American Journal of Analyt ical Chemistry, 2011, 2, 46-55
doi:10.4236/ajac.2011.21005 Published Online February 2011 (http://www.SciRP.org/journal/ajac)
Rapid Separation and Quantification of Iron in Uranium
Matrix by Capillary Zone Electrophoresis (CZE)
Vivekchandra Mishra, Mrinal Kanti Das, Subbiah Jeyakumar, Ramesh Mahadeo Sawant,
Karanam Lakshminarayana Ramakumar
Radioanalytical Chemistry Division, Bhabha Atomic Resea rch Centre, Mumbai, India
Received August 12, 2010; revised November 30, 2010; accepted December 2, 2010
Abstract
A method was developed for rapid separation and determination of iron by employing capillary zone elec-
trophoresis (CZE) technique with direct UV detection. Iron could be separated from matrix uranium by di-
rect injection of dissolved sample solution into capillary using a mixture of 10 mM HCl and 65 mM KCl (pH
= 2) as background electrolyte (BGE) at an applied voltage of 15 kV. The developed method has a very high
tolerance for the matrix element U (100 mg/mL) and as such may not need prior separation of iron from the
matrix. Iron could be separated with better than 95% recovery. The method showed a linear calibration over
a concentration range 1-50 ppm of Fe(III). The migration times for the iron peak were reproducible within
1% for both pure Fe(III) and in presence of matrix uranium (80 mg/mL). The precision (RSD, n = 22) of
peak area obtained for 1 ppm of iron was 3.5%. The limit of detection (LOD) (3) was 0.1 ppm and the ab-
solute LOD was 9 × 10-14 g considering the sample injection volume of 1.5 nL. The developed method has
been validated by separating and determining iron in two certified reference materials of U3O8. The method
was applied for the determination of iron in different uranium based nuclear materials. The CZE method is
versatile for routine analysis as it is simple, rapid and has simple sample preparation procedure.
Keywords: Capillary Zone Electrophoresis, Iron, Uranium, Nuclear Industry, Specification Analysis
1. Introduction
Separation and determination of trace elements in nu-
clear materials is important in nuclear industry as their
performance in the reactor critically depends on their
purity [1]. During the initial stages of nuclear fuel fabri-
cation, the designers require the purity check analyses of
the unfinished fuel with respect to a few elements like Fe
and Ca [2]. Depending upon the concentration of iron,
the designers are required to take adequate measures to
achieve the desired purity.
Several methods are available for the determination of
iron in nuclear materials. The methods involving ICP-
AES [3,4], ICP-MS [5], spectrophotometry [6], ion chro-
matography [7,8] and HPLC [9] are well known. A dis-
advantage of these methods is the associated laborious
and time consuming sample preparation procedures and
the need for prior separation of matrix uranium using
solvent extraction or ion exchange before carrying out
the measurement. Also analysis of a matrix-matched
reference material every time along with samples is in-
dispensable to confirm the quantitative recovery of iron.
This necessitates a simple, reliable, rapid and specific
method to minimize the labour involved and to achieve
maximum sample throughput.
In this context, the application of CZE for metal ion
determination deserves increased attention due to its
speed, high separation efficiency, resolving power, mini-
mal sample and reagent requirements. Ever since the first
paper concerning the indirect detection of inorganic ions
by using CE in 1967 [10], the application of CE to the
separation and determination of inorganic substances has
increased rapidly and many papers have been published
and some of critical reviews have also been summarized
[11-15]. Different approaches have been reported to real-
ize the full potential of CE for separation of metal ions by
CE. These include employing non-aqueous media [16,17],
use of partial or complete complexation techniques [18,
19], using CE in conjunction with ICP-MS [20-22] and
use of quantitative microchip CE [23]. Despite the in-
creased applications to inorganic materials, the applica-

V. MISHRA ET AL.
47
tion of CE in nuclear industry has been somewhat limited
[24,25].
The separation principle of CE is based on the differ-
ential electrophoretic mobility of charged compounds.
The differences in the mobility of the analytes are related,
in turn, to their charge densities, i.e., the charge-to-mass
ratio. One of the problems in the analysis of metal ions
by CE is that most of transition metal cations have al-
most the same mobility due to their similar size and
identical charge. Obviously, the enhancement of separa-
tion selectivity is the only alternative to achieve a satis-
factory resolution. Generally, there are two main ap-
proaches in this direction that imply the addition of a
complexing ligand to either the carrier electrolyte or a
sample solution before introduction into the capillary
[26]. In the first case, the mobility of sample cations to-
ward the cathode can be selectively moderated due to the
partial complexation within the capillary, followed by
the formation of metal complexes of different stability
and thereby effective charge [27,28]. The second ap-
proach provides the complete pre-capillary or on- capil-
lary conversion of metal ions into stable, charged com-
plexes, which can move with different mobilities de-
pending on their charge, size and stability [29]. The latter
approach was exploited in the present investigation.
Several CZE methods have been reported for the de-
termination of iron in various matrices like water, elec-
troplating baths and cyanide complexes [30-34]. CZE
separation of Fe(II) and Fe(III) was demonstrated after
complexing with o-phenanthroline and EDTA respec-
tively and the same indicates separation feasibility of
Fe(III) as its EDTA complex [35]. Another study reports
the separation of Fe(III) as its DTPA complex [36]. In
these reported methods, for the determination of either
total iron or speciation, Fe-aminocarboxylic acid anionic
complex was separated from other metal-aminocarbo-
xylic acid anionic complexes. It would be difficult to
apply these methods directly for the nuclear materials
like uranium compounds because U(VI) also forms ani-
onic complex with aminocarboxylic acids. Between the
aminocarboxylic acid complexes of U(VI) and Fe(III),
the net negative charge of the U(VI) complex is more
than that of the Fe(III)-complex (eg. Fe(III) [33] and
U(VI) [37] with EDTA form Fe-EDTA- and
UO2-EDTA2- respectively). Due to the higher charge on
the uranyl complex, the migration order follows uranium
and then iron and large amount of uranium in the sample
solution produces a bulk and long tailing peak, possibly
masking the iron peak or affecting its recovery. It is
worth mentioning that while separating the metal-ami-
nocarboxylic acid anionic complexes, it would be diffi-
cult to achieve a large separation factor to overcome the
problem of peak masking. Moreover, in these methods
higher pH electrolyte media (typically pH 6-10) are used
for the effective coating of CTAB on the capillary sur-
face and subsequent formation of double layer necessary
for reversing the EOF. Under such pH conditions, the
bulk uranium will undergo hydrolysis significantly and
severely obstruct the movements of ions inside the capil-
lary path. This is because unlike the hydroxycarboxylic
acids, the aminocarboxylic acids cannot prevent the hy-
drolysis of metals at higher pH. Moreover, the pH of the
sample cannot be kept low (pH < 3) as effective com-
plexation does not occur between metal and aminocar-
boxylic acids.
The aim of the work was to develop a method for the
separation and quantification of iron from uranium. As
direct injection of the sample solution into capillary is
envisaged, significant reduction in analysis time is possi-
ble which is desirable during process control operations.
At the same time, interferences from other metal cations
can be eliminated due to the anionic complex formation
with chloride ligand.
2. Experimental
2.1. Instrumentation
Separations were performed on a commercial capillary
electrophoresis apparatus (Prince Technologies, CEC-770,
Netherlands) equipped with a photodiode array detector.
Fused silica capillaries of 50 µm i.d. and 60 cm long
were used. A capillary having 75 µm i.d. and 60 cm long
was also used. Sample injection was done in hydrody-
namic mode by the application of 50 mbar pressure for
0.2 min on the sample vial which corresponds to a sam-
ple volume of 1.5 nL. System DAX software was used
for data acquisition. Direct UV detection was performed
at 214 nm. All the experiments were conducted at room
temperature (25-27˚C). The pH measurements were done
with a pH meter (Eutech, Tutor-model, Malaysia).
2.2. Reagents and Solutions
Standard stock solution of Fe(III) was prepared by dis-
solving Fe(NO3)3•9H2O (99.99%, Aldrich Chemicals,
USA) in 0.01N HNO3. Subsequently, the working stan-
dards were made from the stock by appropriate dilutions.
High purity HCl and HNO3 acids (suprapure, MERCK,
Germany) were used for the sample and electrolyte
preparations. Potassium chloride (GR grade, MERCK,
Germany) was used. Nuclear grade UO2 (NFC, India)
was used for the preparation of standard uranium solu-
tion and the concentration of U was obtained from biam-
perometric determination [38]. All solutions, electrolytes

48 V. MISHRA ET AL.
and standard solutions were prepared with ultrapure wa-
ter (18 MΩ) obtained from a MilliQ-Academic System
(Millipore, India).
2.3. Procedure for Conditioning Capillary and
Sample Injection
Procedures for conditioning capillary, rinsing with elec-
trolyte, hydrodynamic injection of the samples and re-
rinsing were programmable and these are defined for
sequential execution. Briefly, the capillary was washed
with 0.2 M NaOH for 15 min and with water for 10 min
followed by rinsing with BGE for 15 min. About 1.5nL
of the sample solution was injected into the capillary
hydrodynamically by applying 50 mbar pressure for 0.2
min. A potential of 15 kV was applied during the sample
run and in between two sample runs, the capillary was
again rinsed with BGE for 5 min to minimize memory
effect, if any.
3. Results and Discussion
Generally, capillary zone electrophoresis (CZE) is em-
ployed for the separation of metal ions with comparable
concentrations. Separating trace amounts of a metal
cation present in another metal matrix is a difficult task
in any high performance separation technique. However,
achieving such separations offer advantages like hassle
free sample preparation viz. no need of matrix separation,
no uncertainty on analyte recovery and reduction in
overall analysis time. Since the dissolved uranium sam-
ple solution will have high ionic strength, the direct in-
jection of sample into the capillary can cause certain dif-
ficulties such as 1) undesirable peak broadening may
occur if the conductance of sample zone (λs) becomes
higher than the conductance of the BGE (λB), 2) change
in the zeta potential and variation in the magnitude of
EOF occur due to adsorption of matrix cation on the sur-
face of capillary and altering the migration time of all
species significantly, 3) poor precision on the migration
time due to the delayed desorption of the matrix cations
resulting in long tailing or asymmetry, 4) poor separation
efficiency with band broadening (the plate height in-
creases when sample concentration is more) due to over-
loading of the capillary with matrix ions and 5) severe
efficiency loss due to large difference in viscosities of
sample and BGE.
Despite these difficulties, direct separation of trace
analyte in presence of bulk matrix is feasible in CZE
provided the method satisfies two conditions: 1) the ana-
lyte and the matrix elements should have large difference
in their relative mobilities and 2) the mobility of the
analyte should be faster than matrix element. Under these
conditions, the analyte reaches the detector much earlier
than the matrix element and therefore, it will be free
from matrix effects. In the present case, since both uranyl
and ferric ions (in their hydrated form) have little differ-
ence in their charge densities, it is difficult to separate
them under normal conditions and addition of a com-
plexing agent is essential for their separation. The or-
ganic ligands reported in literature for complexing Fe(III)
in CE studies [30-36,39] may not be suitable in the pre-
sent case due to the reasons already mentioned above.
