International Journal of Analytical Mass Spectrometry and Chromatography, 2013, 1, 61-71 Published Online September 2013 (
Separation Behavior of U(VI) and Th(IV) on a Mixed
Ion Exchange Column Using 2,6-Pyridine Dicarboxylic
Acid as a Complexing Agent and Determination of
Trace Level Thorium in Uranium Matrix Employing
High Performance Ion Chromatography
Vaibhavi V. Raut1, S. P. Roy2, M. K. Das1, S. Jeyakumar1*, K. L. Ramakumar3
1Radioanalytical Chemistry Division, Bhabha Atomic Research Centre, Mumbai, India
2Product Development Division, Bhabha Atomic Research Centre, Mumbai, India
3Radiochemistry and Isotope Group, Bhabha Atomic Research Centre, Mumbai, India
Email:,,, *,
Received July 24, 2013; revised August 25, 2013; accepted September 25, 2013
Academic Editor: Prof. N. Sivaraman, HBNI,
India and Head, SCSS, Chemistry Group, Indira Gandhi Centre for Atomic Research, Kalpakkam-603102, INDIA
Copyright © 2013 Vaibhavi V. Raut et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Retention behavior of U(VI) and Th(IV) as their 2,6-pyridine dicarboxylic acid (PDCA) complexes on reversed phase
and ion exchange (cation, anion and mixed ion exchange) columns was studied and based on the results, a simple ion
chromatography method for the determination of trace level thorium in uranium oxide using 0.075 mM 2,6-pyridine
dicarboxylic acid (PDCA) and 1 M KNO3 in 1.2 M HNO3 as eluent (flow rate 1 mL/min) was proposed. The advantage
of the developed method is that the separation of uranium matrix is not required prior to the ion chromatographic de-
termination of trace Th. Separation was carried out on a mixed ion exchange stationary phase and a 104 M arsenazo (III)
solution was used as post column reagent for detecting the separated metal ions. The separation of Th from uranium
using PDCA in the present investigation is attributed through cation exchange mechanism. A calibration plot was con-
structed by following the standard addition method over the concentration range of 0.25 to 10 ppm of Th in the presence
of uranium matrix, which resulted in a linear regression coefficient of 0.9978. The precision of the method was better
than 5% and the LOD for Th was found to be 0.1 ppm (S/N = 3). The method has been validated by comparing the re-
sults with the results obtained from ICP-MS analysis where the Th is separated from the uranium matrix. The pro-
posed method is simple, rapid, accurate and cost effective compared to techniques like ICP-MS or ICP-AES and is suit-
able for the routine kind of analysis.
Keywords: Ion Chromatography; Uranium; Thorium; 2,6-Pyridine Dicarboxylic Acid
1. Introduction
The development of fast breeder reactors in India is in-
evitable as India has limited resources of uranium and the
vast thorium resources require breeder cycle for exploita-
tion [1,2]. The Proto Type Fast Breeder Reactor (PFBR),
a first commercial fast breeder reactor of India, is under
construction and it will use a mixed uranium-plutonium
oxide (MOX) as fuel. Thorium is a trace impurity associ-
ated with all uranium based fuels [3]. Thorium, on irra-
diation with a neutron produces 233U, is a useful fissile
isotope. Small amounts of 232U isotope are also produced
along with 233U. 232U is an undesirable isotope as its ra-
dioactive decay chain produces short-lived and gamma
emitting radionuclides 212Bi and 208Tl, leading to high
radiation exposure [4]. Hence, the uranium based fuels
that are used in the fast reactors has stringent specifica-
tion for Th and in India, the limit is fixed as 200 ppmw
[5]. Hence, determination of trace Th in uranium is in-
dispensible in the chemical quality control of fast reactor
*Corresponding author.
opyright © 2013 SciRes. IJAMSC
Although different analytical techniques are available
for the determination of Th at trace levels [6-8], they
require quantitative separation of Th from uranium ma-
trix which is a prerequisite for analytical determination.
An ICP-MS method [9] was earlier reported from our
laboratory for the determination of Th in uranium matrix,
which employed a two-stage separation sequence in-
volving an initial solvent extraction for the removal of
uranium matrix followed by an ion exchange purification
to get the required quality of sample solutions suitable
for ICP-MS analysis. The conventional separations fol-
lowed by instrumental analysis employing ICP-MS or
AES are time consuming and expensive. To adopt on
routine basis, a reliable, cost effective and rapid analyti-
cal method is desirable.
Ion chromatography (IC) and High performance liquid
chromatography (HPLC) are promising techniques for
the separation of polyvalent metal cations by adding
suitable complexing agents in the mobile phase [10-13].
Many chromatographic studies have been reported on the
selective separation of Th and U in variety of samples
using reversed phase surfaces with weak organic acids as
chelating agents [14,15]. In these separations, the extent
of separation depended on the nature of the metal com-
plexes formed and their conditional stability constants
[16]. Interestingly, it is reported that the elution order of
U and Th on a reversed phase surface could be altered by
selecting a suitable complexing agent. For example, an
elution order of Th followed by U was observed in a RP
surface when α-hydroxy isobutyric acid (α-HIBA) was
used as a complexing agent [17,18]. However, this elu-
tion order got reversed when mandelic acid (MA) was
used in place of α-HIBA [19]. The change in elution or-
der may be explained on the basis of differences in the
complexing abilities of the ligands with respect to indi-
vidual metal cation, effective charge, thermodynamic and
kinetic stabilities and hydrophobicities of the metal com-
plexes formed.
Although α-HIBA and MA are widely used as com-
plexing agents in the separation of U and Th, 2,6-pyri-
dine dicarboxylic acid (PDCA), a non-hydroxy dicar-
boxylic acid, has been identified as a promising ligand
for the separation of heavy metal ions [20-23]. The re-
versed phase surfaces dynamically modified with PDCA
[22,24] showed stronger retention for U(VI) than Th(IV)
due to the high affinity exhibited by the UO2
2+ ion for
PDCA [20]. A study reports [22] that dynamic modifica-
tion of a small PRP column with 0.1 mM PDCA in 1 M
KNO3 and 0.5 M HNO3 could separate some of the acid
hydrolysable heavy metal ions like U, Th, Zr, etc. It also
states that the lanthanides and other transition metal ca-
tions have no retention on the column and get excluded
at the solvent front.
