Simultaneous speciation of inorganic selenium and tellurium by
inductively coupled plasma mass spectrometry following selective
solid-phase extraction separation
Chunhai Yu,a Qiantao Cai,*a Zhong-Xian Guo,a Zhaoguang Yanga and
Soo Beng Khoob
a
Centre for Advanced Water Technology, Innovation Centre (NTU), Block 2, Unit 241,
18 Nanyang Drive, Singapore 637723. E-mail: [email protected];
Fax: 165-67942791; Tel: 165-67943705
b
Department of Chemistry, National University of Singapore, Kent Ridge, Singapore
119260
Received 26th August 2003, Accepted 15th December 2003
First published as an Advance Article on the web 13th February 2004
A new method was developed for the simultaneous determination of inorganic tellurium and selenium species in
waters by inductively coupled plasma mass spectrometry (ICP-MS) following selective solid-phase extraction
(SPE) separation. Under acidic conditions, only selenium(IV) and tellurium(IV) formed complexes with ammonium
pyrrolidine dithiocarbamate (APDC), and the complexes were completely retained on a non-polar C18 cartridge.
Te(VI) and Se(VI) passed through the cartridge and remained as free species in the solution, thereafter being
determined by ICP-MS. Se(IV) and Te(IV) concentrations were obtained as the respective differences between total
selenium and Se(VI), and total tellurium and Te(VI) concentrations. The detection limits (3s) are 7 ng L21 selenium
and 3 ng L21 tellurium. Factors affecting the separation and detection were investigated. Coexisting ions did not
show significant interferences. This method has been successfully applied to the inorganic selenium and tellurium
speciation analysis of water samples with spiked recoveries of 82.3–106%.
DOI: 10.1039/b310318h
Introduction
410
Selenium is an essential micronutrient for living organisms
including humans. The optimum selenium intake from diet is
necessary to achieve good health and to reduce the risk of
certain diseases. However, it is recognized to be toxic at high
concentrations. Its biogeochemical behavior, nutritional bioavailability and toxicity are largely dependent on its chemical
species.1 A lot of organoselenium compounds, such as
selenoamino acids (selenocysteine, selenomethionine) occur
in plants, microorganisms and in their related environments,
e.g. sediments and biological fluids, as a result of the biomethylation of inorganic selenium species.2,3 However, selenium is
primarily present as inorganic species, i.e., selenite and selenate
ions, in water and soil.2 Therefore, speciation of inorganic
selenium has received extensive attention in recent years.
Most of the works were focused on hyphenated techniques, i.e.,
on-line coupling of a liquid chromatographic (LC) system with
an element specific detector such as atomic absorption spectrometry (AAS),4 atomic fluorescence spectrometry (AFS)5,6
and inductively coupled plasma mass spectrometry (ICPMS).7–12 Chromatographic techniques based on anion
exchange,4–9 cation exchange,8 reversed-phase6,10,12 and ionpairing11 were usually used for selenium species separation.
ICP-MS demonstrated its unique advantages of excellent
detection limits and multi-element capability and has therefore
been mainly used for selenium detection in recent years.7
Capillary zone electrophoresis (CZE), coupled with ICP-MS
detection,13 has also been used for selenium speciation. Besides
the hyphenated techniques described above, the hydride
generation (HG) technique in combination with AAS, AFS,
AES and ICP-MS, has become one of the useful alternatives
for selenium speciation.14 One potential drawback of HG is
that it requires selenium species to be in a particular oxidation
state before a hydride can be formed effectively. Cathodic
stripping voltammetry (CSV) was also applied to measuring
J. Anal. At. Spectrom., 2004, 19, 410–413
Se(IV) and Se(VI).15 Chung et al developed a procedure to separate Se(IV) and Se(VI) by liquid–liquid extraction prior to AAS
determination.16
Tellurium is usually associated with selenium in minerals
and the earth’s crust at trace or ultra-trace levels, and has
similar chemical and physical characters. However, it is usually
regarded as a non-essential element and is harmful to
humans.17 Speciation of tellurium can be performed on the
basis of on-line hyphenation of chromatographic separations
and element-specific detectors,13,18 and graphite furnace AAS
combined with liquid–liquid extraction.16
Up to now, Se(IV) and Se(VI) have been the primary selenium
species of interest in aqueous ecosystems.5,7,9,19 In nature,
tellurium is present in the form of tellurite [TeO322, Te(IV)] and
tellurate [TeO422, Te(VI)] and its concentration in natural
waters is generally at trace levels (sub-ng L21).20 Therefore,
accurate and highly sensitive methods are needed for the
simultaneous speciation of inorganic Te and Se in all compartments of aquatic ecosystems.
On the basis of our previous work for Te speciation,21 the
aim of this study was to develop a simple and highly sensitive
method for the simultaneous determination of inorganic Se and
Te species in waters. The heart of the approach is the use of
non-polar silica-based C18 sorbent-containing SPE cartridges
for selective retention of APDC complexed Se(IV) and Te(IV).
Te(VI) and Se(VI) did not form APDC complexes and passed
through the cartridge as free species in the solution, thereafter
being determined by ICP-MS.
Experimental
Instrumentation
An Agilent 7500 quadrupole ICP-MS (Yokogawa Analytical
System, Kyoto, Japan), used for elemental detection, was
equipped with a dual-pass quartz spray chamber (Scott type,
This journal is ß The Royal Society of Chemistry 2004
water-cooled to 2 uC), a quartz concentric nebulizer, a standard
quartz torch (2.5 mm id), a nickel sampling cone and a nickel
skimmer cone. The optimized operating parameters, unless
otherwise stated, were as follows: rf power, 1350 W; plasma
gas, 15 L min21 Ar; auxiliary gas, 0.9 L min21 Ar; carrier gas,
1.20 L min21 Ar; make-up gas, zero; sampling depth, 6.9 mm;
monitoring isotopes, 82Se and 128Te; detector mode, pulse; and
integration time, 0.5 s. An ASX-100 autosampler (Cetac
Technologies, Omaha, NE, USA) with Teflon vials was used
for sample introduction and directly connected to the
concentric nebulizer.
A VacMaster-10 sample processing station (Supelco, PA,
USA) with adjustable speed was used to push sample flowing
through the cartridges. The C18-bonded silica gel-based
cartridges (with sorbent mass of 500 mg, reservoir volume of
10 mL) were obtained from International Sorbent Technology
(Mid-Glamorgan, UK). Each cartridge was preconditioned
with 2 mL of methanol (HPLC grade, Merck) and then 2 mL of
deionized water before use. Samples were filtered through
0.2 mm Nylon membrane disc filters (Whatman, MI, USA)
prior to testing. All labware, such as bottles and volumetric
flasks, used in this study were made of polypropylene, fluoropolymer or low-density polyethylene (Nalgene, Rochester, NY,
USA). Experiments were carried out at ambient temperature of
24 ¡ 2 uC.
Chemicals and materials
All chemicals were analytical reagent grade unless otherwise
stated. Reagent grade water with a specific resistance of
18.2 MO cm or greater was obtained from a Milli-Q water
purification system (Millipore, Bedford, MA, USA). Stock
standards (1000 mg L21, as Se or Te), of Se(IV), Se(VI), Te(IV)
and Te(VI) were obtained by respectively dissolving appropriate amounts of Na2SeO3, Na2SeO4?10H2O, Na2TeO3 and
Na2TeO4?2H2O (all from Aldrich) in water, and storing the
solutions in a refrigerator at 4 uC. Working solutions were
prepared daily by appropriate dilutions of stock solutions. An
APDC solution of 1.0% (w/v) was prepared daily by dissolving
an appropriate amount of APDC (w98%, Fluka, Buchs,
Switzerland) in water. Ultrapure nitric acid (Ultrex II,
J.T. Baker, NJ, USA) was used for sample acidification.
Procedure
For determination of Se(VI) and Te(VI), 20 mL of the filtered
water sample, 0.40 mL of concentrated nitric acid and 1.0 mL
of 1.0% APDC solution were added to a 50 ml snap cap vial.
The vial was then shaken for 3 min and kept for about 10 min
at room temperature to allow the complete complexation of
Se(IV) and Te(IV) by APDC. Thereafter, 10.0 mL of the solution
was allowed to flow through the preconditioned cartridge at a
flow rate of 1.0 mL min21. The effluent from the cartridge was
collected in a vial for subsequent ICP-MS quantification.
For determination of total selenium and tellurium, a portion
of the filtered water sample was acidified to 2% (v/v) HNO3,
and then directly introduced into the ICP-MS. The concentrations of Se(IV) and Te(IV) were then calculated as the respective
concentration differences between total selenium and Se(VI),
and total tellurium and Te(VI).
Results and discussion
Optimization of ICP-MS operating parameters
The determination of selenium by ICP-MS suffered from two
main problems.7,10,11,19 The first is its lower degree of
ionization in the plasma due to the high first ionization
potential of Se. The second is the elemental sensitivity
distributed over six isotopes among which the most abundant,
80
Se (49.7%) and 78Se (23.6%), isotopes suffered from strong
background interferences due to the 40Ar40Ar1 and 40Ar38Ar1
dimers. The detection of nuclides for Se at m/z 78 or 80 was
hence made impossible, as the resolution of quadrupole mass
spectrometers was not sufficient to resolve atomic and
molecular species having the same nominal mass. Therefore,
owing to its much higher signal-to-noise (S/N) ratio, the
isotope 82Se (9.2%) was generally selected for selenium
determination, although it has an abundance of only 9.2%
and may suffer from other interferences, such as 1H81Br,
resulting from a co-existing substance rather than argon
polyatomic species. As for the tellurium determination, the
most abundant isotopes 128Te and 130Te (31.7% and 33.8%,
respectively) did not show significant differences in terms of
their sensitivities and interferences from others. 128Te was
monitored throughout this study. The ICP-MS operating
parameters such as rf forward power, torch position, plasma,
auxiliary and carrier gas flow, as stated in the Experimental
section, were selected on the basis of the best net ion signals of
Se and Te and their S/N ratio.
Owing to the use of nitric acid in standard and sample
preparation, etc., and APDC as chelating reagent, their effects
on the net Se and Te ion intensities were examined. The results
showed the net Se and Te signals decreased with increase of
nitric acid concentration up to 0.5 mol L21. For example,
the net signals of 5.0 mg L21 Te(IV) in nitric acid of 0.2 or
0.5 mol L21 decreased by 25% and 45%, respectively, in
comparison with that in 0.002 mol L21 nitric acid. The
magnitude of suppression was found to be indifferent to both
Te(IV) and Te(VI), and also both Se(IV) and Se(VI). It should be
noted that the acid effects on the net Te and Se ion signals were
strongly dependent on the instrumental operating conditions
employed. The final concentration of nitric acid was controlled
to be 0.3 mol L21, i.e., 2% (v/v), which was usually the acid
concentration employed for water sample preservation. The
presence of APDC up to 0.15% did not show significant effects
on the intensities of Se(VI) and Te(VI). However, the existence of
APDC ¢ 0.02% resulted in suppression of Te(IV) and Se(IV)
responses. Meanwhile, addition of APDC also caused greater
variations in Te(IV) and Se(IV) signals. For example, the net
signals of 5.0 mg L21 Te(IV) in the presence of 0.04% and 0.06%
APDC were suppressed by 10.8% and 14.7%, respectively. Such
effects of APDC on Te(IV) and Se(IV) signals could be attributed
to the formation of water insoluble Te(IV)–APDC and Se(IV)–
APDC complexes under the conditions employed.
Effects of acidity on retention and separation of selenium and
tellurium
The influences of acidity on the retention efficiency of Te(IV),
Te(VI), Se(IV) and Se(VI) on the non-polar C18 cartridge in the
presence and absence of APDC were investigated over the
acidity range 0.02–1.0 mol L21 H1. The acidity values were
adjusted with nitric acid prior to the addition of APDC (final
concentration of 0.05%). The retention efficiency of each
species on the C18 cartridge was determined by measuring the
concentration difference of the species in the original solution
and the effluent solution.
The experimental results indicated that, in the absence of
APDC, only 0–4.4% of Se(IV), 0–3.2% of Se(VI), 0–0.2% of Te(IV)
and 0–3.5% of Te(VI) were retained on the C18 column in the
acidity range studied. Under the conditions investigated, Se(IV),
Se(VI), Te(IV) and Te(VI) mainly existed as water soluble species
H2SeO3, HSeO42, Te(OH)31 and Te(OH)6, respectively,22 and
were therefore not retained on the non-polar cartridge. On the
other hand, as shown in Fig. 1, Se(IV) and Te(IV), in the presence
of APDC as a chelating agent, were retained at 99.7–100% and
98.3–100%, respectively, while Se(VI) and Te(VI) fully passed
through the cartridge and remained as free species in the
solutions.
The retention of Te(IV) and Se(IV) in the presence of APDC
J. Anal. At. Spectrom., 2004, 19, 410–413
411
Fig. 1 Effect of acidity on the retention of 5.0 mg L21 Se(IV), Se(VI),
Te(IV) and Te(VI) on the C18 cartridge in the presence of 0.05% APDC as
chelating reagent.
Fig. 2 Effect of APDC concentration on the retention percentage of
Se(IV) and Te(IV) on the C18 cartridge.
Influence of foreign ions
was due to the formation of organic solvent-extractable Te(IV)–
APDC and Se(IV)–APDC complexes.23 The complexes were
larger neutral molecules and less polar than Te(OH)31 and
H2SeO3, and thus were more strongly retained by the non-polar
C18 sorbent. APDC was known to form complexes with Se(IV)
and Te(IV) selectively, which can be extracted into organic
solvents such as carbon tetrachloride and methyl isobutyl
ketone.23 However, the extraction efficiency was considerably
influenced by the pH or acidity of the aqueous solution. Se(VI)
and Te(VI) did not form complexes with APDC under the same
conditions and therefore could not be extracted into the
organic solvents23 and likewise could not be retained by the
non-polar C18 cartridge. For liquid–liquid extraction of Se(IV),
Se(VI), Te(IV) and Te(VI) using APDC, Chung et al.23 found that
Se(IV) and Te(IV) were quantitatively extracted into chloroform–
carbon tetrachloride mixed solvent over a wide acidity range
from 5 mol L21 HCl to pH 7, while Se(VI) and Te(VI) were only
slightly extractable in strongly acidic solution (¢4 mol L21
HCl).
The results shown in Fig.1 would therefore further suggest
similarities in the retention behaviors of Se(IV), Se(VI), Te(IV) and
Te(VI) on the C18 cartridge in the presence of APDC to their
liquid–liquid extraction characteristics. In consideration of the
use of HNO3 in sample preservation, sample and standard
preparation, etc., and its effect on the ICP-MS signal of Te as
described previously, selective and quantitative separation of
Se(IV), Se(VI), Te(IV) and Te(VI) could be performed in 0.3 mol L21
HNO3 (2% V/V).
Influence of APDC concentration
To quantitatively retain Se(IV) and Te(IV) as their complexes
with APDC on the non-polar C18 cartridge, an excess of APDC
and appropriate ratios of APDC : Te(IV) and APDC : Se(IV)
would be necessary for complete formation of Se(IV)–APDC
and Te(IV)–APDC complexes. The optimal concentration of
APDC for simultaneous separation of Se(IV) and Te(IV)
was examined at Se(IV) and Te(IV) concentrations of 50 and
500 mg L21 with different concentrations of APDC. The results
are shown in Fig. 2. At both concentrations studied, Se(IV) and
Te(IV) were completely retained (¢95.0%) on the C18 column
when the APDC concentration was above 0.03%. In consideration of the competitive complexation with other metal
ions in practical cases, 0.05% APDC should be generally
sufficient and therefore used to retain Se(IV) and Te(IV)
selectively, quantitatively and simultaneously on the C18
cartridge from aqueous solution.
412
J. Anal. At. Spectrom., 2004, 19, 410–413
Trace metal ions can interfere if, under the experimental conditions, they effectively compete for complexation of APDC
and are retained on the non-polar C18 cartridge. Other
materials may also interfere by competitive complexation
and masking of Te(IV) and Se(IV) to form complexes not
retainable on the non-polar C18 cartridge. In consideration of
applying the developed method to environmental source water
samples, the main ions coexisting in water were investigated for
their interferences with the simultaneous determination of Se(IV),
Se(VI), total Se, Te(IV), Te(VI) and total Te. When the recommended
procedure was applied to the simultaneous determination of
2.40 mg L21 Se(IV), 3.00 mg L21 Se(VI), 1.50 mg L21 Te(IV) and
2.00 mg L21 Te(VI), the presence of 0.50 mg L21 As(III), Sb(III),
Cu(II), Ni(II), Co(II), 78 mg L21 Na1, 46 mg L21 K1, 34 mg L21
NH41, 38 mg L21 SO422, 8.0 mg L21 Br2, 39 mg L21 H2PO42,
24 mg L21 HCO32, 24 mg L21 Ac2, 248 mg L21 NO32 and
142 mg L21 Cl2 gave no observable interferences. It should be
noted that other water-soluble organic selenium species and
organic tellurium species, if existing in the samples, may yield
certain error on the selenium and tellurium speciation results,
respectively. Fortunately, inorganic selenium and tellurium
were the mostly dominant forms in waters, while organic
species were seldom found.2,5,7,20
Method performance in inorganic Se and Te speciation of water
samples
Selective retention of Se(IV) and Te(IV) in the presence of APDC
on the non-polar C18 cartridge could be applied to separate
Se(IV) and Te(IV) from Se(VI) and total Te present in the aqueous
sample. The respective differences in retention of Se(IV) and
Se(VI), and Te(IV) and Te(VI), in the presence of APDC, therefore
make it possible to develop a simple, sensitive and simultaneous speciation method of inorganic selenium and tellurium.
The concentrations of Se(VI) and Te(VI) in the aqueous phase can
be determined by ICP-MS after SPE separation of Se(IV) and
Te(IV). The Se(IV) and Te(IV) concentrations can be obtained as
the respective differences between total selenium and Se(VI), and
total tellurium and Te(VI) concentrations. The total tellurium
and selenium concentrations were obtained by ICP-MS from
the filtered original water sample.
By using the proposed method, calibration curves for Se(VI),
total Se, Te(VI) and Te(VI) were established with calibration
levels of 0, 0.5, 2.0 and 5.0 mg L21 with linear regression
correlation coefficients of greater than 0.995 for all the species.
The calibration curves were still linear even up to 50.0 mg L21 in
the pulse detector mode. The limits of detection (LOD) by the
described procedure were calculated to be 7 ng L21 for Se and
Table 1 Speciation and recoveries of Se(IV), Se(VI), total Se, Te(IV), Te(VI) and total Te in water samples (n ~ 3)
Analytical resulta
Spiked recovery
Se(IV)
Samples
Drinking water
Product water
Raw water
Waster water
a
Total
Se/mg L21
0.342
0.144
0.438
1.312
¡
¡
¡
¡
0.021
0.008
0.044
0.062
Se(VI)/mg L21
0.298
0.141
0.432
1.162
¡
¡
¡
¡
Se(VI)
Te(IV)
Te(VI)
Total Te/ Te(VI)/ Te(IV)/ Added/ R
Added/ R
Added/ R
Se(IV)/mg L21 mg L21 mg L21 mg L21 mg L21 (%) mg L21 (%) mg L21 (%)
0.018 0.044 ¡ 0.027
0.007
N.D
0.056
N.D
0.097 0.150¡0.115
N.D
N.D
N.D
N.D
N.D
N.D
N.D
N.D
N.D
N.D
N.D
N.D
3.0
3.0
3.0
3.0
96.3
91.3
99.0
93.6
3.5
3.5
4.0
4.0
98.9
95.5
90.3
89.6
1.5
1.5
1.5
1.0
99.9
82.3
94.4
106
Added/ R
mg L21 (%)
2.0
1.5
2.0
1.5
92.8
106
97.3
96.5
The mean ¡ standard deviation from triplicate measurements; N.D ~ not detectable, less than detection limit.
3 ng L21 for Te, based on 3 times the standard deviations of the
respective blanks.
2
Application to analysis of water samples
4
The method was applied to the simultaneous determination of
Se(IV), Se(VI), total Se, Te(IV), Te(VI) and total Te in water samples
from different sources. Meanwhile, recovery was also examined
by spiking water samples with various Se and Te species at
different concentration levels. The results summarized in
Table 1 demonstrate that Se(IV), Se(VI), total Se, Te(IV), Te(VI)
and total Te in the water samples can be successfully
determined by ICP-MS based on the simultaneous selective
retention of Se(IV) and Te(IV) in the presence of APDC on the
non-polar C18 cartridge and therefore their separation from
Se(VI) and Te(VI). For Se(VI) and total Se at sub-mg L21 levels, the
determination precisions varied between 5–13%. The relative
standard deviations (RSDs) on the determination of Se(IV) in
two samples were 60% and 77%, respectively. The relatively
larger RSDs were due to trace levels of Se(IV) and the combined
measurement uncertainty from the determination of Se(VI)
and total Se. Nevertheless the method’s precision should be
acceptable for such trace levels. Recoveries in ranges of 93.6–
99.0%, 89.6–98.9%, 82.3–106% and 92.8–106% were obtained
for Se(IV), Se(VI), Te(IV) and Te(VI), respectively, at concentration
levels of 1.0–4.0 mg L21. These recovery results could also
indicate that no significant oxidation and reduction occurred
for the spiked selenium and tellurium species during the sample
preparation and SPE separation.
3
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Acknowledgements
Chunhai Yu is grateful to the Institute of Environmental
Science and Engineering, Singapore (formerly known as
Environmental Technology Institute) for providing a scholarship to carry out this research work.
References
1 J. M. Besser, T. J. Canfield and T. W. la Point, Environ. Toxicol.
Chem., 1993, 12, 57–72.
19
20
21
22
23
G. A. Pedersen and E. H. Larson, Fresenius’ J. Anal. Chem., 1997,
358, 591–598.
P. T. Zawislanski, S. M. Benson, R. Terberg and S. E. Borglin,
Environ. Sci. Technol., 2003, 37, 2415–2420.
Y. Q. Zhang and W. T. Frankenberger, Jr., Sci. Total Environ.,
2001, 269, 39–47.
D. Wallschlager and N. S. Bloom, J. Anal. At. Spectrom., 2001, 16,
1322–1328.
J. L. Gomez-Ariza, D. Sanchez-Rodas, M. A. C. de la Torre,
I. Giraldez and E. Morales, J. Chromatogr. A, 2000, 889, 33–39.
D. Wallschlager and R. Roehl, J. Anal. At. Spectrom., 2001, 16,
922–925.
P. Moreno, M. A. Quijano, A. M. Gutierrez, M. C. Perez-Conde
and C. Camara, J. Anal. At. Spectrom., 2001, 16, 1044–1050.
E. Vassileva, A. Becker and J. A. C. Broekaert, Anal. Chim. Acta,
2001, 441, 135–146.
J. M. Marchante-Gayon, I. Feldmann, C. Thomas and
N. Jakubowski, J. Anal. At. Spectrom., 2001, 16, 457–463.
B. Gammelgaard, L. Bendahl, U. Sidenius and O. Jons, J. Anal.
At. Spectrom., 2002, 17, 570–575.
J. Zheng, M. Ohata, N. Furuta and W. Kosmus, J. Chromatogr. A,
2000, 874, 55–64.
C. Casiot, O. F. X. Donard and M. Potin-Gautier, Spectrochim.
Acta, Part B, 2002, 57, 173–187.
A. Chatterjee, Y. Shibata, A. Tanaka and M. Morita, Anal. Chim.
Acta, 2001, 436, 253–263 and references cited therein.
I. D. Gregori, M. G. Lobos and H. Pinochet, