Talanta 78 (2009) 290–294

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

Comparison of ICP-MS and SIMS techniques for determining uranium isotope

ratios in individual particles
F. Esaka ∗ , M. Magara, C.G. Lee, S. Sakurai, S. Usuda, N. Shinohara
Research Group for Analytical Science, Japan Atomic Energy Agency, 2-4 Shirakata-Shirane, Tokai, Naka, Ibaraki 319-1195, Japan

a r t i c l e i n f o a b s t r a c t

Article history: The determination of uranium isotope ratios in individual particles is of great importance for nuclear
Received 15 October 2008 safeguards. In the present study, an analytical technique by inductively coupled plasma mass spectrom-
Received in revised form etry (ICP-MS) with a desolvation sample introduction system was applied to isotope ratio analysis of
10 November 2008
individual uranium particles. In ICP-MS analysis of individual uranium particles with diameters ranging
Accepted 10 November 2008
from 0.6 to 4.2 ␮m in a standard reference material (NBL CRM U050), the use of the desolvation system for
Available online 21 November 2008
sample introduction improved the precision of 234 U/238 U and 236 U/238 U isotope ratios. The performance
of ICP-MS with desolvation was compared with that of a conventionally used method, i.e., secondary ion
Keywords:
Uranium
mass spectrometry (SIMS). The analysis of test swipe samples taken at nuclear facilities implied that the
Isotope ratio performance of ICP-MS with desolvation was superior to that of SIMS in a viewpoint of accuracy, because
Particles the problems of agglomeration of uranium particles and molecular ion interferences by other elements
ICP-MS could be avoided. These results indicated that ICP-MS with desolvation has an enough ability to become
SIMS an effective tool for nuclear safeguards.
© 2008 Elsevier B.V. All rights reserved.

1. Introduction Recently, inductively coupled plasma mass spectrometry (ICP-

Isotope ratio analysis of individual particles is increasingly
important in the field of nuclear safeguards. Here, individual ura-
nium particles in environmental samples taken at nuclear facilities
are measured to reveal the presence of undeclared nuclear activities
[1]. For this purpose, two kinds of analytical techniques are uti-
lized, i.e., a fission track method combined with thermal ionization
mass spectrometry (FT-TIMS) [2–4] and secondary ion mass spec-
trometry (SIMS) [5–7]. The FT-TIMS technique allows us to identify
uranium particles easily and to measure isotope ratios precisely.
However, it needs a nuclear reactor to irradiate samples with ther-
mal neutrons for obtaining fission tracks and it is complicated and
time-consuming to prepare samples. The SIMS technique is efficient
in analyzing samples when uranium particles are large enough both
in numbers and size. There is a difficulty in analyzing 234 U/238 U
and 236 U/238 U isotope ratios accurately because of isobaric inter-
ferences caused by molecular ions of other elements. In a previous
paper, we developed a combination method of particle manipu-
lation and SIMS (manipulation-SIMS), removing the interferences
from other elements in other particles [7]. A remaining problem is
the interferences from other elements in the same particle.
MS) has been increasingly applied for isotope ratio analysis of
long-lived radionuclides [8]. The high sensitivity suggests that it
will become an effective analytical tool for individual particles.
Boulyga and Prohaska [9] applied laser ablation – ICP-MS for deter-
mination of uranium isotope ratio in individual particles with
sub-millimeter in diameter. Further studies would be, however,
needed to measure smaller particles precisely. Degueldre et al.
[10] performed ICP-MS analysis for single uranium particles intro-
duced as diluted suspensions. However, it is difficult to apply these
approaches to isotope ratio analysis because only transient signals
are obtained.
In a previous work [11], we developed an experimental tech-
nique for isotope ratio analysis of individual particles, including
particle manipulation and chemical dissolution of individual par-
ticle for subsequent isotope ratio analysis with ICP-MS. This
technique also has a possibility to remove isobaric interfer-
ences from other elements existing in the same particle by a
combination with chemical separation. Although the developed
technique enabled us to acquire stable signals needed for isotope
ratio measurement, it was difficult to obtain precise results for
234 U/238 U and 236 U/238 U isotope ratios due to the lack of sensitiv-

∗ Corresponding author. Tel.: +81 29 282 6165; fax: +81 29 282 6950.
E-mail address: esaka.fumitaka@jaea.go.jp (F. Esaka).
ity.
In the present study, a desolvation sample introduction system
was used to improve sensitivity in ICP-MS analysis for individual
particles. Moreover, the performance for isotope ratio analysis of
individual particles was evaluated by using test swipe samples

0039-9140/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.talanta.2008.11.011
 F. Esaka et al. / Talanta 78 (2009) 290–294 291

Fig. 1. (a) The 238 U ion signal intensities measured for individual NBL CRM U050 particles by ICP-MS with () and without () desolvation. (b) Example of the SEM images
of measured uranium particles.

containing uranium particles and compared with that of SIMS and nium particle to be analyzed was manipulated with the glass
manipulation-SIMS methods. needle attached to the manipulator and transferred onto the blank
planchet. For test swipe samples, uranium particles without manip-
2. Experimental procedure ulation were also prepared for SIMS analysis.