Other than RP columns, ion exchange columns were
also used for the separation of U and Th. The ion chro-
matographic separation of U and Th using pure ion ex-
change separation is difficult due to the large differences
in their distribution ratios. This is because the UO2
2+ and
Th4+ ions have considerable differences in their ion ex-
change affinity. Separation of U(VI) as UO2
2+ on a low
capacity strong cation exchanger has been reported [25].
Even though few studies reported the separation of U and
Th from water and geological samples using a pure ca-
tion exchange column with gradient elution of a mobile
phase consisted of HCl and Na2SO4 [26,27], use of hy-
drochloric acid as eluent was not recommended for rou-
tine analysis due to its corrosiveness, which demanded
utmost care on the instrument maintenance. Earlier in our
laboratory, we have separated uranium and thorium using
a short length cation exchange column using a mobile
phase consisting of 0.08 mM PDCA in 0.24 M KNO3 and
0.22 M HNO3 (pH ~ 0.6) [23]. This method was found
suitable for the rapid separation of both U and Th at trace
level concentrations; however, it would not be possible to
extend the procedure when one of the metal ions was at
higher concentrations.
The studies that reported the basis of RP-surface modi-
fications and ion exchange separations were capable of
providing good separation between U and Th, provided
their concentrations were comparable. However, the sepa-
ration of U and Th is a difficult task when one of these
metal ions presents in large amount. When determining
one metal ion at trace level in the presence of large ex-
cess of other, the system loses its capacity, resulting in
poor resolution between the analytes [21]. The most
practicable way is to remove matrix element by means of
conventional methods like solvent extraction, column
chromatography and ion exchange prior to the IC or
HPLC. Such sample preparation procedures are not only
time consuming and laborious but also susceptible to
cause inconsistency in the recoveries of the analyte of
interest. Moreover, methods that follow pre-separation
followed by instrumental analysis demand the use of
certified reference materials to assess the recoveries.
Since the objective of this study is to separate trace
thorium from the dissolved uranium oxide samples by
direct sample injection, one of the desired conditions to
realize the objective is to have an elution order of Th(IV)
followed by U(VI) so as to get the elution of Th first and
then the matrix U. This elution order can be obtained by
separating them on a reversed phase column (modified
and unmodified) using α-HIBA as eluent [11]. But, the
separation of Th in presence of bulk U could not be real-
ized by following the reported procedures. A good sepa-
ration of lanthanides, Th and U at relatively large con-
centrations was reported [17] where the C-18 RP column
was coated with bis-2-ethylhexyl succinamic acid (BE-
HSA). However, the same study showed that it could not
separate Th from U when the uranium concentration in
the sample was more than 5 mg/mL. Mandelic acid, an-
Copyright © 2013 SciRes. IJAMSC
V. V. RAUT ET AL. 63
other complexing agent frequently used in the separation
of U and Th, could not be used as it brings reversed order
of elution i.e. U followed by Th and both uranium and
thorium mandelate complexes interact hydrophobically
with the reversed phase surface and exhibit significant
retentions. A similar elution order was reported in the
case of cation exchange chromatography as uranium
forms weaker cationic species with the frequently used
eluents like hydrochloric and nitric acids [28]. This elu-
tion order may be useful in the present case only when
U(VI) has no or least interaction with stationary phase so
that it can get eluted at the solvent front. The realization
of this condition obliviously depends on the selection of
an appropriate stationary phase and a suitable complex-
ing agent. Ion exchange separation of metal ions can be
affected by adding weak organic acids in the mobile pha-
se, which reduces or alters the effective charge of the me-
tal ions by forming either cation, neutral or anion com-
plexes and the nature and stability of the metal com-
plexes would decide the method of ion exchange (anion
or cation) separation. Nowadays, stationary phases hav-
ing both cation and anion exchange capacities are found
to be more useful in separating heavy and transition me-
tal ions [29] because the separation is controlled by the
concentrations of free metal ions and the various insitu
metal complexes formed, which are in an equilibrium
with each other [30]. One such mixed ion exchanger
namely Ion Pac CS5A has been widely used for the sepa-
ration of lanthanides and transition metal cations in sev-
eral matrices [31-33].
Hence it is decided to examine Ion Pac CS5A mixed
ion exchanger column in the present investigation for the
separation of traces of Th from bulk of U as it 1) contains
both strong anion and cation exchangers (anion exchange
capacity: 40 µeq/column and cation exchange capacity:
20 µeq/column), and 2) displays anion and cation ex-
change properties while separating the metal cations by
choosing an appropriate complexing agent [34]. In the
present investigation, PDCA has been chosen as com-
plexing agent as it has high affinity for complex forma-
tion with polyvalent metal cations. The main advantages
in using PDCA as a complexing agent are: 1) it can be
used with low pH or high acid concentration and 2) the
concentration of PDCA required in the mobile phase to
facilitate the separation is very low (~105 M) compared
to the concentrations of the hydroxy carboxylic acids
(~101 M) required. Since the nuclear fuel samples (mg
amounts) are often dissolved in mineral acids and subse-
quently aliquots are made out of this solution for wet
analyses, it is desirable to use PDCA as complexing
agent, which would enable the direct injection of the
samples into the system.
The aim of this investigation was to develop a rapid
and reliable method for the separation and determination
of trace level thorium in uranium oxide using a mixed
ion exchanger column and PDCA as a complexing
2. Experimental
2.1. Instrumentation
The IC system used in the present study was DX500
(Dionex CA, USA). It consisted of a solvent delivery
pump (GP-50) and an absorption detector (AD20). A
pneumatic post column reagent addition facility consist-
ing of a 1000 mL reservoir bottle, gas pressure gauge to
control the flow rate of the gas, a T-junction and a 75 µL
capacity reaction coil was placed between the separator
column and the detector. The samples were injected
through a 50 µL loop fitted to a Rheodyne six port inject-
tor. The separator columns used were 1) Ionpac CS5A
(4.6 × 250 mm) with CG5A guard column (4.6 × 50 mm)
(Dionex CA, USA), 2) Ionpac AS11 HC (4.6 × 250 mm)
with AG11 guard column (4.6 × 50 mm) (Dionex CA,
USA) and 3) IC Pak Cation, 4.6 × 50 mm (Waters, USA).
The pH measurements were made on a CL-5 pH meter
(Toshniwal, India).
2.2. Reagents
All the reagents were of analytical grade (AR). Standard
stock solution of uranium was prepared by dissolving
nuclear grade U3O8 and its concentration was determined
by biamperometry method [35]. Standard stock solution
of Th (97.8 mg ml1) was prepared by dissolving nuclear
grade Th(NO3)4 (IRE, India) in 3 M HNO3 and the Th
concentration was determined by complexometric titra-
tion with EDTA using xylenol orange as indicator. 2,6-
pyridine dicarboxylicacid (PDCA) (99%, Aldrich Chemi-
cal corporation Inc.) was used as such. 0.01 M PDCA
solution was prepared in water and appropriate volume
of this solution was used in the preparation of mobile
phase. Nitric acid was of suprapure grade (Merck, India).
All solutions were prepared using high purity deionised
water (18.2 M) obtained from a Milli Q water system
(Millipore, USA). The eluents were filtered through ny-
lon 6,6, membrane filters (0.45 µm) prior to their use. A
post column reagent of 4.5 × 105M Arsenazo-III in 0.1N
HNO3 was added through a PEEK make T-junction with
a flow rate of 0.8 ml min1 for the detection of the sepa-
rated metal ions at 655nm.
3. Results and Discussion
3.1. Preliminary Investigation
Earlier we reported the separation of U and Th as their
PDCA complexes using cation exchange column with a
mobile phase of 0.08 mM PDCA in 0.24 M KNO3 and
Copyright © 2013 SciRes. IJAMSC
0.22 M HNO3 (pH ~ 0.6) [23]. A study that dealt with the
chelation ion chromatography (CIC) separation could
successfully separate Th(IV) and U(VI) on a neutral
polystyrene resin column dynamically modified with
PDCA [14]. Based on the reported studies, to start with,
an eluent composition of 0.1 mM PDCA in 1M KNO3
and 0.5 M HNO3 (pH ~ 0.3) was arbitrarily used with the
mixed ion exchanger column viz. Ionpac CS5A. With
this eluent, there can be three possible mechanisms in
which the metal complexes can interact with the station-
ary phase and they are: 1) cation-exchange 2) anion-ex-
change provided the metal ions form anionic complexes
with PDCA and 3) chelation ion chromatography. Inter-
estingly, unlike the modified RP column, the mixed ion
exchange column exhibited strong retention for Th (IV)
whereas U(VI) got eluted immediately after the solvent
front. Hence, the elution order observed in the present
case is U(VI) followed by Th (IV), which is similar to
that of the elution order obtained with strong cation ex-
change stationary phase [23,26,27]. This indicates that in
the case of mixed ion exchanger, the chelation exchange
mechanism responsible for retaining U (VI) and Th (IV)
on the RP substrate is insignificant, probably the mixed
ion exchanger column was not sufficiently modified with
PDCA to get the elution order similar to that of the RP
column. Since the observed elution order is similar to
that of the order obtained with cation exchange separa-
tion, it is necessary to study the role of anion exchanger
in the column, if any.
3.2. Effect of Concentration of PDCA
In order to understand the effect of PDCA concentration
on the retention of both U(VI) and Th(IV), concentration
of PDCA in the mobile phase was varied from 0.04 to 0.2
mM by keeping the concentrations of other two compo-
nents namely KNO3 (1 M) and HNO
3 (0.5 M)
0.04 0.08 0.12 0.16 0.20
Retention tim e
[P D C A] m M
Figure 1. Effect of concentration of PDCA on the retention
of U and Th separated on a mixed ion changer column. The
concentrations of KNO3 and HNO3 in the eluents were kept
as 1 M and 0.5 M respectively.
constant. The variation in pH of the eluent was insignifi
cant while varying the PDCA concentration because the
concentra- tion of HNO3 in the eluent was large com-
pared to the concentration of PDCA. The retention trend
obtained is shown in Figure 1. Increasing PDCA con-
centration in the mobile phase had virtually no effect on
the retention of U(VI) whereas the retention of Th(IV)
was drastically reduced.
The PDCA concentration was not increased beyond
0.2 mM because 1) the resolution between the two peaks
was poor and 2) the detection sensitivity of the Th(IV)-
arsenazo-III complex reduced significantly beyond 0.15
mM PDCA. The observed retention behavior of U(VI)
shows that it has least retention on the column and this
can occur only when U(VI) forms a stable neutral com-
plex with PDCA. This is because the formation of either
a cation or an anion complex of U(VI)-PDCA would
have exhibited significant retention since the column has
both cation and anion exchange sites. The equilibrium
distribution of PDCA (uncomplexed) species differs with
pH. PDCA is a dicarboxylic acid having pK1 = 2.10 and
pK2 = 4.68 for an ionic strength of 0.1 M [36]. It has
been reported that PDCA (H2L) predominantly exists as
HL form when the eluent pH 1 whereas at higher (al-
kaline) pH, it can be transformed into L2 form [34].
Since the eluent pH <1, the existence of HL form is
predominant than that of L2 form of the ligand. Hence,
U(VI) may be forming probably a 1:2 with HL or 1:1
complex with L2 form of PDCA. Though we cannot
offer any evidence for the formation of the neutral com-
plex(es), the strong retention of U(VI)-PDCA complex
on a reversed phase surface [14] supports this assumption.
Theoretically, though a neutral complex in equilibrium
with either its pure metal ions or other kind of complex
cations can have feeble interaction on cation exchange
sites provided the complex formed is kinetically labile. In
the present case, the complex formed has no interaction
with both anion and cation exchangers and this may be
due to high stability of the U-PDCA neutral complex
formed. Since there was no significant change in the re-
tention of U(VI) while increasing PDCA concentration in
the eluent, the formation of weak anionic complex of U-
PDCA was not appreciable in the present case.
