Exhaustive removal of thorium and uranium traces from neodymium by liquid extraction

Article abstract of DOI:10.1134/S1066362211030076

The Th content in commercially available Nd₂O₃ samples with the main substance content of 99.0-99.998% was found to be 2-8, and that of U, 3-230 ppb. Therefore, to obtain Nd meeting the requirements of the experiment on studying neutrinoless double β-decay, it should be purified to reduce the content of Th and U impurities by a factor of 10 ³-10⁴. To this end, the extraction of Nd, Th, and U from hydrochloric acid media with solutions of trioctylphosphine oxide (TOPO) in toluene was studied. The distribution ratios of Th increase with a decrease in the HCl concentration in the initial aqueous solution. With an increase in the Cl^- concentration in the aqueous phase, the U distribution ratios decrease, probably because of a decrease in the concentration of the free extractant due to the extraction of Nd and HCl. On the contrary, the distribution ratios of Th increase with an increase in the Cl^- concentration in the aqueous phase, with the slope of the straight line in the coordinates logD_Th-log[Cl^-]_aq close to 7, which may be due to coextraction of Th with Nd in the form of the complex ThNdCl ₇. The enthalpies of formation of the extractable complexes of Th and U were determined from the temperature dependence of the extraction of Nd, Th, and U chlorides with a 0.1 M solution of TOPO in toluene. The optimal extraction system was chosen for Nd purification to remove traces of Th and U: organic phase, 0.1 M solution of TOPO in toluene; aqueous phase, 2.4 M NdCl₃ + 0.1 M HCl. From the initial aqueous solution contaning 574 ppt Th and 2837 ppt U, by single extraction with an equal volume of 0.1 M TOPO in toluene, an aqueous solution containing <10 ppt Th and 31 ppt U (detection limit 10 ppt) was obtained. By semicountercurrent extraction, from the initial aqueous NdCl₃ solution containing 200 ppb Th, the raffinate containing <10 ppt Th was obtained in one extraction step. The results obtained confirm the possibility of exhaustive removal of Th and U impurities (to the level of ≤1 ppt) from Nd by extraction with TOPO solutions from chloride solutions.

Full text of DOI:10.1134/S1066362211030076

ISSN 1066-3622, Radiochemistry, 2011, Vol. 53, No. 3, pp. 269–277. © Pleiades Publishing, Inc., 2011.
Original Russian Text © N.A. Danilov, Yu.S. Krylov, V.I. Zhilov, A.Yu. Tsivadze, E.V. Sal’nikova, I.R. Barabanov, L.B. Bezrukov, G.Ya. Novikova,
E.Ya. Yanovich, C. Cattadori, S. Nisi, M. Di Vacri, C. Salvo, A. Ianni, 2011, published in Radiokhimiya, 2011, Vol. 53, No. 3, pp. 229–236.
Exhaustive Removal of Thorium and Uranium Traces
from Neodymium by Liquid Extraction
N. A. Danilov*a, Yu. S. Krylova, V. I. Zhilova, A. Yu. Tsivadzea, E. V. Sal’nikovaa,
I. R. Barabanovb, L. B. Bezrukovb, G. Ya. Novikovab, E. Ya. Yanovichb, C. Cattadoric, S. Nisid,
M. Di Vacrid, C. Salvod, and A. Iannid
a Frumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences,
Leninskii pr. 31, block 4, Moscow, 119991 Russia; * e-mail:danilov@ipc.rssi.ru
b Institute for Nuclear Research, Russian Academy of Sciences, Moscow, Russia
c National Institute for Nuclear Physics, Milan Bicocca, Italy
Gran Sasso National Laboratory, National Institute for Nuclear Physics, L’Aquila, Italy
d
Received May 4, 2010
Abstract—The Th content in commercially available Nd2O3 samples with the main substance content of 99.0–
99.998% was found to be 2–8, and that of U, 3–230 ppb. Therefore, to obtain Nd meeting the requirements of
the experiment on studying neutrinoless double β-decay, it should be purified to reduce the content of Th and U
impurities by a factor of 103–104. To this end, the extraction of Nd, Th, and U from hydrochloric acid media
with solutions of trioctylphosphine oxide (TOPO) in toluene was studied. The distribution ratios of Th increase
with a decrease in the HCl concentration in the initial aqueous solution. With an increase in the Cl– concentra-
tion in the aqueous phase, the U distribution ratios decrease, probably because of a decrease in the concentra-
tion of the free extractant due to the extraction of Nd and HCl. On the contrary, the distribution ratios of Th
increase with an increase in the Cl– concentration in the aqueous phase, with the slope of the straight line in the
coordinates logDTh–log[Cl–]aq close to 7, which may be due to coextraction of Th with Nd in the form of the
complex ThNdCl7. The enthalpies of formation of the extractable complexes of Th and U were determined
from the temperature dependence of the extraction of Nd, Th, and U chlorides with a 0.1 M solution of TOPO
in toluene. The optimal extraction system was chosen for Nd purification to remove traces of Th and U: organic
phase, 0.1 M solution of TOPO in toluene; aqueous phase, 2.4 M NdCl3 + 0.1 M HCl. From the initial aqueous
solution contaning 574 ppt Th and 2837 ppt U, by single extraction with an equal volume of 0.1 M TOPO in
toluene, an aqueous solution containing <10 ppt Th and 31 ppt U (detection limit 10 ppt) was obtained. By
semicountercurrent extraction, from the initial aqueous NdCl3 solution containing 200 ppb Th, the raffinate
containing <10 ppt Th was obtained in one extraction step. The results obtained confirm the possibility of ex-
haustive removal of Th and U impurities (to the level of ≤1 ppt) from Nd by extraction with TOPO solutions
from chloride solutions.