Hence, it was proposed to consider chloride as ligand as
well as carrier electrolyte since its complexation ability
with UO2
2+ and Fe3+ ions is different and they tend to
form cationic complexes of different charges. Moreover,
the chloride medium enables the direct detection of se-
lected metal ions [40,41] including iron.
At lower concentrations of chloride ion (<0.5 M), the
uranyl ion predominantly forms UO2Cl+ complex [42,43]
in the pH range 2-3. Further, the formation of anionic
complexes viz. UO2Cl3
-, UO2Cl4 2-, and UO2Cl5 3- occurs
only when the chloride concentration exceeds 5 M.
Similarly, though Fe(III) forms FeCl2+, FeCl2
+, FeCl3 and
FeCl4
- complexes, only FeCl2+ is predominant at low
chloride concentrations [44]. The formation constants
(log values) for the [UO2Cl]+ and [FeCl]2+ complexes
(0.17 and and 1.52 respectively) show that the Fe+3 has
more affinity towards chloride complex formation [43,45]
than UO2
2+. Therefore, at low chloride concentrations
UO2
2+ and Fe+3 form [UO2Cl]+ and [FeCl]2+ complexes
respectively and their electrophoretic mobilities will
change significantly due to charge differences and their
separation from each other may become feasible. Com-
plex formation between chloride and divalent transition
metal ions such as Zn(II), Cu(II), Cd(II), Mn(II) etc., is
readily achieved and the complexes are remarkably sta-
ble [46]. These complexes are either anionic or neutral in
nature [47] and do not cause any interference. The ab-
sorbance of the peak for iron as [FeCl]2+ complex was
measured at 214 nm [45].
3.1. Optimization of Background Electrolyte
A mixed solution of HCl and KCl was selected as BGE.
The HCl solution is the carrier electrolyte and it prevents
the hydrolysis of iron, uranium and their chloride com-
plexes by providing lower pH whereas KCl provides
adequate chloride (ligand) ions inside the capillary to
prevent the decomposition of complexes. All the separa-
tions were carried out with 15 kV applied voltage.
Initially the separation of Fe(III) was performed with
10-2, 10-3
, 10-4 and 10-5 M HCl solutions corresponding to
pH 2, 3, 4 and 5 without KCl. This was done to explore

V. MISHRA ET AL.
49
the feasibility of separating iron with low ionic strength
electrolytes, which minimize joule heat generation. The
standard iron solutions were also prepared in the respec-
tive HCl solutions in order to have almost identical
composition in the BGE and standards. Figure 1 shows
the overlay of electropherograms obtained for each car-
rier electrolyte. It is seen from the figure that decreasing
the pH of the carrier electrolyte decreased the migration
time of Fe(III) peak with substantial peak broadening.
This was unexpected because increasing pH of the car-
rier electrolyte typically causes reduction in the migra-
tion time. However, the change in migration time in this
case may be possibly due to following three reasons.
Firstly, on increasing the pH, the silanol group of the
capillary dissociates considerably resulting in negative
charges on the surface of fused silica. These negatively
charged sites attract the free metal ions of iron. Hence,
transient retention of iron on the capillary surface occurs,
which leads to increase in migration time, band broad-
ening and loss in sensitivity. Secondly, at higher pH so-
lutions the extent of complex formation is affected due to
lower concentrations of chloride. Thirdly, when the pH is
above 2, the hydrolysis of Fe(III) and its chloride com-
plexes may probably occur and form neutral species (hy-
droxides) as reported in the literature [44].
The influence of electrolyte pH on the separation did
not result in any definite conclusion because the total
chloride concentration at each pH condition was not kept
constant. Hence, the influence of pH on the separation of
iron at constant chloride concentration and the effect of
chloride concentration under different pH conditions
were examined. For this, three sets of solutions corre-
sponding to pH 2, 3 and 4 were prepared. In each set, the
pH was maintained at the pre-selected value by keeping a
fixed concentration of HCl but the total chloride content
was varied from 30 to 100 mM by adding calculated
amount of KCl. With pH 2 electrolyte, on increasing
chloride concentration, the migration time of iron in-
creased slightly but not significantly and at the same time,
the corresponding peak area increased up to a chloride
concentration of 75 mM (10 mM HCl + 65 mM KCl).
Further increase in chloride concentration caused only
marginal changes in the peak areas (Figure 2). On the
other hand, electrolytes of pH 3 and pH 4 brought about
severe base line drift and huge noise when the chloride
concentration was increased.
During the separation, there may be a little change in
the electrolyte pH due to applied potential and protona-
tion of silinol groups. This change can alter the migration
time of iron. However, this effect was found to be insig-
nificant when iron was separated with electrolytes of pH
1.8, 2.0 and 2.2 as these electrolytes brought about mi-
gration time of iron peak within 1% precision (RSD, n =
5 in each pH condition). Based on these, a BGE of 10
mM HCl and 65 mM KCl (pH 2) was chosen for further
studies.
Figure 1. Electropherograms obtained for a standard iron solution (25 mg•L-1 ) with BGEs of different pH. 1). 10-2 M HCl
BGE corresponding to pH 2; 2). 10-3 M HCl BGE corresponding to pH 3; 3). 10-4 M HCl BGE corresponding to pH 4 and 4).