In the case of Th(IV), it exhibited significant retention
on the mixed ion exchanger column. Elution with in-
creasing PDCA concentration in the mobile phase de-
creased the retention of Th(IV). The proposed cation
exchange models in literature [37] report that increasing
the concentration of complexing agent decreases the re-
tention time of metal ions due to their formation as neu-
tral or weak anionic complexes. Whereas in anion ex-
change model reports that only strong anion metal-ligand
complexes can interact with the anion exchange sites and
hence, increasing the concentration of complexing agent
Copyright © 2013 SciRes. IJAMSC
V. V. RAUT ET AL. 65
in the eluent would increase in the retention time. In or-
der to conclude about the probable nature of the metal
complexes and their interaction with ion exchange sites,
it is necessary to study the retention of U or Th with the
components of eluent individually.
Initially, separations were carried out by varying the
concentration of PDCA between 0.02 mM to 0.4 mM in
0.2 M HNO3 (necessary to avoid hydrolysis of U and Th)
without taking KNO3 in the mobile phase. It was ob-
served that with 0.02 mM PDCA, U appeared at 4.0
minutes whereas the same appeared at 2.7 minutes when
the PDCA concentration was increased to 0.1 mM and
subsequent increase of PDCA up to 0.4 mM did not
change the retention time significantly. In contrast,
Th(IV) peak did not appear at all. A similar observation
was made when pure HNO3 (0.5 to 2 M) or pure KNO3
(0.2 to 1.8 M) was used as eluent. Eluents with mixture
of HNO3 and KNO3 of varying concentrations were also
used and there was no difference in the retention of U(VI)
compared to that of the pure HNO3/KNO3. However,
interestingly it was observed that Th(IV) appeared when
the pH of the mixed eluent was less than 1. Further, it
was seen that the elution of thorium was so slow that it
resulted in a broad peak. Therefore, it may be concluded
that in the absence of PDCA in the eluent, the thorium
elution with KNO3+H NO3 may be due to the formation
of its nitrato complex and the formation of such nitrato
complex is pH dependent.
Therefore, it is seen from the experiments carried out
with PDCA, KNO3 and HNO3 individually as well as
with various combinations that for the separation of
Th(IV), for the purpose of analytical determination, a
mobile phase with the combination of all three compo-
nents is required. The observations showed that at this
low concentration of PDCA, the elution of Th(IV) oc-
curred only when a KNO3 is added along with minimum
0.1 M HNO3 concentration. It has been observed that
eluents with PDCA concentration above 0.08 mM were
not found suitable for separating trace Th in the presence
of large excess of uranium. Therefore, the PDCA con-
centration has been fixed at 0.075 mM. This concentra-
tion would reduce the interactions between UO2
complex with the stationary phase, which is desirable to
realize the elution of U peak immediately after or at the
excluded peak.
3.3. Effect of Concentration of KNO3
The retention trends of U(VI) and Th(IV) shown in Fig-
ure 1 suggest that U(VI) predominantly forms a neutral
complex and gets eluted immediately after the excluded
peak and Th(IV) forms weak neutral or anionic complex
and most probably, the mechanism responsible for the
retention of Th(IV) on the column is due to cation ex-
change. However, addition of relatively high concentra-
tion of KNO3 may suppress the ion exchange process due
to the presence of large concentration of K+ ions (acts as
competing cation). It was observed that the retention time
of U(VI) was unaltered whereas the retention time of
Th(IV) decreased when the concentration of KNO3 (0.2
to 1.8 M) was increased in the mobile phase and by
keeping the PDCA and HNO3 concentrations as 0.075
mM and 0.2 M, respectively.
The retention time of Th(IV) decreases drastically in
the concentration range of 0.2 to 1.2 M and subsequently
there was a little decrease up to 1.8 M. The drastic re-
ducetion in retention time may be attributed due to two
reasons: 1) increasing [KNO3] increases the ionic strength
of the eluent. While using cation-exchange column with
complexing eluent, the affinity of metal ions for ion-
exchange resins decreases as ionic strength increases and
therefore, the retention time decreases and 2) 3
from KNO3 may be competing with PDCA in complex
formation. To identify the role of KNO3, separations
were carried out with PDCA and HNO3 combinations
where the nitrate was supplied in the form of HNO3. For
instance, instead of 0.5 M HNO3 + 0.5 M KNO3 + 0.075
mM PDCA (pH 0.3) an eluent of 0.075 mM PDCA in 1
M HNO3 was prepared and pH was adjusted by adding
LiOH. It was seen that with the latter mobile phase com-
position, Th could not be eluted from the column. This
suggests that K+ is acting as competing ion which re-
duces the bare metal ions (U and Th) interactions with
ion exchange sites.
Equation for cation (complex ion) exchange is expre-
ssed as
loglogloglog log
DK anCnA
 
where Dm is the distribution coefficient of the metal in
presence of complexing agent, KMA is the selectivity co-
efficient between the metal and competing cation, aM(L) is
complex formation coefficient, C is the capacity of the
ion exchanger and [A+] is the concentration of the com-
peting cation.
The distribution coefficient (Dm) of the metal is a func-
tion of the concentration of the competing cation (A+)
and that of the complexing agent in the mobile phase.
Since in the present case the competing cations in the
mobile phase are K+ and H+, the elution mainly depends
on the concentration of K+ ions as it has high selectivity
over H+. On increasing the KNO3 concentration in the
eluent decreased the retention times of only Th(IV)
whereas the retention times of U(VI) was unaffected
throughout the concentration range (Figure 2(a)). Th(IV)
retention times decreased steadily as the KNO3 concen-
tration was raised up to 1.2 M and subsequent increment
of rise in KNO3 concentration did not reduce the reten-
tion times appreciably. As this trend is expected in the
case of cation exchange of Th(IV), a plot of log k (reten-
Copyright © 2013 SciRes. IJAMSC
0.0 0.4 0.81.2 1.6 2.0
Reten tion time
0.2 0.4 0.6 0.8 1.0 1.2
log k'
Figure 2. (a) Effect of concentration of KNO3 on the reten-
tion of U and Th separated on a mixed ion exchanger col-
umn (the concentrations of PDCA and HNO3 in the eluents
were 0.075 mM and 0.5 M respectively) and (b) plot of log k
(retention factor) of Th against [KNO3].
tion factor) against KNO3 concentration over the range of
0 - 1.2 M brought a straight line with negative slope
(Figure 2(b)) indicating the separation is predominantly
by cation exchange mechanism. The optimum KNO3
concentration was fixed as 1.0 M as this provided the
desired separation between U(VI) and Th(IV).