Keywords: extraction, TOPO, diluents, neodymium, thoriuml, uranium
DOI: 10.1134/S1066362211030076
One of the central problems of the modern low-
energy physics is search for double neutrinoless
β-decay (0ν2β) of nuclei. The double β-decay is a
process in which the decaying nucleus emits two elec-
trons and transforms into a nucleus with a charge
higher by two units (in case when it is energetically
possible). Two types of double β-decay are possible:
The decay of the first type is allowed within the
framework of the modern physics and was found for a
number of nuclei. However, search for the decay of the
second type is more interesting. It is forbidden in the
modern elementary particle theory, because it involves
lepton number violation. Detection of 0ν2β-decay, on
the one hand, would have fundamental consequences
for particle physics and cosmology and, on the other
hand, would allow determination of the nature and
mass of neutrino, because the period of the 0ν2β-decay
–
(1.1)
(1.2)
(Z, A) → (Z + 2, A) + 2e + 2ν (2ν2β),
–
(Z, A) → (Z + 2, A) + 2e(0ν2β).
269

270
DANILOV et al.
is directly related to the neutrino mass. The most
promising candidates for studying 0ν2β-decay are iso-
topes whose nuclei are characterized by the highest
probability of such decay, e.g., 124Sn, 150Nd, and 100Mo,
and by the highest decay energy, because in this case
the discrimination of the external and internal back-
ground of the recording detectors is considerably sim-
plified [1]. Recently several research teams started
works aimed to detect 0ν2β-decay of 150Nd. This iso-
tope has very high decay energy (3.376 MeV, one of
the largest values) and therefore is one of the most
promising candidates; in this case, the external radia-
tion background in an underground laboratory will be
lower than with other isotopes [2]. As an experimental
procedure, it is suggested to introduce 150Nd into a liq-
uid organic scintillator in which the sought-for decay
will be detected. We developed a liquid organic Nd-
containing scintillator on the basis of solutions of solid
dried neodymium 2-methylvalerate and 3,5,5-trimeth-
ylhexanoate in trimethylbenzene (TMB) or linear al-
kylbenzene (LAB). The Nd concentration in the ready
scintillator is 6.5 g l–1. The results obtained by now
allowed the upper limit for the nucleus lifetime with
EXPERIMENTAL
In our study we used an aqueous NdCl3 solution
prepared by dissolving 99.99% pure Nd2O3 (Sigma
Aldrich) in hydrochloric acid (ultrapure grade).
As extractant we used trioctylphosphine oxide
(TOPO, Sigma Aldrich, >98%) dissolved in various
organic diluents. The diluents [dodecane, toluene, lin-
ear alkylbenzene (LAB) C6H5R, where R = C10–C14,
white spirit (mixture of СnH2n+2 hydrocarbons, where
n = 9.6, containing up to 12% aromatic components)]
were preliminarily purified by passing through col-
umns packed with Al2O3. All the reagents used in the
study were of ultrapure grade, and their preliminary
analysis for Th and U by ICP MS showed the absence
of these elements (detection limit 10 ppt).
The Th and U concentrations in all the objects were
determined by ICP MS with an Agilent 7500a device.
The uncertainty in determination of the impurities was
±20%. The Nd concentration was determined by com-
plexometric titration with EDTA using Xylenol Or-
ange indicator in acetate buffer (pH 5.8). The HCl con-
centration in aqueous NdCl3 solutions was determined
by alkali titration with Bromothymol Blue indicator
and in the presence of NaF. Most of the experiments
on studying the distribution of elements between the
aqueous and organic phases were performed at room
temperature (22 ± 1°С) in glass or polyethylene sepa-
rating funnels with manual stirring. The glassware
used in the experiments was washed before use three
times with hydrochloric acid (ultrapure grade) and then
with ultrapure water. The influence of temperature on
the distribution of Th, U, and Nd was studied in tem-
perature-controlled vessels (±0.1°С) under stirring
with a magnetic stirrer.
respect to 0ν2β-decay to be estimated at T1/2
>
1022 years. Further enhancement of the sensitivity re-
quires a considerable decrease in the background of the
detecting system, primarily associated with impurities
of U and Th (and elements of their radioactive series)
in Nd. Estimations show that, to attain the requred sen-
sitivity (1025 years), the impurity of U and Th in the
ready scintillator should not exceed 10–15 g g–1 (ppq).
Taking into account dilution, the Th impurity in the
initial Nd should not exceed 10–12 g g–1, and the limit-
ing impurity is Th because of hard radiation of prod-
ucts in the Th decay chain (Eβ > 5 MeV). At the same
time, determination of the Th content in commercially
avaliable Nd2O3 samples of 99.0–99.998% purity re-
vealed the presence of Th (2–8 ppb) and tens times
higher content of U. Therefore, Nd should be prelimi-
nary purified to reduce the content of U and Th impuri-
ties by a factor of ~104. The goal of this study is the
development of methods for extractive purification of
Nd to the required level [10–12 g g–1 (ppt)].
When performing a semicountercurrent process for
Nd purification to remove Th, the fraction of Th (σ)
remaining in the aqueous phase after passing a definite
volume W of the organic phase was calculated by the
formula [3, 4]
σ = [Th]aq/[Th]in = exp(– WDTh) /[V0(1 + rDTh )],
The presence of microimpurities of 40K and 176Lu
(present in the initial Nd) in the scintillator does not
give rise to serious experimental problems, because the
decay energies of 40K (Еγ = 1.461 MeV) and 176Lu
(Еβ = 0.6 MeV, Еγ = 0.307 MeV) are considerably
lower than that of 150Nd (Е = 3.376 MeV). Therefore,
in this paper we do not consider separation of these
impurities.
where [Th]ni is the Th concentration in the initial aque-
ous solution of NdCl3; [Th]aq, Th concentration in the
aqueous NdCl3 solution after passing a volume W of
the organic phase; W, extractant volume passed
through the cell; DTh, Th distribution ratio; V0, volume
of the aqueous NdCl3 solution used for the calculation
(200 ml); and r, volume ratio of the phases in the
emulsion, ΔW : Vaq (r = 0.1).
RADIOCHEMISTRY Vol. 53 No. 3 2011

EXHAUSTIVE REMOVAL OF THORIUM AND URANIUM TRACES
271
Table 1. Results (content, ppm) of determining various im-
purities in two Nd2O3 samples of different purity grades by
ICP MS
RESULTS AND DISCUSSION
For Nd purification, it was appropriate to choose
Nd2O3 samples with the lowest content of Th, U, and
other undesirable impurities.
Element Nd2O3, 99.99% (Sigma Aldrich) Nd2O3, 99.0 %
La
K
Ce
Pr
Sm
Eu
Er
8.3
19.0
0.5
3.0
1.0
9.6
31.0
76.0
Preliminary analysis of Nd2O3 (seven samples from
different suppliers) of 99.0–99.998% purity for the
content of Th, U, and a number of other impurities was
performed with an HPGe detector arranged in a spe-
cially equipped low-background room of the Gran
Sasso underground laboratory (Italy) and using ICP
MS. Both methods gave consistent results. The Th
concentration in Nd2O3 samples varied within 2–8 ppb,
which approximately corresponds to the activity of 8–
30 mBq kg–1. The U content varied in a wide range,
from 3 to 230 ppb (Table 1). In addition, in the 99.0%
pure sample the Pr content was inadmissibly high. The
Pr absorption peak near 440 nm (Fig. 1) negatively
affects the scintillator transparence (the light sensitiv-
ity maximum for the photoelectron multipliers used in
the study is in the range 400–450 nm).
11 000.0
1.0
0.048
2.0
0.040
2.0
Tm
Lu
Th
U
0.32
0.68
1.3
0.003
0.23
0.045
0.002
0.003
tion should be performed using NdCl3 solutions of
concentration as high as possible. In the process, the
distribution ratio of Th (DTh) should be maximal, and
that of Nd (DNd), minimal, so as to reduce the Nd loss
in the course of purification to a minimum.
For further experiments we chose Nd2O3 supplied
by Sigma Aldrich, because in this sample the content
of the undesirable impurities was the lowest. However,
even in this case it was necessary to decrease the Th
concentration by a factor of 103–104.
Third, the mineral acid (HCl) concentration in the
aqueous NdCl3 solution used for the purification
should be minimal, so as to avoid neutralization of its
excess in the subsequent synthesis of neodymium car-
boxylate.
The most suitable methods for purification of Nd
and other REEs are chromatography and liquid extrac-
tion. Our choice of the purification method was pri-
marily governed by requirements associated with the
convenience of the subsequent synthesis of neodym-
ium carboxylate used for preparing the liquid organic
Nd-containing scintillator and with minimization
of additional operations with the purified NdCl3, so as
to reduce to a minimum the risk of its secondary con-
tamination.
In view of these requirements and available data,
we believe that the most convenient will be liquid ex-
traction of Th and U from an aqueous NdCl3 solution
with neutral organophosphorus compounds.
It is known from the literature [5–10] that Th and U
are well extracted from HCl solutions with solutions of
trioctylphosphine oxide (TOPO). The distribution ra-
tios of Th (1 g l–1 in the initial solution) increase with
First, it is preferable to synthesize neodymium car-
boxylate used for preparing the scintillator by the reac-
tion of the ammonium carboxylate with an aqueous
NdCl3 solution. The use of neodymium nitrate for this
purpose is undesirable, because the oxidizing action of
nitrate anion in its contact with the organic scintillator
deteriorates its transparency and light output. Although
in most cases extractive separation of REEs is per-
formed using nitrate media, in our case it is appropriate
to separate Th and U impurities from Nd using chlo-
ride media, so as to avoid the subsequent conversion of
nitrates to chlorides.
Fig. 1. Optical density of aqueous NdCl3 solutions: (1) Nd
from Sigma Aldrich, [Nd] = 52.6 g l–1; (2) 99.0% pure Nd,
[Nd] = 53.4 g l–1.
Second, from the economic viewpoint, the purifica-
RADIOCHEMISTRY Vol. 53 No. 3 2011

272
DANILOV et al.
Table 2. Dependence of the distribution ratio DTh on
the concentration ratio [HCl]aq : [NdCl3]aq in the initial aque-
ous phase. Organic phase: 0.2 M TOPO in toluene. Initial
aqueous phase: solution of NdCl3 + HCl + 100 ppm Th;
∑[Cl–]aq = 6 M
the concentration ratio of NdCl3 and HCl in the initial
solution affects DTh at a constant [Cl–] = 6 M, which
corresponds to the maximum of the Th distribution
ratios in the extraction from hydrochloric acid solu-
tions [5–10] with [Th] = 1 g l–1. To confirm the possi-
bility of using the suggested procedure for Nd purfica-
tion in a wide range of Th concentrations and to obtain
reliable values of the distribution ratios, to the initial
NdCl3 solutions we added Th in amounts of 100 ppm
and 200 ppb. The results we obtained show (Tables 2,
3) that, with a decrease in the HCl concentration and
an increase in the NdCl3 concentration in the initial
aqueous solution, DTh and DNd regularly increase,
reaching a maximum at a minimal HCl concentration.
The Th recovery in one step under these conditions
was close to 99.8%. At the same time, the HCl extrac-
tion increases with an increase in its concentration in
the aqueous phase. At [HCl]aq = 4.5 M, 0.2 M HCl
passes to the organic phase, which corresponds to the
ratio [TOPO] : [HCl]о = 1 : 1. This fact suggests for-
mation in the organic phase of the compound
[TOPO·HCl]о. This may be responsible for the ob-
served decrease in the Th distribution ratios with an
increase in the HCl concentration in the aqueous
phase, because the concentration of the “free” extrac-
tant in the organic phase decreases (a part of TOPO is
bound in the complex TOPO·HCl). At high HCl con-
centration in the aqueous phase, apparently, formation
of the complexes TOPO·2HCl is possible (Table 4). In
the presence of 1 M NdCl3 in the aqueous phase, the
distribution ratios of HCl are higher than under similar
conditions but without NdCl3. This may be due to the
salting-out effect of NdCl3 toward HCl. Thus, the high-
est distribution ratios of the limiting radioactive
Th impurity are observed at the maximal concentration
of NdCl3 and minimal concentration of HCl in the
aqueous phase in a wide range of Th concentrations
(Tables 2, 3). This fully meets the above-formulated
conditions of the convenience of the subsequent syn-
[NdCl3]aq, [HCl]aq, [Th]aq after
Degree of Nd
purification, %
DTh
M
M
extraction, ppm
2.6
2.0
1.66
1.0
0.33
0.1
0.1
1.0
3.0
5.0
6.0
0.03
0.18
0.33
1.5
3.3
8.3
>1000.0
554.0
293.0
65.7
>99.9
99.8
99.6
98.5
96.7
91.6
29.3
11.0
0.0
an increase in the HCl concentration in the aqueous
phase, reaching a maximum (DTh ≈ 100) at its concen-
tration of 6 M, after which they start to decrease. Addi-
tion of a salting-out agent, LiCl, also leads to a consid-
erable increase in DTh. Data on the extraction of mi-
croamounts of many elements from HCl solutions with
a 5% solution of TOPO in toluene are reported in [9].
The U distribution ratios under the same conditions are
≥103. The distribution ratios of La, Ce, Pr, and Y do
not exceed 10–3. Studies of the extraction from chlo-
ride solutions of macroamounts of some individual
REEs with a 0.5 M solution of tri-1-methylheptyl-
phosphine oxide [11] and of Yb with triisoamyl-
phosphine oxide [12] showed that the REE extraction
was noticeable only at their high concentrations in the
aqueous phase or on addition of salting-out agents. We
found no published data on the extraction of ultrasmall
amounts of Th and U in the presence of high concen-
trations of NdCl3.
Therefore, it was necessary to examine regular
trends in the distribution of microamounts of Th and U
in the presence of high concentrations of Nd and to
find optimal conditions for Nd purification to remove
Th and U impruities. Preliminarily we examined how
Table 3. Dependence of the distribution ratio DTh on the concentration ratio [HCl]aq : [NdCl3]aq in the initial aqueous phase.