10-5 M HCl BGE corresponding to pH 5, Conditions: Capillary: 60 cm × 50 µm, Applied voltage: 15 kV. Detection: direct UV
t 215 nm. a

V. MISHRA ET AL.
50
Figure 2. Effect of total chloride concentration on the peak
area of iron obtained with pH 2. (BGEs were of 10-2 M HCl
(fixed) and varying amounts of KCl). Capillary: 60 cm × 50
µm, Applied voltage: 15 kV; Detection: direct UV at 214
nm.
The migration of the matrix ion, UO2
2+ and other
common ions such as Cu(II), Ni(II), Zn(II), Mn(II),
Co(II), Cd(II), Cr(III), Al(III), Ca(II), Mg(II), Sr(II),
Ba(II), Na and K, which are expected to be present in the
actual samples in very low concentrations, were investi-
gated. The uranyl ion (as UO2Cl+) appeared much later
than iron peak whereas the transition, alkali and alkaline
earth metal cations could not be detected even up to 60
min. at the entire range of absorption measurements from
214 nm to 240 nm. The probable explanation for this is
the migration of these metal complexes (mainly in ani-
onic form) towards the anode and also the EOF at pH 2
is expected to be significantly small.
The peaks of Fe(III) and U(VI) were confirmed by in-
jecting the standard mixture solutions having varying
concentrations.
3.2. Optimization of Applied Voltage
The relation between efficiency (N) and applied voltage
is expressed as, N = µepV/2DS where N is efficiency, V is
applied voltage and Ds is the diffusion co-efficient [48].
Higher values of N can be achieved by the application of
high voltage because it reduces the dispersion time of the
sample zone. Hence, the effect of applied voltage on the
mobilities of iron and uranium ions was investigated by
varying the potential from 5 to 30 kV. There was a con-
siderable change in the mobilities of Fe and U and their
peak reproducibility when the potential was raised be-
yond 20 kV. It was also seen that at higher voltages the
resolution between Fe and U was poor due to peak
broadening. A plot of current Vs applied voltage showed
that the current and voltage was non linear beyond 18 kV
indicating the inability of capillary to dissipate the Joule
heat. Therefore, an applied voltage of 15 kV was adopted
as optimal. Typical electropherograms obtained for
Fe(III) of different concentrations under the optimized
conditions have been shown in Figure 3.
3.3. Matrix Element Tolerance
The tolerance of matrix element on the separation and
determination of Fe(III) was investigated by spiking 50
ppm of Fe(III) into uranium solutions of different con-
centrations (0.1-90 mg of U/mL). An un-spiked uranium
solution was treated as blank. A standard solution con-
taining 50 ppm of Fe(III) was taken as a calibration
standard for determining concentration of Fe(III) in the
iron-spiked uranium solutions. The values obtained for
the spiked solutions are in good agreement with the ex-
pected value and the overall precision and accuracy was
better than 5% as shown in Table 1. The studies also
show the recovery of iron in the presence of bulk ura-
nium was unaffected and the recovery of iron was found
between 97 and 103%. An electropherogram obtained for
5ppm of Fe(III) with 80mg of U/mL is also shown in
Figure 3.
3.4. Validation of Method
3.4.1. Specificity of CZE Method
Owing to the minute differences in the charge densities
of many transition metal cations including iron, they may
interfere with iron. Therefore, the migrations of some
selective metal cations were examined. The iron peak
was totally interference free from the metal ions like
Cu(II), Pb(II), Ni(II), Zn(II), Mn(II), Co(II), Cd(II),
Cr(III), alkali, alkaline earth metals as they were not
detected up to a concentration of 300 ppm of each metal
ion. Hence, the CZE separation is highly specific for
Fe(III).
Table 1. Recoveries of spiked iron in uranium solutions of
various concentrations.
No Uranium
(ppm)
Fe spiked in U
solution (ppm)
Fe determined
by CE (ppm)a
% of
recovery
1 10 50 48.6 ± 2.4 97.2
2 1000 50 48.3 ± 2.7 96.6
3 10000 50 51.9 ± 2.9 103.8
4 50000 50 47.9 ± 3.1 95.8
5 75000 50 52.3 ± 2.6 104.6
6 90000 50 51.8 ± 2.8 103.6
amean value of three determinations (n = 3).
C
opyright © 2011 SciRes. AJAC

V. MISHRA ET AL.
51
3.4.2. Linearity
A linear relationship between peak area and concentra-
tion was obtained in the 1-50 ppm of Fe(III) and the cor-
relation coefficient obtained in this case was 0.9995.
3.4.3. Limit of Detection (LOD) and Limit of
Quantification (LOQ)
The detection limit for a signal-to-noise ratio of 3 for
Fe(III) was 0.1 ppm. The absolute LOD is 9 × 10-14 g;
considering the sampling volume as 1.5 nL. An e-gram
obtained for 0.08 ppm of iron, which is very close to the
detection limit was also shown in Figure 3. The limit of
quantification was approximately 0.6 ppm (basis of S/N
ratio of 6).
3.4.4. Reproducibility
The reproducibility was studied by making ten consecu-
tive runs of two standard solutions of iron, in which the
first solution has only 5 ppm of iron and the second has 5
ppm of iron in presence of 80 mg of U/mL of uranium.