3.4. Effect of Concentration of HNO3
Addition of HNO3 and its concentration plays an impor-
tant role in controlling the effective ligand concentration
in the mobile phase, which decides the nature of metal-
complex. In addition, HNO3 prevents the hydrolysis of
metal cations as these heavy metal ions belong to acid
hydrolysable group. Moreover, the H+ ions from the ni-
tric acid also act as the competing ion during the elution
of metal cations or cation complexes. To investigate the
influence of HNO3 concentration on the elution, two sets
of experiments were conducted. In the first set, HNO3
concentration was varied over a range between 0.5 and
1.5 M.
It was expected that increasing acid concentration
would reduce the effective concentration of ligand, whi-
ch would affect the complexation of U(VI) leaving more
metal cations and in such case, the retention times for
both Th(IV) and U(VI) are expected to be increased.
However, there was no appreciable change in the reten-
tion times observed. Interestingly, there was a steady im-
provement in the symmetry of Th(IV) peak was observed.
In the second set, the pH of the mobile phase was varied
from 0.3 to 4.0. For preparing the eluents, dilute nitric
acid of appropriate concentration was used and final pH
adjustment was done with dilute LiOH solution. Increas-
ing eluent pH caused the retention times of Th(IV) to
decrease and of U(VI) to increase. In addition, the or-
der of elution reversed from pH 2 onwards (Figure 3).
There are two ways in which this can be explained. First,
increase of eluent pH leads to increase in the effective
ligand (HL) concentration and therefore, U(VI) tends
form anion complex, which interacts with anion ex-
changer resulting increasing trend in the retention. On the
other hand, it has not resulted in change in the complex
forming behavior of Th(IV) significantly and hence, re-
tention time decreased. There could also be a possibility
of column modification due to the hydrophobic interac-
tions between the phenyl group of PDCA and the poly-
stryrene-divinyl benzene skeleton of the resin. On in-
creasing pH of the eluent, the dissociation of acid groups
on the immobilized chelating ligand produces an increase
in the conditional stability constants of the surface metal
complexes and it causes the retention times of the metal
ions to increase [15]. Between Th(IV) and U(VI), it has
been reported that the U(VI)-PDCA (1:1) complex has
high stability constant (log K1 = 4.72) [38] and hence, it
is more retained than Th(VI). The existence of this
mechanism would have been confirmed only by carrying
out separations at higher pH levels. However, such stud-
ies are not affordable in this case as the hydrolysis of
heavy metal ions cannot be prevented.
From the analytical point of view, considering the de-
sired order of elution, separation factor and other chro-
matographic parameters, it is desirable to have higher
concentration of HNO3 and accordingly the HNO3 con-
centration was fixed as 1.2 M in the mobile phase.
3.5. Elution with a Pure Cation Exchange
Preceding investigations suggest that the elution behavior
of Th(IV) was mainly due to the cation exchange mecha-
nism. On the other hand U(VI) was eluted as its neutral
complex, which gets eluted immediately after the ex-
cluded peak indicating that it has least interaction with
both cation and anion exchangers. It has been seen that a
Copyright © 2013 SciRes. IJAMSC
V. V. RAUT ET AL. 67
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
log k' (retention factor)
Eluent pH
Figure 3. Effect of eluent pH on the retention of U and Th
on a mixed ion exchanger column. The [PDCA] and [KNO3]
were 0.075 mM and 1.0 M respectively.
Figure 4. Typical chromatogram obtained for a standard
solution of U(VI) and Th(IV) using cation exchanger col-
umn. Peaks (1) U(VI) – 3 ppm and (2) Th(IV) 3 ppm. Col-
umn: IC Pak Cation (4.6 × 50 mm). Eluent: 0.1 mM PDCA
and 0.12 M KNO3 in 0.1 M HNO3; Flow rate: 1 ml/min.
Detect io n: P ost col um n addi ti on of 4.5 × 1 0-5 M arsenazo-III
and detection at 656 nm.
mobile phase of 0.1 mM PDCA and 0.12 M KNO3 in 0.1
M HNO3 brought good separation between U(VI) and
Th(IV) and Figure 4 shows a typical chromatogram ob-
tained. The order of elution was as same as that of the
one obtained with the mixed ion exchanger column. Ef-
fects of concentrations of PDCA and KNO3 on the reten-
tion of Th(IV) showed similar trend in both mixed ion
exchange and cation exchange columns. This suggests
that the separation of Th(IV) is mainly due to cation ex-
change. In order to understand the separation with mixed
ion exchange column and the role of anion exchange, it is
necessary to carry out separations with pure anion ex-
change column too.
3.6. Separation with Pure Anion Exchange
When similar investigations were carried out with a pure
anion exchange column (IonPac AS11 HC column), the
elution order was reversed where Th peak appeared im-
mediately after the excluded peak followed by U. while
increasing the PDCA concentration (from 0.08 - 0.5 mM)
in the mobile phase it was observed that there were no
significant changes in the retention time of both U and
Th. This indicates that up to 0.5 mM PDCA both U and
Th are not forming stable anionic complexes. This ob-
servation suggests that Th may be forming weak anionic
and neutral complexes with PDCA. In the present case,
the appreciable retention of U cannot be explained on the
basis of formation of anionic complex because the reten-
tion time did not increase while increasing the PDCA
concentration. Probably the neutral U-PDCA complex
may be interacting on the surface of the column hydro-
phobically as it showed in the case of reverse phase sur-
face. Therefore, it may be concluded that the cation ex-
change sites of the mixed ion exchange column play
predominant role in the separation of U and Th. However,
the anion exchange sites present in the mixed ion ex-
change column may be reducing the hydrophobic inter-
action of U-PDCA neutral complex, though we cannot
offer any evidence.
3.7. Effect of Other Metal Ions
Since the uranium oxide fuel samples are expected to
have other metallic impurities like Fe, Cd, Cu, Ca, Mg,
Ni, Cr, Mn, Mo, rare earth elements etc. (within the spe-
cification limits), separations were carried out in the pre-
sence of these metal ions to observe the interference ef-
fects, if any, during the separation of Th. For this purpose,
two synthetic sample solutions were prepared and used.
The first solution was a mixture of thorium with transi-
tion metals (Ca and Mg were also added) and the second
one consisted thorium and mixture of all lanthanides. The
individual metal ion to Th concentration ratio was kept
ten in both the solutions. It has been observed that the
first solution did not give any peak whereas the second
solution containing mixture of lanthanides showed a
small triplet peak after the uranium peak with very poor
detection sensitivity. Therefore, the formulated eluent
composition was found to be suitable for the interference
free determination of Th. A typical chromatogram ob-
tained for a standard solution containing U and Th is
shown Figure 5.
3.8. Effect of Matrix ion on the Separation of Th
Under the optimized separation conditions, the uranyl ion
eluted very close to the excluded peak and the retention
capacity factor (k) for uranium was 0.2. Hence, sepa-
Copyright © 2013 SciRes. IJAMSC
Figure 5. Chromatogram obtained for a standard mixture
solution containing U and Th. Peaks (1) U(VI) 1 ppm and (2)
Th(IV) 2.5 ppm. Column: IonPac CS5A (4 × 250 mm). Elu-
ent: 0.075 mM PDCA and 1M KNO3 in 1.2 M HNO3. Flow
rate: 1 mL/min. Detection: Post column addition with 4.5 ×
10-5 M Arsenazo-III in 0.1 N HNO3 at 656 nm; flow rate: 0.8
ration of trace Th from bulk uranium may be feasible
because most of the uranium may get excluded along
with solvent front, which can reduce the possibility of
column swamping. Several synthetic samples in dupli-
cates having U/Th ratios 1, 2, 20, 200,400 and 1.8 × 104
were prepared and injected for Th separation. The obser-
ved recoveries for Th in each case have been listed in
Table 1 and the recoveries of Th were found to be more
than 90%, which showed that the method has high
tolerance for uranium and hence, the direct separation of
Th from the dissolved uranium samples is feasible.
3.9. Calibration and Method Validation
Since the success of an analytical method is critically
dependent upon how accurately the concentrations of the
analytes are known and how closely the matrix of the
standards match with that of the samples to be analyzed,
construction of a calibration plot in presence of matrix is
desirable as the instrumental response is read in an al-
most identical chemical environment for both samples
and calibration standards. This will compensate the ma-
trix effect on the measurements. Therefore, high purity
UO2 was dissolved in nitric acid and the solution was
made up to a known volume. Several aliquots were pre-
pared and varying volumes of a standard Th solution
were spiked except for one aliquot. The spiked aliquots
were made up to a known volume keeping a final con-
centration of U in all aliquots at 87 mg/mL. A calibration
plot was constructed between the peak areas and their
respective Th concentrations in the presence of matrix
uranium (87 mg of U/mL of the standard solution), which
showed a linear relation with a regression coefficient of
0.9975 and the plot obtained is shown in Figure 6.
The Th content of the uranium blank solution was
calculated from the intercept of the plot. The limit of
detection (LOD) was calculated as 0.1 ppm of Th using
S/N = 3 formula. The method was validated by analyzing
a U3O8 reference material (ILCE-IV) employing both the
present method and ICP-MS (a well established proce-
dure [9,39] for the separation of trace Th from the ura-
nium matrix prior to the ICP-MS analysis was followed).
Accurately weighed quantities (~0.4 g) of this U3O8 ref-
erence material were dissolved in Conc.HNO3 and heated
to dryness. Further they were re-dissolved in 5 mL of 0.5
M HNO3 and injected into the IC for separation. Another
set of dissolved samples were subjected to solvent ex-
traction for the removal of matrix U using aliquat-336
[40] and the final separated fractions were taken in 0.5 M
HNO3. The values obtained are listed in Table 2 and they
are in good agreement with the mean value reported by
the ICP-MS method. The chromatograms obtained for a
direct injection is shown Figure 7.
3.10. Procedure and Sample Analysis
A sample size of 0.4 - 0.5 g was dissolved in Conc.HNO3
Table 1. Effect of concentration of U (matrix) on the sepa-
ration of Th.
[Th] µg/mL
S-1 5 5 1 4.9 98.0
S-2 10 5 2 4.78 95.6
S-3 100 5 20 4.81 96.2
S-4 1000 5 200 4.85 97.0
S-5 2000 5 400 4.65 93.0
S-6 88,000 5 1.8 E4 4.73 94.6
R^2 = 0.99751
B = 118.87797
Peak A rea
Conc. of Th in ppm
Figure 6. Calibration plot obtained for Th over a concen-
tration range of 0.25 - 10 ppm in the presence of 87 mg of U/
mL of the standard solutions.
Copyright © 2013 SciRes. IJAMSC
V. V. RAUT ET AL. 69
Table 2. Results obtained for a U3O8-refer ence material.
[Th4+] by IC
(ppmw)b Mean Mean reported
D-1 75.5
D-2 72.7
D-3 74.6
D-4 71.8
D-5 75.2
74.0 ± 1.6
72.5 ± 2.2
S-1 72.3
S-2 69.8 71.4 ± 1.3
S-3 72.7
S-4 70.9
a0.5 g of sample dissolved and made up to ~20 g by 0.1N HNO3 and 50 µL
of this solution was injected into IC. beach value quoted is a mean ob-
tained from three replicate injections. c0.5 g of sample was dissolved in HCl
and U was separated by aliquat-336 [40]in 6 M HCl.
Figure 7. Typical chromatograms obtained for a U3O8-Re-
ference material sample (A) direct injection of dissolved
uranium solution (B) sample injected after removal of ura-
nium (matrix) by solvent extraction.
and evaporated to dryness. Further, it was re-dissolved in
0.5 M HNO3 to have a final volume of 5 mL and 50 µL
of this solution was injected into IC for separation. A
mobile phase of 0.075 mM PDCA and 1M KNO3 in
1.2M HNO3 was used at a flow rate of 1 mL/min
forseparation. The separated fraction of metals are de-
tected and measured by post column addition of 4.5 ×
104 M arsenazo-III at a flow rate of 0.8 mL/min. The
results obtained for typical samples are listed in Table 3.
4. Conclusion
The proposed method was successfully used for the de-
termination of trace levels of Th in uranium matrix with-
out employing prior separation of matrix. The method is
simple, rapid, accurate and promising for its adaptation
in the routine quality control analysis of uranium based
Table 3. Results obtained for uranium samples.