Organic phase: 0.2 M TOPO in toluene. Aqueous phase: solution of NdCl3 + HCl + 200 ppb Th; ∑[Cl–]aq = 6 M
[NdCl3]aq, [HCl]aq, [Th]aq after extraction,
Degree of Nd
DTh
[Nd]o, g l–1 [HCl]o, M
DHCl
D
Nd × 104
M
M
0.1
0.75
1.5
3.0
4.5
ppb
0.2
purification, %
2.21
1.75
1.5
1.07
0.5
1000.0
600.0
434.0
91.7
99.9
99.8
99.8
98.9
98.6
3.5
0.01
0.08
0.12
0.14
0.2
0.1
0.12
0.087
0.049
0.046
110.0
16.2
8.8
0.098
0.46
2.18
2.8
0.41
0.19
0.12
0.035
8.32
4.85
71.4
RADIOCHEMISTRY Vol. 53 No. 3 2011

EXHAUSTIVE REMOVAL OF THORIUM AND URANIUM TRACES
Table 4. Isotherms of HCl extraction with TOPO solutions in toluene
273
0.2 M TOPO + HCl
0.2 TOPO + HCl + 1 M NdCl3
0.1 M TOPO + HCl
[HCl]aq,
M
[HCl]o, M
0.0026
0.023
0.088
0.15
DHCl
[HCl]o, M
0.04
0.086
0.18
0.24
–
DHCl
4.2 × 10–2
4.5 × 10–2
4.7 × 10–2
4.3 × 10–2
–
[HCl]o, M
0.0012
0.0093
0.004
DHCl
0.96
2.02
4.15
5.74
8.8
2.6 × 10–3
1.1 × 10–2
2.1 × 10–2
2.6 × 10–2
2.63 × 10–2
3.2 × 10–2
1.25 × 10–3
4.6 × 10–3
9.7 × 10–3
1.44 × 10–2
1.4 × 10–2
2.1 × 10–2
0.0082
0.116
0.22
0.36
11.2
–
–
0.24
Table 5. Influence of the TOPO concentration in toluene on the extraction of Nd from the aqueous phase NdCl3 + 0.1 M HCl
0.52 M TOPO
[Nd]о, g l–1
0.2 M TOPO
[Nd]о, g l–1
0.1 M TOPO
[Nd]о, g l–1
[Nd]aq, g l–1
DNd
DNd
DNd
186.0
210.0
260.0
315.0
350.0
1.01
2.34
6.64
5.4 × 10–3
1.1 × 10–2
2.6 × 10–2
5.9 × 10–2
0.017
0.069
0.42
3.47
4.5
3.1 × 10–5
3.3 × 10–4
1.6 × 10–3
1.1 × 10–2
1.3 × 10–2
0.008
0.009
0.064
1.32
4.3 × 10–5
1.5 × 10–5
2.5 × 10–4
4.2 × 10–3
5.7 × 10–3
18.5
2.0
ride concentration in the aqueous phase [Cl–]aq, the
higher DTh; (3) at constant chloride conce[Cl–]aq, the
slope of the dependence in the logDTh–log[TOPO]o
coordinates should correspond to the number of TOPO
molecules incorporated in the extractable complex (n);
(4) at a constant TOPO concentration in the organic
phase [TOPO]o, the slope of the dependence in the co-
ordinates logDTh–log[Cl–]aq should correspond to the
number of Cl– anions incorporated in the extractable
complex.
thesis of neodymium carboxylate. We chose the mini-
mal concentration [HCl]aq = 0.1 M to prevent possible
hydrolysis of Th. In the process, up to 4.5 g l–1 of Nd
passes into the organic phase at the extractant concen-
tration of 0.2 M. With an increase in the TOPO con-
centration in the organic phase, the Nd concentration
in it (and correspondingly the loss) increases also
(Table 5). Extraction of thorium chloride with TOPO
can be described by Eq. (2):
Th + 4Cl–aq + nTOPOо = [ThCl4·nTOPO]o.
(2)
4+
aq
As can be seen (Fig. 2), DTh increases with an in-
crease in the TOPO concentration, with the slope of
the straight line in the coordinates logDTh–log[TOPO]
close to 2, which suggests solvation of the extractable
Th complex with two TOPO molecules.
The extractoion constant
Keоx= [ThCl4·nTOPO]o/([Th4 +]aq[Cl–]4aq[TOPO]оn). (3)
On the other hand,
As seen from Fig. 2 and Table 6, the use of even
0.05 M TOPO in toluene allows recovery of more than
99.5% (DTh ≈ 350) of the initial Th amount. In so do-
D
Th = [ThCl4·nTOPO]o/[Th4 +]aq.
(4)
(5)
Substituting (4) in (3), we obtain
Kex = DTh /([Cl–]4aq[TOPO]оn).
Taking the logarthm of (5), we obtain
logDTh = logKex + 4log[Cl–]aq + nlog[TOPO]о.
(6)
Analysis of Eq. (6) allows the following conclu-
sions: (1) The higher the TOPO concentration in the
organic phase, the higher DTh; (2) The higher the chlo-
Fig. 2. Influence of the TOPO concentration in toluene on
the distribution ratios of Th in the extraction from the aque-
ous phase containing 2.2 M NdCl3 + 200 ppb Th.
RADIOCHEMISTRY Vol. 53 No. 3 2011

274
DANILOV et al.
Table 6. Distribution ratios of Th in the extraction with
TOPO solutions in toluene from the aqueous phase contain-
ing 2.26 M NdCl3 + 0.1 M HCl + 300 ppb Th
can be readily stripped to more than 99% by single
washing with 0.1 M HCl at Wо : Vaq = 2 : 1 and can be
returned into the process.
[Th]aq, ppb, after
extraction
Th recovery,
%
[[TOPO]o, M
DTh
A study of the U distribution under similar condi-
tions showed that its distribution ratios were close to
~1000, with 99.9% recovery in one step.
0.1
0.05
0.45
0.90
700.0
350.0
99.85
99.71
We also examined the influence of the Cl– concen-
tration on the Th and U distribution. The concentration
of chloride anions in the aqueous phase was adjusted
by adding NdCl3. With an increase in the NdCl3 con-
centration in the aqueous phase, Th and U behave dif-
ferently. The U distribution ratios decrease with an
increase in the NdCl3 concentration in the aqueous
phase (Fig. 3). At the same time, the Th distribution
ratios increase, with the slope of the linear portion of
the logDTh–log[Cl–]o plot close to 7.
Table 7. Diluent effect on the Nd purification to remove Th.
Organic phase: 0.1 M TOPO solutions in various diluents;
aqueous phase: 2.2 M NdCl3 + 0.1 M HCl + 200 ppb Th
[Nd]o, [Nd]o, % of
Th recovery,
%
Diluent
Toluene
DTh
g l–1
1.4
initial
0.38
0.26
0.21
0.56
0.88
666.0
125.0
137.0
166.0
183.0
99.85
99.2
99.28
99.4
Trimethylbenzene 0.84
White spirit
LAB
Decane
0.66
1.77
2.8
A decrease in the U distribution ratios with an in-
crease in the NdCl3 concentration in the aqueous phase
is attributable to a decrease in the concentration of the
free extractant due to extraction of Nd and HCl. On the
contrary, the very large slope for Th can be accounted
for by coextraction of Th with Nd in the form of the
complex NdThCl7. Taking into account data shown in
Figs. 2 and 3, the reaction of the Th extraction can be
presented in the form
99.45
ing, the Nd loss decreases to 0.28% of its content in
the initial aqueous solution. We examined the diluent
effect on the Th distribution ratios. As diluents we took
substances used for preparing scintillators, for which it
was known that the Th content in them was below
1 ppt. As can be seen (Table 7), with any diluent more
than 99% of the initial Th amount can be removed
in one step, but the best results were obtained with
toluene.
Th4+ + Nd3+ + 7Cl– + 2ТОPО = (ThNdCl7·2ТОPО)о.
However, additional independent studies are re-
quired to make final conclusions on the composition of
the extractable complex.
To separate Th and U impurities from Nd, we chose
the following extraction system. Organic phase: 0.1 M
solution of TOPO in toluene; aqueous phase: solution
of 2.4 M NdCl3 + 0.1 M HCl. With this extraction sys-
tem, it is possible to efficiently recover Th and U into
the organic phase with minimal loss of Nd. It should
be noted that the Nd extracted into the organic phase
It is known that, to decrease the extractant viscosity
and attain complete backwashing, extraction is per-
formed at elevated temperatures [13]. Previously Fide-
lis and Siekierski [14, 15] examined the influence of
temperature on the REE extraction with tributyl phos-
phate and HDEHP from nitric acid solutions. They
showed that the enthalpies of extraction in these sys-
tems are negative. Sato and Yamatake [5] studied the
influence of temperature (10–40°С) on the extraction
of U and Th with a 0.02 M solution of TOPO in kero-
sene from aqueous solutions containing 4 M HCl and
4 M LiCl + 0.1 M HCl; they found that the enthalpies
of extraction were 29.05 and 19.7 kJ mol–1, respec-
tively.
A study of the temperature dependence of the ex-
traction showed that the distribution ratios of U and Th
decrease with increasing temperature, whereas the Nd
extraction is practically independent of temperature
Fig. 3. Influence of the [Cl–] concentration in the aqueous
phase on the distribution ratios of (1) U and (2) Th. Initial
concentrations of U and Th 200 ppb each. Extractant:
0.1 M TOPO in toluene.
RADIOCHEMISTRY Vol. 53 No. 3 2011

EXHAUSTIVE REMOVAL OF THORIUM AND URANIUM TRACES
275
Table 8. Influence of temperature on the distribution coeffi-
cients of Th, U, and Nd. Organic phase: 0.1 M TOPO in
toluene; aqueous phase: 2.4 M NdCl3 + 0.1 M HCl +
200 ppb Th + 200 ppb U
(Table 8). Negative enthalpies of extraction, obtained
from the data given in Fig. 4, show that the extraction
is exothermic. The quantities –ΔH for Th and U, found
using the equation ΔlogD/Δ(1/T) = –ΔH/(2.303R), are
11.4 and 15.9 kJ mol–1, respectively, which indicates
that NdCl3 exerts a weaker salting-out effect than does
LiCl [5].