The precisions (% RSD, n = 10) obtained for peak area
for the first and second solutions were 0.99 and 1.88 re-
spectively as shown in Figure 4(a). However, replicate
analysis (n = 20) of a 1 ppm iron standard solution under
the optimized conditions brought about a precision (%
RSD) of 3.5 as shown in Figure 4(b). Since the analyte
in the real sample is present in uranium matrix, the re-
producibility of migration time was also studied with the
same two 5 ppm iron solutions mentioned above. The
precisions (% RSD, n = 10) of migration time obtained
for the pure iron and with uranium were 0.97 and 1.02
respectively (Table 2).
3.4.5. Comparative Analysis with Certified Reference
Materials
To check the recovery of iron and to validate the devel-
oped procedure, two reference materials of U3O8 matrix
(namely ILCE-4 and ILCE-5), which were indigenously
prepared, characterized and certified by the Department
of Atomic Energy, India [49] were used for the trace
elements.
Accurately weighed quantities of the standards (in the
range of 0.3 - 0.4 g) were dissolved in 5 mL of high pu-
rity Conc. HNO3 and the solution was heated to near
dryness using a hot plate under controlled heating and
this procedure was repeated twice to ensure the complete
dissolution. Further 3.5 mL of 0.01 N HCl (pH 2) acid
solution was added at warm condition. The pH of the
solution was measured using a combined pH electrode
meant for small volume measurements and the pH was
adjusted using 0.01 N HCl solution, if required. Prior
to sample injection, the solution was transformed into a
standard volumetric flask and made up to 5 mL using
0.01 N HCl.
The dissolution was carried out with nitric acid be-
cause it oxidizes iron and uranium to their highest oxida-
tion states as Fe(III) and U(VI) respectively. Since both
elements exhibit their highest oxidation states, their
Figure 3. Electropherograms obtained for a standar d solution of Fe(III) and Uranium. (A) Fe(III) (5 µg/mL) + U(VI) (80,000
µg/mL) standard solution; (B) Fe(III) standard solution (1µg/mL); (C) Fe(III) (0.08 µg/mL) standard solution. BGE: 10 mM
HCl in 65 mM KCl (pH 2), Conditions: Capillary: 60 cm × 50 µm, Applied voltage: 15 kV. Detection: direct UV at 214 nm.

V. MISHRA ET AL.
52
(a)
(b)
Figure 4. (a) Reproducibility of Fe(III) peak area (n = 10)
obtained for 5 µg/mL of Fe with and without presence of
80,000 µg/mL of U; (b) for a 1 ppm Fe st andard sol ution (n =
22). Conditions are as same as in Figure 3.
Table 2. Reproducibility of migration time of Fe(III) (5
ppm) with and without presence of 80 mg of U/mL (n = 10).
Migration time (min)
Fe(III) alone Fe(III) with 80 mg of U (VI)/mL
3.088 (RSD, 0.97%) 3.133 (RSD, 1.09%)
combination cannot be a redox couple. Further the for-
mation of Fe(II) is not possible as it has least stability in
nitric and hydrochloric acid media and therefore, the
sample dissolution ensures all iron in its +3 state. The
entire sample preparation process typically takes around
30 minutes whereas in other methods like ICP-AES,
ICP-MS and ion chromatography etc. lasts for few hours
as these methods require the separation of matrix ele-
ment by following solvent extraction or ion exchange
procedures. Table 3 compares the results obtained for
the two reference materials and the values obtained by
this method are in good agreement with the certified
values.
3.5. Real Sample Analysis
The method was applied to nine uranium based fuel
samples from an advanced fuel fabrication facility. The
samples are either in metallic or oxide form of uranium.
Samples were dissolved by following the procedure as
described above for the standard reference materials
(ILCE-4 and ILCE-5). Depending on the concentration
of Fe obtained from the first run of the sample, further
dilution was carried out, to adjust the analyte concentra-
tions to the linearity range of method calibration. Table
4 lists the typical iron contents determined in the samples,
Table 3. Comparison between iron determination via the
developed method and certified method.
Concentration of Fe (ppmw)
No. Reference
Std. codeobtainedmean Certified
% of
deviation
163.2
160.3
167.1
161.6
1. ILCE-IV
165.5
163.5 ± 2.8 170 3.8
278.8
276.9
270.9
275.4
2. ILCE-V
274.0
275.2 ± 2.9 290 5.4
Table 4. Results obtained for the samples.
Sample code Fe (ppm)a
M1 882 ± 52
M2 1311 ± 70
M3 753 ± 37
M4 801 ± 36
M5 913 ± 46
U1 318 ± 19
U2 723 ± 33
U3 626 ± 29
U4 389 ± 18
amean value of three determinations (n = 3).

V. MISHRA ET AL.
53
Figure 5. A typical electropherogram obtained for a sample solution. Conditions are same as in Figure 3.
where the sample codes “M” and “U” refer to metallic
and oxides of uranium respectively.
3.6. Advantages of the CZE Method
1) Specific for iron and no interference from transition
and other common metal ions.
2) Possesses high tolerance for background matrix and
therefore, pre separation of matrix is not required.
3) Very simple sample preparation.
4) Simple and rapid separation. Typically it takes
around 10 minutes for a sample run.
5) Provides good precisions on migration time as well
as peak area of the analyte.
6) Recovery of the analyte is more than 95%.
7) Generation of minimum analytical waste as there
was not any additional wet handling other than the sam-
ple dissolution.