Sample Conc. of Th (ppmw)a
UO2 – SG 18.3 ± 1.3
U3O8 – NFC-1 15.7 ± 1.4
U3O8 – NFC-2 23.6 ± 1.7
amean of three analyses.
5. Acknowledgements
The authors thank Dr. B. S. Tomar, Head, Radioanalyti-
cal Chemistry Division, BARC for his support. They also
thank Mr. V. G. Mishra, RACD for his help.
[1] H. J. Bhabha and N. B. Prasad, “A Study of the Contribu-
tion of Atomic Energy to a Power Programme in India,”
Proceedings of the Second United Nations International
Conference on the Peaceful Uses of Atomic Energy, Ge-
neva, Vol. 1, September 1958, pp. 89-101.
[2] M. Dayal, S. R. Paranjape, B. Singh and N. B. Prasad,
“Study of Fuel Cycles with Reference to a Power Pro-
gramme,” Proceedings of the Second United Nations In-
ternational Conference on the Peaceful Uses of Atomic
Energy, Geneva, Vol. 13, September 1958, pp. 184-194.
[3] T. K. Mukherjee, “Characterization and Quality Control
of Nuclear Fuels (CQCNF-2002),” Allied Publishers,
New Delhi, 2003.
[4] IAEA-TECDOC-1450, “Thorium Fuel Cycle-Potential
Benefits and Challenges”, IAEA, Austria, 2005, pp. 68-
[5] V. Venugopal, “Chemical Quality Control of Nuclear
Materials,” Indian Association for Nuclear Scientists and
Allied Chemists (IANCAS) Bulletin, Vol. 4, Mumbai,
[6] S. Joannon, P. Telouk, C. Pin, “Determination of U and
Th at Ultra-Trace Levels by Isotope Dilution Inductively
Coupled Plasma Mass Spectrometry Using a Geyser-
Type Ultrasonic Nebulizer: Application to Geological
Samples,” Spectrochimica Acta B, Vol. 52, No. 12, 1997.
pp. 1783-1789.
[7] A. G. Adriaens, J. D. Fassett, W. R. Kelly, D. S. Simons
and F. C. Adams, “Determination of Uranium and Tho-
rium Concentrations in Soils: Comparison of Isotope Di-
lution-Secondary Ion Mass Spectrometry and Isotope Di-
lution-Thermal Ionization Mass Spectrometry,” Analy-
tical Chemistry, Vol. 64, No. 23, 1992, pp. 2945-2950.
[8] E. H. Borai and A. S. Mady, “Separation and Quantifica-
tion of 238U, 232Th and Rare Earths in Monazite Samples
by Ion Chromatography Coupled with On-Line Flow Scin-
tillation Detector,” Applied Radiation and Isotopes, Vol.
57, No. 4, 2002. pp. 463-469.
[9] S. B. Deb, M. K. Saxena, B. K. Nagar and K. L. Rama-
Copyright © 2013 SciRes. IJAMSC
kumar, “Determination of Trace Amounts of Thorium in
a Uranium Matrix by Inductively Coupled Plasma Mass
Spectrometry and Validation of the Separation Procedure
by Standard Addition and Tracer Techniques,” Atomic
Spectroscopy, Vol. 29, No. 2, 2008, pp. 39-44.
[10] F. Hao, B. Paull and P. R. Haddad, “Determination of
Thorium and Uranyl in Nitrophosphate Solution by On-
Line Matrix-Elimination Reversed-Phase Chromatogra-
phy,” Chromatographia, Vol. 42, No. 11-12, 1996, pp.
[11] F. Hao, B. Paull and P. R. Haddad, “Determination of
Trace Levels of Thorium(IV) and Uranyl by Reversed-
Phase Chromatography with On-Line Preconcentration
and Ligand Exchange,” Journal of Chromatography A,
Vol. 749, No. 1-2,1996, pp. 103-113.
[12] R. M. Cassidy, “Determination of Rare-Earth Elements in
Rocks by Liquid Chromatography,” Chemical Geology,
Vol. 67, No. 3-4, 1998. pp. 185-195.
[13] P. E. Jackson, J. Carnevale, H. Fuping and P. R. Haddad,
“Determination of Thorium and Uranium in Mineral Sands
by Ion Chromatography,” Journal of Chromatography A,
Vol. 671, No. 1-2, 1996, pp. 181-191.
[14] R. M. C Sutton, S. J. Hill, P. Jones, A. Sanz-Medel and J.
I. Garcia-Alonso, “Comparison of the Retention Behav-
iour of Uranium and Thorium on High-Efficiency Resin
Substrates Impregnated or Dynamically Coated with
Metal Chelating Compounds,” Journal of Chromatogra-
phy A, Vol. 816, No. 2, 1998, pp. 286-291.
[15] R. M. Cassidy and M. Fraser, “Equilibria Effects in the
Dynamic Ion-Exchange Separation of Metal Ions,” Chro-
matographia, Vol. 18, No. 7, 1984, pp. 369-373.
[16] P. Jones and P. N. Nesterenko, “High-Performance Che-
lation Ion Chromatography a New Dimension in the Se-
paration and Determination of Trace Metals,” Journal of
Chromatography A, Vol. 789, No. 1-2, 1997, pp. 413-435.
[17] Ch. S. K. Raju, M. S. Subramanian, N. Sivaraman, T. G.
Srinivasan and P. R. Vasudeva Rao, “Retention Studies
on Uranium, Thorium and Lanthanides with Amide Mo-
dified Reverse Phase Support and Its Applications,” Jour-
nal of Chromatography A, Vol. 1156, No. 1-2, 2007, pp.
[18] M. Akhila Maheswari, D. Prabhakaran, M. S. Subrama-
nian, N. Sivaraman, T. G. Srinivasan and P. R. Vasudeva
Rao, “High Performance Liquid Chromatographic Studies
on Lanthanides, Uranium and Thorium on Amide Modi-
fied Reversed Phase Supports,” Talanta, Vol. 72, No. 2,
2007, pp. 730-740.
[19] F. Hao, B. Paull and P. R. Haddad, “Retention Behaviour
of Thorium(IV) and Uranyl on a Reversed-Phase Column
with Glycolate and Mandelate as Eluents,” Journal of
Chromatography A, Vol. 739, No. 1-2, 1996, pp. 151-161.
[20] M. J. Shaw, S. J. Hill and P. Jones, “Chelation Ion Chro-
matography of Metal Ions Using High Performance Sub-
strates Dynamically Modified with Heterocyclic Carbox-
ylic Acids,” Analytica Chimica Acta, Vol. 401, No. 1-2,
1999, pp. 65-71.
[21] P. Jones, “Major Sensitivity Improvements in Ion Chro-
matography Determinations Involving Post-Column Spe-
ctrophotometric Reaction Detectors through Elimination
of Pump Noise Using a Dual Wavelength Monitoring
Procedure,” Analyst, Vol. 125, No. 5, 2000, pp. 803-806.
[22] M. J. Shaw, S. J. Hill, P. Jones and P. N. Nesterenko,
“Determination of Uranium in Environmental Matrices by
Chelation Ion Chromatography Using a High Performance
Substrate Dynamically Modified with 2,6-Pyridinedi-
carboxylic Acid,” Chromatographia, Vol. 51, No. 11-12,
2000, pp. 695-700.
[23] S. Jeyakumar, V. G. Mishra, M. K. Das, V. V. Raut, R. M.
Sawant and K. L. Ramakumar, “Separation Behavior of
U(VI) and Th(IV) on a Cation Exchange Column Using
2,6-Pyridine Dicarboxylic Acid as a Complexing Agent
and Its Application for the Rapid Separation and Deter-
mination of U and Th by Ion Chromatography,” Journal
of Separation Science, Vol. 34, No. 6, 2011, pp. 609-616.
[24] J. Cowan, M. J. Shaw, E. P. Achterberg, P. Jones and P.
N. Nesterenko, “The Ion Chromatographic Separation of
High Valence Metal Cations Using a Neutral Polystyrene
Resin Dynamically Modified with Dipicolinic Acid,” Ana-
lyst, Vol.125, No. 12, 2000, pp. 2157-2159.
[25] “Determiantion of Uranium and Thorium in Complex
Matrices Using Chelation Ion Chromatography,” Dionex
Application Note, No. 79, 1998, LPN 034670-01.
[26] A. W. Al-Shawi and R. Dahl, “Determination of Thorium
and Uranium in Nitrophosphate Fertilizer Solution by Ion
Chromatography,” Journal of Chromatography A, Vol.
706, No. 1-2, 1995. pp. 175-181.
[27] M. P. Harrold, A. Siriraks and J. Riviello, “High-Perform-
ance Ion Chromatographic Separation of Uranium and
Thorium in Natural Waters and Geological Materials,”
Journal of Chromatography A, Vol. 602, No. 1-2, 1992,
pp. 119-125.
[28] F. Nelson, T. Murase and K. A. Kraus, “Ion Exchange
Procedures. I. Cation Exchange in Concentration HCl and
HClO4 Solutions,” Journal of Chromatography A, Vol. 13,
1964, pp. 503-535.
[29] H. Hojabri, A. G. Lavin, G. G. Wallace and J. M. Riviello,
“Determination of Metal Ions Using Ion Chromatography
and Indirect Amperometric Detection,” Analytical Che-
mistry, Vol. 59, No. 1, 1987, pp. 54-57.
[30] Y. S. Park, S. H. Han, K. S. Joe, T. Y. Eom and G. Lee,
“Retention Behaviour of Transition Metal Ions with Some
Copyright © 2013 SciRes. IJAMSC
Copyright © 2013 SciRes. IJAMSC
Complexing Agents on Cation Exchanger,” Bulletin of
the Korean Chemical Society, Vol. 14, No. 6, 1993. pp.
[31] N. Cardellicchio, P. Ragone, S. Cavalli and J. Riviello,
“Use of Ion Chromatography for the Determination of
Transition Metals in the Control of Sewage-Treatment-
Plant and Related Waters,” Journal of Chromatography A,
Vol. 770, No. 1-2, 1997, pp. 185-193.
[32] E. H. Borai, M. A. Eid and H. F. Aly, “Determination of
REEs Distribution in Monazite and Xenotime Minerals
by Ion Chromatography and ICP-AES,” Analytical and
Bioanalytical Chemistry, Vol. 372, No. 4, 2002, pp. 537-
[33] L. Perna, F. Bocci, L. Aldave de las Heras, J. De Pablob
and M. Betti, “Studies on Simultaneous Separation and
Determination of Lanthanides and Actinides by Ion Chro-
matography Inductively Coupled Plasma Mass Spectro-
metry Combined with Isotope Dilution Mass Spectrome-
try,” Journal of Analytical Atomic Spectrometry, Vol. 17,
9, 2002, pp. 1166-1171.
[34] P. Janvion, S. Motellier and H. Pitsch, “Ion-Exchange Me-
chanisms of Some Transition Metals on A Mixed-Bed
Resin with a Complexing Eluent,” Journal of Chroma-
tography A, Vol. 715, No. 1, 1995, pp. 105-115.
[35] M. Xavier, P. R. Nair, K. V. Lohithakshan, S. G. Marathe
and H. C. Jain, “Determination of Uranium in the Pres-
ence of Iron and Plutonium by Ti(III) Reduction and
Biamperometric Titration,” Journal of Radioanalytical
and Nuclear Chemistry, Vol. 148, No. 2, 1991, pp. 251-
[36] L. G. Sillen and A. E. Martell, “Stability Constants of
Metal Ion Complexes, spl,” Chemical Society, London,
1964, p. 545.
[37] P. R. Haddad and R. C. Foley, “Modelling of Cation Re-
tention in Ion Chromatography Using Fixed-Site and Dy-
namically Coated Ion-Exchange Columns,” Journal of
Chromatography A, Vol. 500, 1990, pp. 301-312.
[38] A. E. Martell and R. M. Smith “Critical Stability Con-
stants,” Vol. I, Amino Acids, Plenum, New York, 1974, p.
[39] “DAE-Interlaboratory Comparison Experiment (ILCE)
for Trace Metal Assay of Uranium,” Phase 2, Document,
BARC, India, 2002.
[40] V. R. Bhandiwad, R. Swarup and S. K. Patil, “Extraction
of Actinides by Quaternary Amines from Hydrochloric
Acid Medium,” Journal of Radioanalytical Chemistry,
Vol. 52, No. 1, 1979, pp. 5-14.