2.1. Samples 2.4. Instrumentation

Uranium particles in a certified reference material (grains of An Apex-Q desolvation system (Elemental Scientific Inc., USA)
U3 O8 powder, CRM U050, New Brunswick Laboratory (NBL), USA) was used for sample introduction to ICP-MS. This system consists
were used in this study. The uranium particles were smeared onto a of a heated cyclonic spray chamber, a Peltier cooled condenser
cotton cloth (TexWipe® TX304, used for safeguards environmental and an ACM Naflon fluoropolymer membrane desolvation mod-
sampling by IAEA inspectors). In addition, two kinds of test swipe ule, allowing high efficient sample introduction. Uranium isotope
samples taken at nuclear facilities were used to compare the per- ratios were measured using a high resolution ICP-MS model ELE-
formance of ICP-MS with SIMS. MENT 1 (Thermo Fisher Scientific, USA), which has an ICP ion source
with Guard electrode and a double-focusing mass spectrometer.
2.2. Sample preparation for ICP-MS The instrument was operated in electric scanning (E-scan) modes
with a mass resolution (M/M) of 300. A PFA micro-flow nebulizer
Particles on the cotton cloth were recovered onto a Si disk (Nihon (self-aspirating mode) was used in combination with a Scott-type
Exceed Co. Ltd., Japan) with a diameter of 25 mm and a thickness of double pass spray chamber cooled at 4 ◦ C. The instrument with-
3 mm by an impaction method [12]. Si chips (5 mm × 5 mm, Semitec out desolvation was tuned daily for optimal 238 U sensitivity (2300
Co. Ltd., Japan) were cleaned, in sequence, with 40%, 2% HNO3 cps/ppt). Four mono-atomic ions (234 U+ , 235 U+ , 236 U+ and 238 U+ )
(high purity, Tama Chemicals Co. Ltd.) and deionized Milli-Q water were measured in a peak-jumping mode. The signal intensity of
238 U16 O+ was also measured for monitoring the stability of the
(18 M). The Si disk and the Si chips were introduced into a cham-
ber of a scanning electron microscope (SEM, JSM-6700F, Jeol Co. instrument. The formation rate of 238 U16 O+ to 238 U+ was between 2
Ltd., Japan) simultaneously. Individual uranium particles on the Si and 3%. The operating condition for the ICP-MS is listed in Table 1.
disk were observed with SEM and identified with energy disper- The interference of 235 U1 H+ on 236 U+ was corrected with prelimi-
sive X-ray (EDX) analysis. Single uranium particle was manipulated
with a glass needle attached to a manipulator and transferred onto Table 1
the center of each Si chip. Teflon bottles with a volume of 8 mL were Analytical condition in ICP-MS.
cleaned with 4% HNO3 and deionized Milli-Q water. The 8 mL of 4%
Parameter Setting
HNO3 was, then, added to the bottle and the count rate at a mass
238 was measured with ICP-MS to check the cleanliness of the bot- ICP-MS
RF power 1148 W
tle. Each Si chip containing a single uranium particle was dropped Cooling gas flow rate 16.0 L/min
into the bottle and 250 ␮L of 40% HNO3 was added. The particle Auxiliary gas flow rate 0.85 L/min
was soaked overnight and dissolved at 55 ◦ C for 4 h under ultrason- Sample gas flow rate 0.79–0.86 L/min
eration. The solution was finally adjusted to 4% HNO3 by adding Solution uptake rate 0.18 mL/min
Sampling time per isotope 50 ms
2.75 mL of deionized Milli-Q water.
Scan per replicate 400
Number of replicates 5
2.3. Sample preparation for SIMS Resolution (M/M) 300

