Investigation on the use of UTEVA as a stationary phase for chromatographic separation of actinides on-line to inductively coupled plasma mass spectrometry

Abstract

A method based on the use of the UTEVA-Spec. Eichrom resin as the stationary chromatographic phase is described. The column was coupled on-line to an ICP-MS detector for the determination of actinides. The method is compared with one based on the use of commercially available cation-exchange stationary phases. The analytical procedure was validated by the use of certified reference materials as well as by other independent analytical techniques. It was demonstrated that the UTEVA-Spec. column can be used at the same time as a pre-concentration column, as well as a chromatographic column, lowering the detection limit of analysis down to a few pg g⁻¹.

---

Introduction

Inductively coupled plasma mass spectrometry (ICP-MS) has become one of the most successful methods in atomic spectroscopy, owing to its high detection power and true multi-elemental capability. Nevertheless, the analytical figures of merit are limited by spectroscopic and non-spectroscopic interference. Spectroscopic interferences are caused by atomic or molecular ions having the same nominal mass as the analyte isotope of interest. Recently, the use of ICP-MS instruments based on double-focusing sector field mass spectrometers (ICP-SFMS) has been proposed in order to overcome such interferences.¹

Quadrupole and sector ICP-MS are very intensively exploited for the determination of longer lived radioisotopes.^2–4

In the case of the spectroscopic interferences due to isobars of different elements, for example, those for transuranium elements or fission products (Table 1), even the resolution of sector field ICP-MS is insufficient. In these cases a chemical separation of the elements is necessary. In our laboratory, the use of a chromatographic separations system coupled on-line to an ICP-MS quadrupole instrument has been exploited for the determination of the complete inventory of fission products and actinides in solutions of spent nuclear fuel as well as in environmental samples.^5–7 The chromatographic phases used consisted of commercially available columns from Dionex and the separation mechanisms were based on cation exchange. Automated radiochemical separation based on the use of Eichrom resins has been recently revised.⁸ The use of UTEVA-Spec. has also been investigated for the separation on-batch of U and Pu from fission product and minor actinides before their measurements by thermal ionization mass spectrometry (TIMS).⁹ In this paper we investigated the use of UTEVA-Spec. as a stationary chromatographic phase coupled on-line with ICP-MS. A method based on extraction chromatography for the simultaneous determinations for Th, Np, U and Pu isotopes was developed. Interferences from Am and Cm could be avoided by exploiting their elution in the dead volume of the column.

Table 1 List of the more common isobaric interferences for transuranium elements and fission products

Mass	Elements
134	Cs–Ba
135	Cs–Ba
137	Cs–Ba
144	Ce–Nd
147	Pm–Sm
151	Sm–Eu
155	Eu–Gd
232	Th–U
238	U–Pu
241	Am–Pu
242	Pu–Cm
243	Am–Cm

---

The method has been compared with that based on the use of the CS10 (Dionex) column for the separation of Np, U and Pu.⁶

Experimental

Instrumentation

An Elan 5000 (Perkin-Elmer SCIEX, Thornhill, Ontario, Canada) ICP-MS instrument was modified for installation in a glove-box as described elsewhere.^10,11 The eluent from the Dionex 4500I (Sunnyvale, CA USA) gradient high-performance liquid chromatographic (HPLC) pump passed inside the box through a Dionex CG2 guard column, which was used only as a security column (i.e., to prevent any liquid flow from the box to the outside). After the inlet, a pneumatically activated four-way high-pressure valve was used for injection prior to separation on the column, which was finally connected directly to the cross-flow nebulizer. Helium gas was used for degassing the eluents and to activate the injection valve from the HPLC pump automatically. The waste from the spray chamber was pumped out with a two-way peristaltic pump (Gilson, Villiers-le-Bel, France) and stored inside the glove-box.

Reagents and materials

Natural element standards (²³²Th, ²³⁸U) were obtained from SPEX (Grasbrunn, Germany) as 1000 µg ml⁻¹ stock standard solutions and diluted as necessary with 3 M nitric acid. Enriched isotopes ²⁴³Am, ²⁴⁸Cm and ²³⁰Th from Amersham (Braunschweig, Germany), a certified enriched ²³³U from IRMM (Geel, Belgium), and a ²³⁷Np standard from Los Alamos National Laboratory (Los Alamos, NM, USA) were used. Certified enriched ²⁴⁴Pu and certified ²³⁸Pu and ²³⁹Pu standards were obtained from the New Brunswick Laboratory (Argonne, IL, USA) and National Institute of Standards and Technology (Gaithersburg, MD, USA), respectively. UTEVA-Spec. resin (100–150 µm) from Eichrom Inc. was used.

