI) complexes with glutardiamidoxime from this work and
the “double” protonation of UO B (eqn (2)) to be 10 , a with carbonate from the literature, the speciation of U(VI)
1
1.9
23
2
reasonable value for the two-step protonation of the nitrogen under seawater conditions (C = 3.3 ppb, Ccarbonate = 0.0023 M)
U
atom in the oxime group.
is calculated and shown in Fig. 6. The speciation was calcu-
lated at two ligand concentrations: C = 0.001 M and 0.01 M.
B
þ
2þ
2
H þ UO2B ¼ UO2ðH2BÞ
ð2Þ
In the presence of 0.001 M glutardiamidoxime (Fig. 6, upper),
nearly all U(VI) is complexed by carbonate at pH 8.3 (98.6%
4−
2−
Based on this analysis and the observed intra-molecular UO
(CO )
2 3 3
2 3 2
and 1.1% UO (CO ) ). Only when pH is above 11,
proton transfer from the oxygen to the nitrogen within the glutardiamidoxime starts to be competitive for complexing
9
,20
oxime group (–NOH) in crystal structures,
we hypothesize uranium. In the presence of 0.01 M glutardiamidoxime (Fig. 6,
5
694 | Dalton Trans., 2013, 42, 5690–5696
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 7 Absorption spectra showing the competition of H
right) with carbonate for complexing uranium. C = 0.05 mM. (left) C
.10 mM, Ccarbonate/C = 0–25; (right) C = 0.10 mM, Ccarbonate/C
for the left figure are from ref. 9.
2
A (left) and H
2
B
=
(
U
A
0
B
B
B
= 0–5. Data
as low as 2.5. Obviously, glutardiamidoxime (H B) is a much
2
weaker competing ligand with carbonate for complexing U(VI)
than glutarimidedioxime (H A) under these conditions.
2
Fig. 6 Speciation of U(VI). C
U
= 3.3 ppb, Ccarbonate = 0.0023 M, C
B
= 0.001 M
2+
2−, 4 – UO
−
(
upper) and 0.01 M (lower). 1 – UO
2
B, 2 – UO
2
H
2
B
, 3 – UO
2
B
2
4−.
HB ,
2 2
2+
2−
5
– UO
2
H
4
B
2
, 6 – UO
2
(CO
3
), 7 – UO
2
(CO
3
)
2
, 8 – UO (CO )
2
3 3
Summary
lower), about 77% U(VI) is still complexed by carbonate at pH 8.3
4−
2−
(
76% UO (CO ) and 1% UO (CO ) ). Only about 22% U(VI) is
2 3 3 2 3 2
Glutardiamidoxime (H B) was synthesized and studied as the
small water-soluble molecular surrogate for the amidoxime-
based sorbents that have been used for the sequestration of
uranium from seawater. Glutardiamidoxime (H B) forms fairly
strong complexes with U(VI), but it cannot effectively compete
2
−
complexed by glutardiamidoxime in the form of UO HB
In contrast, a previous study on the related ligand, glutari-
midedioxime (H A), indicates that, even in the presence of the
.
2
2
9
2
2
lower concentration (0.001 M H A), U(VI) is dominantly com-
2
plexed by the glutarimidedioxime (>80%) and only minor
amounts of U(VI) are in the carbonato complexes at pH 8.3.
with carbonate for complexing U(VI) under seawater pH. In
contrast, glutarimidedioxime (H A), a previously studied cyclic
2
This means that, though glutardiamidoxime (H B) and glutari-
2
ligand related to glutardiamidoxime (H B), was shown to be a
2
2
midedioxime (H A) form U(VI) complexes with comparable
strong complexing ligand that can effectively compete with car-
bonate for U(VI) under seawater pH. On the basis of these
findings, we plan to conduct further studies on the thermo-
dynamics of the complexation of H A and H B with transition
2 2
metals (Fe, Cu, Pb, Ni, V) to evaluate the competition between
U(VI) and transition metals.
In the amidoxime-based sorbents that are prepared by the
current radiation-induced grafting/reaction process, various
configurations of the functional groups could exist, including
strength (Table 1), the former (H B) is a much weaker compet-
2
ing ligand with carbonate than the latter (H A) for complexing
2
U(VI) at seawater pH. The reason for the weaker competing
ability of H B with carbonate at pH 8.3 lies in the fact that H B
2
2
has higher tendency for protonation (higher overall pK values)
a
2 2
than H A. At pH higher than 8.3 when more H B is deproto-
nated, the ability of H B to compete with carbonate for com-
2
plexing U(VI) is expected to be stronger. In fact, the speciation
diagrams in Fig. 6 show that nearly 100% U(VI) would be com-
2
those represented by the cyclic glutarimidedioxime (H A) and
plexed by H B at pH 11 when C = 0.001 M, and at pH 10.5
2
B
“open” glutardiamidoxime (H B). Combining the data from
2
when C
B
= 0.01 M.
9
the present and previous work, we have demonstrated that
The competition between glutardiamidoxime (H
bonate can be further illustrated by the optical absorption
spectra of U(VI) in the presence of glutardiamidoxime (H B) and
carbonate (Fig. 7, right). The spectra showing the competition
2
B) and car-
2
the cyclic glutarimidedioxime (H A) is the preferred configur-
ation for sequestration of uranium from seawater. Results
suggest that, to improve the efficiency of the sequestration of
uranium from seawater, the grafting/reaction process should
be conducted at 80–90 °C to achieve higher yield of the cyclic
2
between glutarimidedioxime (H A) and carbonate from a pre-
2
9
vious study are also shown in Fig. 7 (left) for comparison. At
2
imidedioxime (H A).
A B
the same concentration of the ligands (C = C = 0.1 mM), the
left figure shows that, when the ratio of Ccarbonate/C is as high
A
as 10, the intensity of the band for the U(VI)/H
∼290 nm) is still significant. In contrast, the right figure shows
that the bands for the U(VI)/H B complexes (∼300 nm and This work was supported by the Uranium Resources Program,
2
A complexes
Acknowledgements
(
2
3
50 nm) are barely identifiable when the ratio of Ccarbonate/C
B
is Fuel Cycle Research and Development Program, Office of
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 5690–5696 | 5695

View Article Online
Paper
Dalton Transactions
Nuclear Energy of the U.S. Department of Energy (DOE) under 11 G. Gran, Analyst, 1952, 77, 661.
Contract No. DE-AC02-05CH11231 at Lawrence Berkeley 12 F. Eloy and R. Lenaers, Chem. Rev., 1962, 62, 155.
National Laboratory (LBNL). Single-crystal X-ray diffraction 13 A. R. Forrester, H. Irikawa, R. H. Thomson, S. O. Woo and
data were collected and analyzed at the Advanced Light Source
T. J. King, J. Chem. Soc., Perkin Trans. 1, 1981, 1712.
(ALS). ALS is supported by the Director, Office of Science, 14 J. A. Elvidge, R. P. Linstead and A. M. Salaman, J. Chem.
Office of Basic Energy Sciences, U.S. DOE under Contract
Soc., 1959, 208.
No. DE-AC02-05CH11231.
15 P. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 1739.
16 L. Rao, T. G. Srinivasan, A. Yu. Garnov, P. Zanonato, P. Di
Bernardo and A. Bismondo, Geochim. Cosmochim. Acta,
Notes and references
2004, 68, 4821.
1
OECD, Uranium 2009: Resources, Production and Demand, 17 P. Gans, A. Sabatini and A. Vacca, J. Solution Chem., 2008,
OECD NEA Publication 6891, 2010, 456. 37, 467.
R. V. Davies, J. Kennedy, R. W. McIlroy, R. Spence and 18 N. Durust, M. A. Akay, Y. Durust and E. Kilic, Anal. Sci.,
2
K. M. Hill, Nature, 1964, 203, 1110.
2000, 16, 825.
3
4
K. Schwochau, Top. Curr. Chem., 1984, 124, 91.
H. J. Schenk, L. Astheimer, E. G. Witte and K. Schwochau,
Sep. Sci. Technol., 1982, 17, 1293.
L. Astheimer, H. J. Schenk, E. G. Witte and K. Schwochau,
Sep. Sci. Technol., 1983, 18, 307.
19 A. E. Martell, R. M. Smith and R. J. Motekaitis, NIST Criti-
cally Selected Stability Constants of Metal Complexes Data
Base. NIST Stand. Ref. Database 46, U. S. Department of
Commerce, Gaithersburg, MD, 1998.
20 E. G. Witte, K. S. Schwochau, G. Henkel and B. Krebs,
Inorg. Chim. Acta, 1984, 94, 323.
5
6
7
M. Kanno, J. Nucl. Sci. Technol., 1984, 21, 1.
T. Shimizu and M. Tamada, PROCEEDINGS OF CIVIL 21 S. Vukovic, L. A. Watson, S. O. Kang, R. Custelcean and
ENGINEERING IN THE OCEAN, 2004, 20, 617. B. P. Hay, Inorg. Chem., 2012, 51, 3855.
M. Tamada, N. Seko, N. Kasai and T. Shimizu, Trans. Atom. 22 P. Barber, S. Kelley and R. Rogers, RSC Adv., 2012, 2, 8526.
Energy Soc. Jpn., 2006, 5, 358.
G. Tian, S. Teat, Z. Zhang and L. Rao, Dalton Trans., 2012,
8
9
23 I. Grenthe, J. Fuger, R. J. Konings, R. J. Lemire,
A. B. Muller, C. Nguyen-Trung and H. Wanner, Chemical
thermodynamics of uranium, ed. H. Wanner and I. Forest,
Elsevier Science Publishers B.V., Amsterdam, 1992.
4
1, 11579.
1
0 C. W. Sill and H. E. Peterson, Anal. Chem., 1947, 19, 646.
5
696 | Dalton Trans., 2013, 42, 5690–5696
This journal is © The Royal Society of Chemistry 2013