Apex-Q desolvation system

Particles on the cotton cloth were collected onto a glassy carbon Spray chamber temperature 100 ◦ C
disk (Ø25 mm, Hitachi Chemical Co. Ltd., Japan) by the impaction Condenser temperature 2 ◦C
method. Both the planchet containing particles and a blank disk Nitrogen flow 1.5 mL/min
Additional gas (Ar) 0.08–0.15 L/min
were introduced into the main chamber of the SEM. Each ura-
292
Fig. 2. The precision of (a) 234 U/238 U, (b) 235 U/238 U and (c) 236 U/238 U isotope ratios obtained for individual NBL CRM U050 particles by () manipulation-SIMS, () ICP-MS
and () ICP-MS with desolvation.
narily measured 238 U1 H/238 U ratio, 5.0 × 10−5 . The mass bias factor
assuming the following equation was determined using a 4 pg/mL
NBL CRM U015 standard solution:
Rt = Rm (1 + C)M
where Rt is the certified isotope ratio, Rm is the measured isotope
ratio, M is a difference in mass and C is a mass discrimination
factor.
A secondary ion mass spectrometer (IMS-6f, Cameca, France)
was used for measuring uranium isotope ratios in individual parti-
cles. In SIMS analysis, an O2 + focused beam of 15 keV with a current
of 1.0–20 nA was rastered over a 50 ␮m × 50 ␮m area. Positive sec-
ondary ions emitted from a particle were accelerated through 5 keV.
Transfer optics of 150 ␮m and a field aperture of 1800 ␮m were
used. An energy slit was adjusted for a band pass of 50 eV. The
mass resolution was set at 300. Secondary ions at masses 234,
235, 236, 238 and 239 were counted in an automatic peak-jumping
mode. An electron multiplier was operated in an ion counting mode
with a dead time of 39 ns. Mass calibration was performed each
time before starting measurements of particles. Mass bias correc-
tions were conducted by comparing the measured isotope ratios
for a certified uranium reference material (NBL CRM U350) with
the reference values. The ratio of certified 235 U/238 U to measured
235 U/238 U was around 0.991. The hydride formation rate of ura-
nium (238 UH/238 U) was determined (approximately 4 × 10−4 ) for
each particle and used to subtract the contribution of 235 UH from
the signal of a mass 236.
3. Results and discussion
3.1. Effect of Apex-Q desolvation system in ICP-MS
The Apex-Q desolvation system was used to improve sensitivity
for analysis of individual uranium particles. Prior to the analysis of
uranium particles, procedural blank solution was prepared by per-
forming the same sample preparation procedure without particles.
Count rates on masses 234, 235, 236 and 238 were measured for
procedural blank solution by ICP-MS with and without desolvation
and are given in Table 2. No obvious differences in the count rates
on masses 234 and 236 were found between the results obtained
with and without desolvation. In contrast, the use of the desolva-
tion system caused the increases of the count rates on masses 235
and 238, which would be due to the memory effect of the system.
The 238 U ion signal intensities were measured for individual
U050 particles with the diameters ranging from 0.6 to 4.2 ␮m by
F. Esaka et al. / Talanta 78 (2009) 290–294
ICP-MS with and without desolvation. These results were plotted
against the 238 U amount calculated using the diameter as an aver-
age value between the lengths of major and minor axes of each
particle observed with SEM in Fig. 1(a). Fig. 1(b) shows examples
(1) of the SEM images of the particles. As shown in Fig. 1(a), the 238 U
count rate obtained with desolvation was several times higher than
that without desolvation. The increases of count rates were also
observed on masses 234, 235 and 236. In the result measured with
desolvation, approximate chemical recovery rate was about 90%,
which was calculated using the uranium amount estimated from
particle size and the signal intensity obtained with ICP-MS. This
recovery rate would depend on the chemical composition of par-
ticles to be measured. Therefore, chemical dissolution procedure
should be optimized to perform sensitive measurement.
In isotope ratio analysis, precision is an important factor to
evaluate performance of measurements. In order to compare the
performance between ICP-MS and SIMS, relative standard devia-
tion (R.S.D.) values of uranium isotope ratios obtained by analyzing