Temperature, °C
DTh
537
470
412
338
318
DU
D
Nd × 103
6.93
7.0
7.06
7.22
7.2
30
40
50
60
70
1717
1382
1152
966
Because an experiment in studying double β-decay
requires 12 m3 of the liquid organic Nd-containing
scintillator with a Nd concentration of 6.5 g l–1, the
total amount of Nd to be purified to remove Th and U
will be 78 kg. This corresponds to 210 l of aqueous
NdCl3 solution with a concentration of 2.4 M (366 g l–1).
This volume of NdCl3 can be placed in a plastic vessel
and brought in contact two times in succession with
equal volumes of 0.1 M TOPO or purified by continu-
ous passing of the calculated amount of the extractant
through and aqueous NdCl3 solution (semicounter-
current).
812
radiation. The sample being measured is placed in the
center of the sphere. The detector is arranged in the
Gran Sasso underground laboratory (Italy) at a depth
of 1.4 km. Secondary purification of the NdCl3 solu-
tion obtained with a fresh portion of the extractant was
not performed, because an analytical method of the
required sensitivity was lacking. Moreover, further
purification should be performed in a specially
equipped “clean” room or in specially equipped glove
boxes ensuring maximal protection from the ingress of
Th and U from the environment.
The NdCl3 purification was performed by these two
methods: single extraction in a separating funnel and
semicountercurrent on one extraction step in a com-
mon laboratory room.
Semicountercurrent purification of Nd. The
semicountercurrent method allows exhaustive purifica-
tion of any component of a binary mixture being sepa-
rated. An additional feature of the semicountercurrent
separation is that the degree of recovery of the better
extractable component is maximal at the minimal vol-
ume ratio of phases in the emulsion r = ΔWo/Vaq. Thus,
the separation should be performed at minimal values
of r, taking into account, of course, the balance be-
tween the productive capacity and degree of recovery.
In this case, the exponent in the main equation of semi-
countercurrent extraction is maximal at r → 0 [1, 2].
Taking into account the balance between the produc-
tive capacity and degree of recovery, we chose r = 0.1.
Single extraction of Th and U was performed in a
polyethylene separating funnerl charged with 400 ml
of a 0.1 M TOPO solution in toluene and 400 ml of a
solution of 2.4 M NdCl3 + 0.1 M HCl, containing
574 ppt Th and 2837 ppt U. The aqueous NdCl3 solu-
tion was prepared from Nd2O3 preliminarily purified
by a commercial company to remove Th from the ini-
tial Th content of 6 ppb to its content in the final prod-
uct of 574 ppt. After manual shaking for 5 min and
settling for 30 min, the phases were analyzed for the
Nd, Th, and U content. In the aqueous phase, we found
<10 ppt Th (detection limit 10 ppt) and 31 ppt U. Thus,
the recovery of Th was >97.9% and that of U, 98.9%,
which is somewhat lower than the results obtained in
the preliminary experiments in which the Th and U
concentrations in the initial aqueous solution were
from 100 ppm to 200 ppb (Tables 2, 3). This may be
due to possible ingress of Th and U from the environ-
ment, because manipulations were performed in a
common laboratory room. The solution was delivered
to long-term measurements of ultralow radioactivity
levels with a special ultrasensitive detector (Counting
Test Facility, CTF). This detector contains 4 g of a
scintillating liquid placed inside a transparent Nylon
sphere of radius 1 m and surrounded by 1000 t of ul-
trapure water for the protection from the background
1
2
Fig. 4. Influence of temperature on the distribution ratios of
(1) U and (2) Th in their extraction with a 0.1 M TOPO
solution in toluene from the aqueous phase containing
2.4 M NdCl3 + 0.1 M HCl + 200 ppb Th + 200 ppb U.
RADIOCHEMISTRY Vol. 53 No. 3 2011

276
DANILOV et al.
Table 9. Influence of the extractant volume passed through a semicountercurrent extraction cell containing aqueous solution
of NdCl3 on the fraction of Th remaining in the aqueous phase (σ). V = 200 ml, r = 0.1, DTh = 600. Extractant: 0.1 M TOPO
in toluene; initial aqueous phase: 2.4 M NdCl3 + 0.1 M HCl + 200 ppb Th
Experimental data
[Th]aq, ppb
Calculated values
[Th]aq, ppb
Wо, ml
σ
σ
10.0
50.0
100.0
200.0
300.0
500.0
143
25
2.1
0.09
<0.01
<0.01
0.72
0.125
122
17
0.61
8.5 × 10–2
7.3 × 10–3
5.3 × 10–5
3.9 × 10–7
2.1 × 10–11
1.05 × 10–2
4.5 × 10–4
<5.0 × 10–5
<5 × 10–5
1.46
1.1 × 10–2
7.8 × 10–5
4.2 × 10–9
One-step semicountercurrent process was per-
formed in a cylindrical polyethylene cell equipped
with a polyethylene propeller stirrer. The cell was
charged with 200 ml of an aqueous solution of 2.4 M
NdCl3 + 0.1 M HCl + 200 ppb Th. The organic phase
was fed under the stirrer with a peristaltic pump at a
rate of 100 ml h–1. In the upper part of the cell, we ar-
ranged a polyethylene plate with cylindrical holes
where the emulsion underwent phase separation. The
organic phase was accumulated over the plate and, af-
ter reaching the discharge level, was removed from the
cell. The aqueous phase was returned into the cell. Af-
ter passing a definite volume of the extractant, an
aqueous phase sample was taken, washed twice with
toluene, and analyzed by ICP MS to determine the Th
concentration. The experimental data were compared
with the calculated values. As can be seen (Table 9), to
decrease the Th concentration from 200 ppb in the ini-
tial aqueous solution of NdCl3 to <10 ppt (by a factor
of >104), it is sufficient to pass through the semicoun-
tercurrent step the extractant volume Wо equal to
1.5 volumes of the aqueous solution of NdCl3 present
in this step. This fact confirms high efficiency of the
semicountercurrent separation method.
specially equipped “clean” room with maximum possi-
ble protection from the Th ingress from the environ-
ment.
Thus, we found the optimal conditions for Nd puri-
fication to remove Th and U microimpurities using the
extraction system TOPO–NdCl3–0.1 M HCl. With this
extraction system, we performed the separation of U
and Th microimpurities from Nd by single extraction
in a separating funnel with an equal volume of the ex-
tractant and by semicountercurrent extraction and ob-
tained Nd samples containing the limiting impurity,
Th, in an amount of less than 10 ppt. The extraction
relationships that we obtained for Nd, Th, and U and
the results of Nd purification experiments confirm the
possibility of obtaining Nd samples containing no
more than 1 ppt Th and U, provided that the ingress of
Th and U from the environment is limited to the re-
quired extent.
ACKNOWLEDGMENTS
The study was supported by the Gran Sasso Na-
tional Laboratory of the National Institute for Nuclear
Physics (L’Aquila, Italy) and by the Russian Founda-
tion for Basic Research (project nos. 09-02-92423-K9
and 09-02-00279а).
Certain disagreement between the experimental
data and calculated values (Table 9) may be due both
to design features of the extraction cell and to possible
ingress of Th from the environment in the course of the
experiment. The effect of these factors becomes more
pronounced as the Th concentration in the aqueous
phase decreases. After passing 500 ml of the extractant,
the process was stopped, because the Th content in the
aqueous phase was below the detection limit of the
available analytical procedure. To perform more ex-
haustive removal of Th from Nd, it is necessary to de-
velop a more sensitive procedure for Th analysis. In
addition, further purification should be performed in a
REFERENCES
1. Gehman, V M. et al., ArXiv: 0911.2198[nucl-ex], No-
vember 11, 2009.
2. Barabanov, I. et al., J. Phys., 2008, Conf. Ser. 136, pa-
per 042 088.
3. Korpusov, G.V. and Kuznetsov, G.I., Radiokhimiya,
1974, vol. 5, no. 5, pp. 695–701.
4. Kodina, G.S. and Korpusov, G.V., Radiokhimiya, 2002,
vol. 44, no. 1, pp. 61–64.
5. Sato, T. and Yamatake, M., J. Inorg. Nucl. Chem.,
RADIOCHEMISTRY Vol. 53 No. 3 2011

EXHAUSTIVE REMOVAL OF THORIUM AND URANIUM TRACES
277
1969, vol. 31, pp. 3633–3642.
6. Ross, W.J. and White, J.C., Anal. Chem., 1959, vol. 31,
no. 11, p. 1847.
11. Danilov, N.A., Korpusov, G.V., and Krylov, Yu.S., Ra-
diokhimiya, 2003, vol. 45, no. 2, pp. 128–133.
12. Danilov, N.A., El Fared, S., and Korpusov, G.V., Ra-
diokhimiya, 1992, vol. 34, no. 2, pp. 72–81.
7. Alian, A. and Zaborenko, K.B., Z. Anal. Chem., 1968,
vol. 238, nos. 1–3, p. 318.
13. Mikhailichenko, A.I. and Mikhlin, E.V., Redkoze-
mel’nye elementy (Rare-Earth Elements), Moscow: Me-
tallurgiya, 1987.
8. Mrochec, J.E., Laughlin, J.O., and Banks, G.V.,
J. Inorg. Nucl. Chem., 1965, vol. 27, no. 3, p. 603.
9. Watanabe, K., J. Nucl. Sci. Technol., 1964, vol. 1, no. 5,
pp. 155–162.
14. Fidelis, I. and Siekierski, S., J. Inorg. Nucl. Chem.,
1967, vol. 29, pp. 2629–2635.
10. Danilov, N.A. and Shatokhina, O.B., Izv. Vyssh.
15. Fidelis, I. and Siekierski, S., J. Inorg. Nucl. Chem.,
Uchebn. Zaved., Tsvetn. Metall., 1982, no. 2, pp. 71–75.
1970, vol. 32, pp. 997–1003.
RADIOCHEMISTRY Vol. 53 No. 3 2011