4. Conclusions
A rapid, reproducible and simple CZE method for the
determination of iron in uranium matrix has been dem-
onstrated. The method involved chloride complexation,
where cationic chloride complexes of iron as well as
uranium are selectively formed and this made the CZE
separation of iron from uranium feasible. The chloride
complex of iron allows direct detection at 214 nm with
satisfactory limit of detection. The capability of the
method for determining iron with high background of the
matrix element was proved by analyzing matrix match
reference materials. The method was applied success-
fully to the determination of iron in different uranium
based fuel materials and is adoptable for routine analyses.
5. Acknowledgements
Authors are thankful to Dr. V. Venugopal, Director, RC
& IG, BARC for his kind support. Sincere thanks are due
to Mrs. Vaibhavi V. Raut for carrying out the experi-
ments.
6. References
[1] M. V. Ramaniah, “Analytical Chemsitry of Fast Reactor
Fuels - A Review,” Pure and Applied Chemistry, Vol. 54,
No. 4, 1982, pp. 889-908. doi:10.1351/pac198254040889
[2] H. C. Bailly, D. Menessier and C. Prunier, “The Nuclear
Fuel of Pressurized Water Reactors and Fast neutron Re-
actors: Design and Behavior,” Lavoisier Publishing Inc.,
Paris, 1999.
[3] K. S. Rao, T. Balaji, T. P. Rao, Y. Babu and G. R. K.
Naidu, “Determination of Iron, Cobalt, Nickel, Manga-
nese, Zinc, Copper, Cadmium and Lead in Human Hair
Inductively Coupled Plasma-Atomic Emission Spec-
trometry,” Spectrochim Acta Part B, Vol. 57, No. 8, 2002,
pp. 1333-1338. doi:10.1016/S0584-8547(02)00045-9
[4] M. Gopalakrishnan, K. Radhakrishnan, P. S. Dhami, V. T.
Kulkarni, M. V. Joshi, A. B. Patwardhan, A. Ramanujam
and J. N. Mathur, “Determination of Trace Impurities in
Uranium, Thorium and Plutonium Matrices by Solvent
Extraction and Inductively Coupled Atomic Emission

54 V. MISHRA ET AL.
Spectrometry,” Talanta, Vol. 44, 1997, pp. 169-176.
doi:10.1016/S0039-9140(96)02027-9
[5] H. Vanhoe, C. Vandecasteele, J. Versieck and R. Dams,
“Determination of Iron, Cobalt, Copper, Zinc, Rubidium,
Molybdenum and Cesium in Human Serum by Induc-
tively Coupled Mass Spectrometry,” Analytical Chemis-
try, Vol. 61, No. 17, 1989, pp. 1851-1857.
doi:10.1021/ac00192a014
[6] Z. Marczenko and M. Balcerzak, “Separation, Precon-
centration and Spectrophotometry in Inorganic Analysis,”
Elsevier, 2000, pp. 226-237.
[7] A. Kelkar, A. Prakash, M. Afzal, J. P. Panakkal and H. S.
Kamath, “Determination of Alkali, Alkaline Earth and
Transition Metal Ions in UO2, ThO2 Powders and Sin-
tered (Th,U)O2 Pellets by Ion Chroamtography,” Journal
of Radioanalytical and Nuclear Chemistry, Vol. 284, No.
2, 2010, pp. 443-449. doi:10.1007/s10967-010-0491-y
[8] V. V. Raut, M. K. Das, S. Jeyakumar and K. L. Rama-
kumar, “HPIC Separation and Quantification of Ca and
Fe in Uranium Metal and Alloy Compounds,” Proceed-
ings SESTEC-2008, Delhi, 12-14 March 2008.
[9] R. M. Cassidy and S. Elchuk, “Application of High Per-
formance Liquid Chromatography to the Analysis of Nu-
clear Materials,” Journal of Liquid Chromatography &
Related Technologies, Vol. 4, No. 3, 1981, pp. 379-398.
doi:10.1080/01483918108059943
[10] S. Hjerten, “Free Zone Electrophoresis,” Chroma-
tographic Reviews, Vol. 9, 1967, pp. 122-219.
doi:10.1016/0009-5907(67)80003-6
[11] A. K. Malik, “Metal Analysis with Capillary Zone Elec-
trophoresis,” Methods in Molecular Biology, Vol. 328,
2008, pp. 21-42. doi:10.1007/978-1-59745-376-9_2
[12] J. S. Fritz, “Recent Developments in the Separation of
Inorganic and Small Organic Ions by Capillary Electro-
phoresis,” Journal of Chromatography A, Vol. 884, No.
1-2, 2000, pp. 261-275.
doi:10.1016/S0021-9673(00)00186-2
[13] A. R. Timerbaev, “Capillary Electrophoresis of Inorganic
Ions: An Update,” Electrophoresis, Vol. 25, 2004, pp.
4008-4031. doi:10.1002/elps.200406111
[14] A. R. Timerbaev, “Recent Trends in CE of Inorganic Ions:
From Individual to Multiple Elemental Species Analy-
sis,” Electrophoresis, Vol. 28, 2007, pp. 3420-3435.
doi:10.1002/elps.200600491
[15] A . R. Timerbaev, “Inorganic Species Analysis by CE -
An Overview for 2007-2008,” Electrophoresis, Vol. 31,
2010, pp. 192-204. doi:10.1002/elps.200900397
[16] F. Qu and J. Lina, “Separation of Transition Metals in
Nonaqueous Media with Capillary Electrophoresis,”
Journal of Chromatography A, Vol. 1068, 2005, pp.
169-174. doi:10.1016/j.chroma.2004.10.055
[17] F. Qu, J. Lin and Z. Chen, “Simultaneous Separation of
Nine Metal Ions and Ammonium with Nonaqueous Cap-
illary Electrophoresis,” Journal of Chromatography A,
Vol. 1022, 2004, pp. 217-221.
doi:10.1016/j.chroma.2003.09.039
[18] E. Naujalis and A. Padarauskas, “Development of Capil-
lary Electrophoresis for the Determination of Metal Ions
Using Mixed Partial and Complete Complexation Tech-
niques,” Journal of Chromatography A, Vol. 977, 2002,
pp. 135-142. doi:10.1016/S0021-9673(02)01350-X
[19] J . Gao, X. Sun, W. Yang, H. Fan, C. Li and X. Mao,
“Separation and Determination of Six Metal Cations by
Capillary Zone Electrophoresis,” Journal of the Chilean
Chemical Society, Vol. 53, No. 1, 2008, pp. 1431-1434.
doi:10.4067/S0717-97072008000100020
[20] B. H. Li and X. P. Yan, “Rapid Speciation of Iron by
On-Line Coupling of Short Column Capillary Electro-
phoresis and Inductively Coupled Plasma Mass Spec-
trometry with the Collision Cell Technique,” Journal of
Separation Science, Vol. 30, 2007, pp. 916-922.
doi:10.1002/jssc.200600405
[21] J. E. Sonke and V. J. M. Salters, “Capillary Electrophore-
sis-High Resolution Sector Field Inductively Coupled
Plasma Mass Spectrometry,” Journal of Chromatography
A, Vol. 1159, 2007, pp. 63-74.
doi:10.1016/j.chroma.2007.05.055
[22] R. Kautenburger and H. P. Beck, “Complexation Studies
with Lanthanides and Humic Acid Analysed by Ultrafil-
tration and Capillary Electrophoresis-Inductively Cou-
pled Plasma Mass Spectrometry,” Journal of Chroma-
tography A, Vol. 59, 2007, pp. 75-80.
doi:10.1016/j.chroma.2007.03.092
[23] E. X. Vrouwe, R. Luttge, W. Olthuis and A. van den Berg,
“Rapid Inorganic Ion Analysis Using Quantitative Mi-
crochip Capillary Electrophoresis,” Journal of Chroma-
tography A, Vol. 1102, 2006, pp. 287-293.
doi:10.1016/j.chroma.2005.10.064
[24] G. L. Klunder, J. E. Andrews, P. M. Grant, B. D. Andre-
sen and R. E. Russo, “Analysis of Fission Products Using
Capillary Electrophoresis with On-Line Radioactivity
Detection,” Analytical Chemistry, Vol. 69, 1997, pp.
2988-2993. doi:10.1021/ac970042e
[25] B. Kuczewski, C. M. Marquardt, A. Seibert, H. Geckeis,
J. V. Kratz and N. Trautmann, “Separtion of Plutonium
and Neptunium Species by Capillary Electrophoresis-
Inductively Coupled Plasma-Mass Spectrometry and Ap-
plication to Natural Groundwater Samples,” Analytical
Chemistry, Vol. 75, 2003, pp. 6769-6774.
doi:10.1021/ac0347213
[26] A. R. Timerbaev, “Metal Ions Analysis by Capillary
Electrophoresis: New Possibilities in Separation and De-
tection,” Journal of Capillary Electrophoresis, Vol. 1,
1995, pp. 14-23.
[27] F. Foret, S. Fanali, A. Nardi and P. Bocek, “Capillary
Zone Electrophoresis of Rare Earth Metals with Indirect
UV Absorbance Detection,” Electrophoresis, Vol. 11,
1990, pp. 780-783. doi:10.1002/elps.1150110919
[28] Y. Shi and J. S. Fritz, “Separation of Metal Ions by Cap-
illary Electrophoresis with A Complexing Electrolyte,”
Journal of Chromatography A, Vol. 640, 1993, pp.
473-479. doi:10.1016/0021-9673(93)80217-V
[29] A. R. Timerbaev, O. P. Semenova and O. M. Petrukhin,

V. MISHRA ET AL.
55
“Migration Behaviour of Metal Complexes in Capillary
Zone Electrophoresis: Interpretation in Terms of Quanti-
tative Structure-Mobility Relationships,” Journal of
Chromatography A, Vol. 943, 2002, pp. 263-274.
doi:10.1016/S0021-9673(01)01462-5
[30] S. Pozdniakova, A. Padarauskas and G. Schwedt, “Si-
multaneous Determination of Iron(II) and Iron(III) in
Water by Capillary Electrophoresis,” Analytica Chimica
Acta, Vol. 351, 1997, pp. 41-48.
doi:10.1016/S0003-2670(97)00331-0
[31] B. Y. Deng, C. J. Zeng and W. T. Chan, “Speciation of
Iron Using Capillary Electrophoresis Inductively Coupled
Plasma Atomic Emission Spectrometry,” Spectroscopy
and Spectral Analysis, Vol. 25, 2005, pp. 1868-1871.