individual U050 particles were plotted against the amounts of ura-
nium isotopes and are shown in Fig. 2. The R.S.D. values of 234 U/238 U
and 236 U/238 U isotope ratios obtained by ICP-MS with desolvation
were smaller than those without desolvation. These implied that
the use of the desolvation system improved the precision of these
isotope ratios. In contrast, no obvious improvement was found in
the analysis of 235 U/238 U isotope ratio. This is related to the increase
of 235 U intensity of procedural blank by using the desolvation sys-
tem. Thorough cleaning of the system to reduce the blank value is
necessary to perform more precise analysis.
For all isotope ratios, manipulation-SIMS gave better R.S.D. val-
ues than ICP-MS analysis. However, in the analysis of relatively large
uranium particles, the R.S.D. values obtained by ICP-MS with des-
olvation were nearly equal to those by manipulation-SIMS. This
means that ICP-MS analysis with desolvation has enough perfor-
mance for individual particle analysis. It should be noted that all
Table 2
The procedural blank values on masses 234, 235, 236 and 238 in ICP-MS with and
without desolvation.
Mass Counts/s
Procedural blank Procedural blank with desolvation
234 0.5 ± 0.3a 0.7 ± 0.3
235 0.9 ± 0.5 2.3 ± 1.7
236 0.2 ± 0.2 0.0 ± 0.1
238 12.6 ± 6.6 115.8 ± 18.6
a
One standard deviation.
 F. Esaka et al. / Talanta 78 (2009) 290–294
Fig. 3. Total-reflection X-ray fluorescence spectra of test swipe samples: (a) 1 and
(b) 2 measured at an incident angle of 0.05 ◦ for 1000 s.
R.S.D. values of 235 U/238 U isotope ratios in Fig. 2 were within 5%,
which is an acceptable precision in the analysis of individual ura-
nium particles for nuclear safeguards [13].
3.2. Analysis of test swipe samples
Swipe samples taken at nuclear facilities contain various kinds
of particles having different size, morphology and elemental com-
position. Typically, swipe samples contain a large number of
non-uranium particles and a small number of uranium particles.
In this study, two kinds of test swipe samples were analyzed to
compare performance of SIMS and ICP-MS with desolvation. Fig. 3
Fig. 4. The 234 U/238 U or 236 U/238 U isotope ratio against the 235 U/238 U isotope ratio measured for test swipe sample 1 by: (a) SIMS, (b) manipulation-SIMS and (c) ICP-MS with
desolvation. The error bars represent one standard deviation.
293
represents the total-reflection X-ray fluorescence (TXRF) spectra of
these samples. Here, the particles on Si disks were irradiated with
molybdenum X-rays at an incident angle of 0.05◦ . The TXRF analy-
sis implied that sample 1 contained iron, zinc and a relatively larger
amount of uranium. Sample 2 contained iron, calcium, lead and a
small amount of uranium. The iron and zinc are used for equipment
and the lead is utilized as a shielded material for neutron in nuclear
facilities. Thus, swipe samples taken at nuclear facilities normally
contain these elements.
Fig. 4 shows the 234 U/238 U and 236 U/238 U isotope ratios mea-
sured for sample 1, plotted against 235 U/238 U isotope ratios. For
comparison, the result obtained by SIMS without manipulation,
used as a standard method for nuclear safeguards, is also shown
in Fig. 4(a). The result in Fig. 4(a) indicated that all particles
had almost the same isotope ratios. The average values were
4.22 × 10−4 , 4.00 × 10−2 and 6.50 × 10−5 for 234 U/238 U, 235 U/238 U
and 236 U/238 U isotope ratios, respectively. In contrast, the result
obtained by manipulation-SIMS (Fig. 4(b)) and ICP-MS with desol-
vation (Fig. 4(c)) indicated the presence of uranium particles with
different isotopic ratios. For example, the range of 235 U/238 U isotope
ratio measured by ICP-MS was from 2.52 × 10−2 to 5.14 × 10−2 . In
addition, the 234 U/238 U isotope ratio had a good correlation with
the 235 U/238 U isotope ratio. The correlation between 234 U/238 U and
235 U/238 U isotope ratios is a signature in enrichment plants [1]. The
discrepancy of the results between Fig. 4(a) and (c) is presumably
due to agglomeration of uranium particles having different isotope
ratios. In fact, the averaged 235 U/238 U isotope ratio (4.06 × 10−2 ) in
ICP-MS analysis was well consistent with that (4.00 × 10−2 ) in SIMS
without manipulation.
When the sample has a large number of uranium particles,
agglomeration is a severe problem in SIMS analysis. This can be
avoided by using single particle manipulation before SIMS analy-
sis. Since ICP-MS analytical procedure also has a process of single
particle manipulation, the problem of agglomeration of uranium
particles could be avoided.
The results for sample 2 are shown in Fig. 5. In the result of
SIMS without manipulation (Fig. 5(a)), the variations of 234 U/238 U
and 236 U/238 U isotope ratios were larger, compared to those
in manipulation-SIMS (Fig. 5(b)) and ICP-MS with desolvation
(Fig. 5(c)). This discrepancy can be explained by an effect of molec-
ular ion interferences in SIMS analysis. This sample has a small
amount of uranium and larger amounts of other elements, as shown
in Fig. 3. In particular, lead is known as a typical element giving
molecular ion interferences on uranium mass region [7]. Molecu-
lar ions such as 206 Pb12 C16 O, 207 Pb12 C16 O and 208 Pb12 C16 O can give
294
Fig. 5. The 234 U/238 U or 236 U/238 U isotope ratio against the 235 U/238 U isotope ratio measured for test swipe sample 2 by: (a) SIMS, (b) manipulation-SIMS and (c) ICP-MS with
desolvation. The error bars represent one standard deviation.
interferences on masses 234, 235 and 236, respectively. Uranium
minor isotopes (234 U and 236 U) are easily affected by these inter-
ferences because the count rates of these isotopes obtained in SIMS
analysis are considerably low. The problem of interferences could
also be removed by using single particle manipulation in SIMS and
ICP-MS analyses as shown in Fig. 5(b) and (c). It should be noted
that the correlation between the 234 U/238 U and 235 U/238 U isotope
ratios was clearly observed in Fig. 5(b) and (c).
4. Conclusions
A desolvation system was used for determination of individual
uranium particles in ICP-MS analysis. The improvement of pre-
cision in the analysis of 234 U/238 U and 236 U/238 U isotope ratios
was achieved, though the increase of procedural blank values on
masses 235 and 238 prevented the improvement of precision of
235 U/238 U isotope ratio. The ability of ICP-MS with desolvation
for determination of uranium isotope ratios in individual parti-
cles was compared with that of SIMS by using test swipe samples.
As well as manipulation-SIMS, the use of ICP-MS allowed us to
perform precise isotope ratio analysis without problems of agglom-
eration of uranium particles and molecular ion interferences. In
conclusion, ICP-MS with desolvation has almost the same ability
as manipulation-SIMS, becoming an effective tool for individual
particle analysis for nuclear safeguards.
F. Esaka et al. / Talanta 78 (2009) 290–294
Acknowledgements
This work was supported by the Ministry of Education, Culture,
Sports, Science and Technology of Japan.
References
[1] D.L. Donohue, J. Alloys Compd. 271–273 (1998) 11.
[2] S. Baude, R. Chiappini, Proceedings of Symposium on International Safeguards,
Vienna, Austria, 2001, IAEA-SM-367/10/05.
[3] K.T. Esaka, F. Esaka, J. Inagawa, K. Iguchi, C.G. Lee, S. Sakurai, K. Watanabe, S.
Usuda, Jpn. J. Appl. Phys. 43 (2004) L915.
[4] C.G. Lee, K. Iguchi, F. Esaka, M. Magara, S. Sakurai, S. Usuda, Jpn. J. Appl. Phys. 45
(2006) L1121.
[5] D.S. Simons, G. Gillen, C.J. Zeissler, R.H. Fleming, P.J. McNitt, Secondary Ion Mass
Spectrometry XI, John Wiley & Sons, New York, 1998, pp. 59–62.
[6] G. Tamborini, M. Betti, V. Forcina, T. Hiernaut, B. Giovannone, L. Koch, Spec-
trochim. Acta B 53 (1998) 1289.
[7] F. Esaka, K.T. Esaka, C.G. Lee, M. Magara, S. Sakurai, S. Usuda, K. Watanabe,
Talanta 71 (2007) 1011.
[8] D. Lariviere, V.F. Taylor, R.D. Evans, R.J. Cornett, Spectrochim. Acta B 61 (2006)
877.
[9] S.F. Boulyga, T. Prohaska, Anal. Bioanal. Chem. 390 (2008) 531.
[10] C. Degueldre, P.-Y. Favarger, R. Rossé, S. Wold, Talanta 68 (2006) 623.
[11] X.Z. Zhang, F. Esaka, K.T. Esaka, M. Magara, S. Sakurai, S. Usuda, K. Watanabe,
Spectrochim. Acta B 62 (2007) 1130.
[12] F. Esaka, K. Watanabe, H. Fukuyama, T. Onodera, K.T. Esaka, M. Magara, S. Sakurai,
S. Usuda, J. Nucl. Sci. Technol. 41 (2004) 1027.
[13] V.M. Piksaikin, G.M. Pshakin, V.A. Roshchenko, Sci. Global Secur. 14
(2006) 49.