Nitric acid and hydrochloric acid were Suprapur grade from Merck (Darmstadt, Germany). Oxalic acid, ascorbic acid, hydrogen peroxide, sodium nitrite, sodium sulfite, hydroxyl-amine and Mohr's salt [(NH₄)₂Fe(SO₄)₂·6H₂O] were obtained from Merck (Darmstadt, Germany). Water purified in a Milli-Q system (Millipore, Eschborn, Germany) was used throughout.

All standard solutions, spikes and sample were prepared by dilution by mass in polyethylene bottles. Radioactive samples and standards were treated in a glove-box.

A sediment reference sample, IAEA 135, certified for ²³²Th was used. The sample was prepared in a clean laboratory (class 10-100) using a microwave digestor (Gesa 301, Prolabo, France).

Column preparation

The UTEVA-Spec., TRU and TEVA resins from Eichrom were used to pack the chromatographic analytical columns to be connected on-line to an ICP-MS detector. For this purpose columns of length of 5 cm and 4 mm internal diameter were used.

In particular the UTEVA-Spec. resin, exploited in this investigation, consisted of particles with an external diameter of 100–150 µm.

Quantification procedures

The samples, spiked sample and spikes were injected into the chromatographic system. Integration of chromatographic peaks was performed using the software package Origin™ (Microcal Software Inc., Northampton, MA, USA). Quantitative results based on peak areas were obtained by isotope dilution analysis as well as the standard additions method.

Procedures for separation and pre-concentration

For the normal sample injection in the chromatographic column, a four-way valve as shown in Fig. 1A was used. In the standard configuration, when the valve is in the off position, the syringe in position (3) that fills up the loop (4–8), is used for sample injection. The eluent enters in position (5), flows through the loop (8–4) that contains the sample and flows to the chromatographic column through the exit (1), and then to the ICP-MS instrument.

Fig. 1 Four-way valve configuration for chromatographic separation on UTEVA-Spec: A, normal chromatographic elution and B, on-line preconcentration and successive chromatographic separation.

To carry out the on-line preconcentration of the actinides, the system was modified as follows (Fig. 1B): when the valve is closed (off), the sample solution is loaded onto the column (4–8) by a two-way peristaltic pump. In this way the actinides are retained on the column, while the matrix flows to waste (7). The eluent enters in (5), flows through the connection tube and exits at (1), and then goes directly to the ICP-MS instrument. When the preconcentration operation is finished, the valve is opened and the eluent flows through the chromatographic column (8–4), exits at (1) and goes to the ICP-MS instrument.

Results and discussion

Hydrochloric or nitric acids (more than 2 M) are normally employed in order to retain actinides on the UTEVA-Spec. resin.¹² According to the sample's acidity, Np and Pu can be present in solution in different oxidation states. In fact, Pu can be present as Pu(III), Pu(IV) and Pu(VI), and Np as Np(IV), Np(V) and Np(VI). Since our interest is in performing quantitative determinations of the total elements and not to study their speciation, it was necessary to have them in a single oxidation state. Nevertheless, it has been observed that this resin is very appropriate for performing speciation of actinides and further research is on going in this direction. For both, the oxidation state IV was the choice. This was because: (1) Pu(III) and Np(V) are not retained on the UTEVA-Spec.; and (2) Np(VI) and Pu(VI) have the same elution time as U(VI), and therefore the isobaric interference at mass 238 between U and Pu cannot be resolved. Moreover high concentrations of ²³⁸U cause interference on Np determination at mass 237.^5,6

In order to obtain Pu and Np in the oxidation state IV, different redox agents were tested; reported in Table 2 along with the oxidation states obtained by their use for both Pu and Np. Among the different agents studied, as can be seen from the Table 2, only hydrogen peroxide is capable of giving both elements in the oxidation state (IV). Nevertheless, the procedure employing hydrogen peroxide is quite complicated and it is necessary to bring the solution to the boiling point; a potential hazard when performed in a glove-box on radioactive samples. In view of this, it was decided to employ another procedure consisting of the combined use of two redox agents. According to this procedure, the Mohr's salt was used in order to convert all Np to the Np(IV). The Pu was converted to Pu(III). After 4 min, which is the necessary time to complete the reaction to obtain all Np in the oxidation state (IV), sodium nitrite is added in order to oxidise the Pu(III) to Pu(IV). This procedure is less complicated, but sensitive to the reaction time. In Fig. 2(a) and (b) the kinetic of the reaction after addition of sodium nitrite to the sample treated with Mohr's salt is reported. A solution containing 50 ng g⁻¹ of Np and Pu was analyzed at different times using the UTEVA Spec. resin in the chromatographic column coupled on-line to the ICP-MS detector. As can be seen, after 10 min of reaction with sodium nitrite, the oxidation to state IV is complete for both the analytes.

Fig. 2 (a) Kinetc of reaction of ²³⁷Np with sodium nitrite (redox agent) and (b) kinetic of reaction of ²³⁹Pu with sodium nitrite (redox agent).

Table 2 Different redox reagents and relevant oxidation states of Np and Pu

Redox agent	Plutonium	Neptunium
Ammonium and iron (II) sulfate.	III	IV
Sodium sulfite	III and IV	IV and V
Hydrogen peroxide	IV	IV
Sodium nitrite	IV	IV and V
Ascorbic acid	III	IV

---

Retention and elution of actinides on UTEVA-Spec

U(VI), Np(IV), Pu(IV), and Th(IV) have a high value for the coefficient of distribution on the UTEVA-Spec. in 3 M nitric acid.¹² On the other hand, the three-valent actinides are not retained under these conditions independently of the concentration of the nitric or hydrochloric acid solution.¹² Consequently, Am(III) and Cm(III) are eluted at the dead time of the chromatographic system, thus avoiding isobaric interferences with Pu. When using 2 M hydrochloric acid solution, the distribution coefficient of the tetra- and hexavalent actinides decreases very rapidly. At this concentration of hydrochloric acid, Th, Pu and Np can be eluted. For U it is necessary to use a more diluted solution, e.g., 0.025 M hydrochloric acid. In Table 3, the chromatographic conditions utilized for the separation of U, Np, Pu and Th from the UTEVA-Spec. are reported. The concentration of the hydrochloric acid was changed after 5 min to 0.025 M in order to elute the U. In Fig. 3, the chromatogram obtained for a solution 3 M HNO₃ containing 50 ng g⁻¹ of Th, Np, U and Pu, and 10 ng g⁻¹ of Am, after a previous redox reaction with Mohr's salt and sodium nitrite, is reported. As can be seen, Am is eluted in the void volume of the column, and Th, Np, Pu and U are eluted in sequence in a total time of 500 s.

Fig. 3 Chromatogram of a solution containing Th, Np, Pu and U (50 ng g⁻¹ of each) and Am (10 ng g⁻¹), obtained according to the elution conditions of Table 3.

Table 3 Chromatographic elution program

Step	Time/min	Flux/ml min^−1	Eluent	Valve position	Analytical process
1	0.0	0.5	HNO3, 3 M	Off	Column precondition
2	0.4	0.5	HNO3, 3 M	On	Sample Injection and elements retention
3	1.7	1	HCl, 2 M	On	Th, Np and Pu elution
4	5.0	1	HCl, 0.025 M	On	U elution
5	8.6	1	HNO3, 3 M	Off	Column precondition

---

Use of a complexing agent (oxalic acid) in the eluent solution

The resolution between the peaks of Th, Np and Pu, relative to that of U, can be increased by adding oxalic acid in the eluent solution consisting of hydrochloric acid. The oxalic acid can vary, in different ways, the constant of repartition of the tetra- and hexavalent actinides according to the value of the constant of formation for the different complexes.¹³ Oxalic acid with a concentration range of 0.02–0.1 M was added to the 2 M hydrochloric acid solution. To avoid problems of carbon deposition on the torch and on the cone of the instrument, the maximum concentration of oxalic acid that can be used is 0.1 M. Fig. 4 shows the chromatogram obtained for one solution containing 50 ng g⁻¹ of Th, U, Np and Pu, and 10 ng g⁻¹ of Am, according to the elution program reported in Table 3. The only change was the presence of 0.1 molar oxalic acid in the solution of 2 M hydrochloric acid in step 3. As can be seen in Fig. 4, the separation of U from the other analytes is improved.

Fig. 4 Chromatogram of a solution containing Th, Np, Pu and U (50 ng g⁻¹ of each) and Am (10 ng g⁻¹) adding 0.1 M oxalic acid to the eluent at step 3 of Table 3.

Linear dynamic range

In order to evaluate the linear dynamic range for the different analytes, solutions containing different concentrations varying between 10 and 100 ng g⁻¹ were analyzed. For each value of concentration, two measurements were performed. A decrease in the peak areas was observed due to the loss of the extractant adsorbed on the inert support during subsequent elution runs. For this reason it was decided to add ²³⁰Th, ²³³U and ²⁴⁴Pu to the sample as internal standards. All the calibration curves were obtained by using the ratio between the area of the analyte and the relative spike added to the sample versus the ratio of the concentration between the analyte and the spike. For Np, since no spike is available due to the fact that ²³⁷Np is the only long-lived isotope that can be used with an ICP-MS, the calibration curve was obtained using the spike applied for the U determination. For all four analytes the parameters of the calibration curves, calculated by means of the method of linear regression had r² = 0.999.

Detection limit and precision

The detection limit was calculated by means of repeated measurements of the blank. The blank was constituted by a solution of Mohr's Salt (0.1 M) and sodium nitrite (0.2 M) in HNO₃ (3 M). Assuming that the minimum detectable is equal to three times the standard deviation of the measured signal of the blank, it was possible to calculate the detection limit. Using our experimental conditions the detection limit was typically equal to 0.7 ng g⁻¹. The precision, based on the relative standard deviation of the peak area calculated on the basis of seven repeated measurements, was always less than 2% for the concentration range 10–100 ng g⁻¹.

Validation of the analytical procedure

In order to validate the analytical procedure, two different reference materials were analyzed. One consisted of one sediment (IAEA 135) certified for the Th content, and the other one was a sample analyzed for ²³⁸U and ²³⁸Pu by several analytical techniques.¹⁴

The certified sediment was treated in a clean laboratory class 10-100. The dissolution was performed by means of a digester operating with microwaves, and ultra pure acid. After that, the sample solution was introduced into a glove-box and ²³⁷Np was added in known amounts in order to validate the procedure for this element. The amount of Np added was of the same order of magnitude as the other actinides in the sample solution (several ng g⁻¹). Four different aliquots were prepared from this sample solution. For each of the four aliquots, ²³³U and ²³⁰Th were added in a concentration equal to 10 ng g⁻¹ for isotope dilution analysis. In the other three aliquots, increasing amounts of ²³⁸U, ²³²Th and ²³⁷Np were added in order to calculate the concentration of the three elements by means of the standard additions method. All solutions were analyzed with chromatographic separation using UTEVA-Spec. and ICP-MS. For all aliquots, the reaction with Mohr's salt and sodium nitrite was performed. A 250 µl sample was injected onto the column. In Fig. 5 a typical chromatogram is reported. The appearance of a peak at 300 s is imputed to the change in the eluent at this time (c.f., Table 3 step 4). This peak shape does not influence the quantitative determination of U because it is reproducible in the chromatograms of the spiked samples.

Fig. 5 Chromatogram obtained for the sample IAEA 135.

In Table 4 the parameters of the calibration curves obtained for the different samples are reported together with the experimental values, expected values and certified values in terms of concentration. As can be seen, the experimental values are in good agreement with the expected values.

Table 4 Parameters of the calibration curve. Experimental and expected values, together with interval confidence at 95% for n = 3, obtained for the IAEA 135 sample

Element	Calibration curve	Experimental values/ng g^−1	Expected values/ng g^−1
a Certificate value. b Recommended value. c Added to the sample.
^232Th	y = 0.66 x + 1.34 r^2 = 0.999	202 ± 5.6	215 ± 11^a
^238U	y = 0.69 x + 0.80 r^2 = 0.9990	100 ± 3.4	105^b
^237Np	y = 0.91 x + 3.32 r^2 = 0.9991	325 ± 8.4	337^c

---

The second sample analyzed for the validation of the analytical procedure consisted of a leaching solution containing ²³⁸U and ²³⁸Pu. This sample was previously analyzed in the laboratory by several analytical procedures based on independent techniques.¹⁴ Also, for the calculation of the concentration, the method of internal standard additions was performed. The same experimental conditions used for the analysis of the sample IAEA 135 were also used for the analysis of this leaching solution. In Fig. 6 the chromatogram obtained for this solution is reported. In Table 5 are reported the parameters relative to the calibration curve obtained for ²³⁸U and ²³⁸Pu, together with the concentrations obtained versus the expected concentrations, according to the other measurements performed in the laboratory. Moreover, the same leaching solution was also analyzed using a method developed in the laboratory based on oxidation with silver oxide and chromatographic elution with DAP (40 mM) in HNO₃ (0.6 M) on IonPac CS10.⁶ The results are also compared in Table 5. As can be seen they are all in good agreement. However the accuracy, at least for ²³⁸U, is better when using UTEVA-Spec. than the CS10 column. In the case of UTEVA-Spec., the chromatography based on the extraction mechanism improves the efficiency of elution with respect to that based on a cation-exchange mechanism.

Fig. 6 Chromatogram obtained for the leaching solution.

Table 5 Calibration curve parameters relative to ²³⁸U and ²³⁸Pu, obtained by the standard additions method: measured and expected values, together with standard deviation, for the ²³⁸Pu and ²³⁸U determination in a leaching solution by UTEVA-Spec. and CS10 columns

Element	Calibration curve	UTEVA value/ng g^−1	Expected value/ng g^−1	CS 10 value/ng g^−1
^238Pu	y = 0.94 x + 0.76 r^2 = 0.9991	90 ± 2	92 ± 2	91 ± 2
^238U	y = 0.74 x + 0.79 r^2 = 0.993	19.9 ± 0.4	20 ± 1	21 ± 1

---

Real samples

To verify the feasibility of the application of the proposed analytical procedure to a real sample, Np, U and Pu were determined for an irradiated nuclear fuel sample, which was dissolved in a hot cell following the standard procedure.¹⁵

The quantity of Np, U and Pu was measured according to the standard additions method. Fig. 7A shows a typical chromatogram for Np determination. The peak relevant to U found by scanning mass 237 is due to the interference of ²³⁸U on ²³⁷Np and is a good example of the need of a chromatographic separation between U and Np for samples of high U concentration.

Fig. 7 Chromatogram obtained for a spent fuel solution. A, Separation Np/U. The peak relevant to U at mass 237 is due to the interference of ²³⁸U on ²³⁷Np. The others U isotopes are also present. B, Separation Pu/U.

The parameters for the Np calibration curve are: y = 1.06 x + 1.39, r² = 0.999₇. Calculation of Np concentration, according to the above mentioned parameters, gives 13.2 ± 0.2 ng g⁻¹.

Fig. 7B shows the chromatogram relative to ²³⁹Pu and ²³⁵U determination in the diluted solution. Table 6 reports the calculated concentrations, referred to the diluted spent fuel solution analysed, for ²³⁵U and ²³⁹Pu as well as the total U and Pu concentrations, obtained by the analytical procedure proposed as well as by TIMS.

Table 6 ²³⁵U and ²³⁹Pu concentrations, together with the standard deviations, in the diluted solution, obtained by the analytical procedure proposed and by TIMS

Element	Analyzed solution concentration
	Proposed analytical procedure	TIMS
^235U	69 ± 2 ng g^−1	67.2 ± 0.6 ng g^−1
^239Pu	90 ± 2 ng g^−1	94.7 ± 0.5 ng g^−1
Utotal	10 ± 1 µg g^−1	10.5 ± 0.3 µg g^−1
Putotal	0.18 ± 0.02 µg g^−1	0.18 ± 0.01 µg g^−1

---

In order to obtain the isotopic composition for the irradiated fuel solution, Pu and U isotopes were measured. The values in weight % are reported in Table 7 and are in agreement with those obtained by isotope dilution thermal ionization mass spectrometry after chemical separation of U and Pu.

Table 7 Pu and U isotopic composition obtained from an irradiated nuclear fuel, analyzed by the analytical procedure proposed and by TIMS

Isotope	Weight (%)		Isotope	Weight (%)
	Proposed analytical procedure	TIMS		Proposed analytical procedure	TIMS
^238Pu	3.1 ± 0.5	3.4 ± 0.4	^234U	0.03 ± 0.01	0.014 ± 0.005
^239Pu	50 ± 7	51.3 ± 3.6	^235U	0.7 ± 0.1	0.64 ± 0.05
^240Pu	24 ± 3	23.2 ± 1.9	^236U	0.7 ± 0.1	0.60 ± 0.05
^241Pu	14 ± 2	13.5 ± 0.7	^238U	98.5 ± 1.4	98.7 ± 0.9
^242Pu	9 ± 1	8.4 ± 0.5

---

Pre-concentration

To lower the analytical procedure detection limit down to the pg g⁻¹ level, the chromatograph–valve–column–ICP-MS system was modified to carry out an on-line pre-concentration of an actinide aqueous solution. The scheme shown in Fig. 1 is used.

In this way a pre-concentration factor that is proportional to the sample volume flowing through the column is obtained. The advantage is evident when very dilute aqueous solutions have to be analysed. Fig. 8A shows a chromatogram of a waste solution from a uranium enrichment plant. The concentration was about 1 ng g⁻¹, and the chromatogram was obtained with the methodology proposed with the standard configuration. As can be seen, the analytical signal is very disturbed with a low signal-to-noise ratio, and, in this condition, it is difficult to determinate the concentration. Fig. 8B shows the chromatogram of 5 ml of the same solution, analyzed with the preconcentration system described above. In both the measurements the same eluents and conditions were used. In the second case a larger signal-to-noise ratio was obtained and consequently the concentration calculated has a lower error.

Fig. 8 Chromatogram obtained for a waste solution containg 1 ng g⁻¹ of Th and U, respectively: A, without on-line pre-concentration and B, with on-line pre-concentration.

Conclusion

The UTEVA-Spec. resin based on the extraction mechanism can be used for packing chromatographic columns to be used on-line with an ICP-MS detector. The advantage of this column is the selectivity for actinides. By adding a complexing agent such as oxalic acid to the 2 M hydrochloric acid eluent solution, the separation between uranium and the other actinides is improved. Therefore the method is applicable to samples containing a large amount of uranium. Of particular interest is the possibility of performing on-line pre-concentration, using the same column both for pre-concentration and chromatographic separation. Moreover the method can be exploited to perform speciation studies of actinides.

acknowledgement

Maria Betti and Roger Fuoco dedicate this paper to Prof. P. Papoff on the occasion of his 80th birthday. The authors thank Dr C. Apostolidis for useful discussions.

References

1. N. Jakubowsky, L. Moens and F. Vanhaecke, Spectrochim. Acta, Part B, 1998, 53, 1739.
2. J. S. Becker, R. S. Soman, K. L. Sutton, J. A. Caruso and H. J. Dietze, J. Anal. At. Spectrom., 1999, 14, 933.
3. J. S. Becker, H. J. Dietze, J. A. McLean and A. Montaser, Anal. Chem., 1999, 15, 3077.
4. J. S. Becker and H. J. Dietze, J. Anal. At. Spectrom., 1999, 14, 1493.
5. M. Betti, J. Chromatogr. A, 1997, 789, 369.
6. J. M. Barrero Moreno, M. Betti and J. I. García Alonso, J. Anal. At. Spectrom., 1997, 12, 355.
7. J. M. Barrero Moreno, M. Betti and G. Nicolaou, J. Anal. At. Spectrom., 1999, 14, 875.
8. J. W. Grate and O. B. Egorov, Anal. Chem., 1998, 70, 3920.
9. C. Apostolidis, R. Molinet, P. Richir, M. Ougier and K. Mayer, Radiochim. Acta, 1998, 83, 21.
10. J. I. García Alonso, F. Sena, Ph. Arboré, M. Betti and L. Koch, J. Anal. At. Spectrom., 1995, 10, 381.
11. M. Betti, J. I. García Alonso, Ph. Arboré, L. Koch and T. Sato, in Application of Plasma Source Mass Spectrometry, ed. G. Holland and N. Eaton, Royal Society of Chemistry, London, 1993, p. 205..
12. E. P. Horwitz, M. L. Dietz, R. Chiarizia, H. Diamond, A. M. Essling and D. Graczyk, Anal. Chim. Acta, 1992, 266, 25.
13. J. J. Katz, G. T. Seaborg and L. R. Moor, The Chemistry of Actinide Elements, Chapman and Hall, London, New York, 1986, vol. 2, p. 1497..
14. D. Solatie, P. Carbol, M. Betti, F. Bocci, T. Hiernaut, V. Rondinella and J. Cobos, Fresenius' J. Anal. Chem., 2000, 368, 88.
15. J. M. Barrero Moreno, J. I. García Alonso, Ph. Arboré, G. Nicolaou and L. Koch, J. Anal. At. Spectrom., 1996, 11, 929.

---

Footnote

† Present address: European Commission, JRC, Institute for Health and Consumer Protection, TP 260, 21020 Ispra, Italy

---