[32] P. Blatný, F. Kvasnička and E. Kenndler, “Trace Deter-
mination of Iron in Water at the µg/l Level by On-line
Coupling of Capillary Zone Electrophoresis with UV
Detection of the EDTA-Fe(III) Complex,” Journal of
Chromatography A, Vol. 757, 1997, pp. 297-302.
doi:10.1016/S0021-9673(96)00668-1
[33] B. Baraj, M. Martinez, A. Sastre and M. Aguilar, “Si-
multaneous Determination of Cr(III), Fe(III), Cu(II) and
Pb(II) as UV-absorbing EDTA Complexes by Capillary
Zone Electrophoresis,” Journal of Chromatography A,
Vol. 695, 1995, pp. 103-111.
doi:10.1016/0021-9673(94)01078-S
[34] G. Owens, V. K. Ferguson, M. J. Mclaughlin, I. Singleton,
R. J. Reid and F. A. Smith, “Determination of NTA and
EDTA and Speciation of Their Metal Complexes in
Aqueous Solution by Capillary Electrophoresis,” Envi-
ronmental Science & Technology, Vol. 34, 2000, pp.
885-891. doi:10.1021/es990309m
[35] S. Schäffer, P. Gared, C. Dezael and D. Richard, “Direct
Determination of Iron(II), Iron (III) and Total Iron as
UV-absorbing Complexes by Capillary Electrophoresis,”
Journal of Chromatography A, Vol. 740, 1996, pp.
151-157. doi:10.1016/0021-9673(96)00121-5
[36] W. Buchberger and S. Mülleder, “Determination of Che-
lating Agents and Metal Chelates by Capillary Zone Elec-
trophoresis,” Mikrochimica Acta, Vol. 119, 1995, pp.
103-111. doi:10.1007/BF01244859
[37] P. Kuban, P. Kuban and V. Kuban, “Simultaneous Capil-
lary Electrophoretic Separation of Small Anions and
Cations after Complexation with Ehylenediamine-
tetraacetic Acid,” Journal of Chromatography A, Vol.
836, 1999, pp. 75-80.
doi:10.1016/S0021-9673(98)00980-7
[38] M. Xavier, P. R. Nair, K. V. Lohithakshan, S. G. Marathe
and H. C. Jain, “Determiantion 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, 1991, pp. 251-256.
doi:10.1007/BF02060358
[39] I. Ali, V. K. Gupta and H. Y. Aboul-Enein, “Metal Ion
Speciation and Capillary Electrophoresis: Application in
the New Millennium,” Electrophoresis, Vol. 26, 2005, pp.
3988-4002. doi:10.1002/elps.200500216
[40] B. Baraj, A. Sastre, A. Merkoci and M. Martinez, “De-
termination of Chloride Complex of Au(III) by Capillary
Zone Electrophoresis with Direct UV Detection,” Journal
of Chromatography A, Vol. 718, 1995, pp. 227-232.
doi:10.1016/0021-9673(95)00655-9
[41] B. Baraj, A. Sastre, M. Martinez and K. Spahiu, “Simul-
taneous Determination of Chloride Complexes of Pt(IV)
and Pd(II) by Capillary Zone Electrophoresis with Direct
UV Absorbance Detection,” Analytica Chimica Acta, Vol.
319, 1996, pp. 191-197.
doi:10.1016/0003-2670(95)00472-6
[42] C. Hennig, K. Servaes, P. Nockemann, K. V. Hecke, L. V.
Meervelt, J. Wouters, L. Fluyt, C. Görller-Walrand and R.
V. Deun, “Species Distribution and Coordination of
Uranyl Chloro Complexes in Acetonitrile,” Inorganic
Chemistry, Vol. 47, 2008, pp. 2987-2993.
doi:10.1021/ic7014435
[43] C. Hennig, J. Tutschku, A. Rossberg, G. Bernhard and A.
C. Scheinost, “Comparative EXAFS Investigation of Ura-
nium(VI) and –(IV) Aquo Chloro Complexes in Solution
Using a Newly Developed Spectroelectrochemical Cell,”
Inorganic Chemistry, Vol. 44, 2005, pp. 6655-6661.
doi:10.1021/ic048422n
[44] N. J. Welham, K. A. Malatt and S. Vukcevic, “The Effect
of Solution Speciation on Iron-Sulphur-Arsenic-Chloride
Systems at 298K,” Hydrometallurgy, Vol. 57, 2000, pp.
209-223. doi:10.1016/S0304-386X(00)00121-3
[45] W. Liu, B. Etschmann, J. Brugger, L. Spiccia, G. Foran
and B. McInnes, “UV-Vis Spectrophotometric and XAFS
Studies of Ferric Chloride Complexes in Hyper-Saline
LiCl Solutions at 25-90˚C,” Chemical Geology, Vol. 231,
2006, pp. 326-349. doi:10.1016/j.chemgeo.2006.02.005
[46] W. Walkowlak, D. Bhattacharyya and R. B. Grieves,
“Selective Foam Fractionation of Chloride Complexes of
Zinc(II), Cadmium(II), Mercury(II) and Gold(III),” Ana-
lytical Chemistry, Vol. 48, 1976, pp. 975-979.
doi:10.1021/ac60371a036
[47] J. Bjerrum, “Determiantion of Small Stability Constants,
A Spectrophotmetric Study of Copper(II) Chloride Com-
plexes in Hydrochloric Acid,” Acta Chemica Scandi-
navica - Series A, Vol. 41, 1987, pp. 328-334.
doi:10.3891/acta.chem.scand.41a-0328
[48] H. Engelhardt, W. Beck and T. Schmitt, “Capillary Elec-
trophoresis,” Friedrich Vieweg & Sohn Verlagsgesell-
schaft mbH, Braunschweig/Wiesbaden, 1996.
[49] DAE-Interlaboratory Comparison Experiment (ILCE) for
Trace Metal Assay of Uranium, Phase-II Experimets:
Development of Certified U3O8 Reference Materials,
Document, BARC, India, February 2002.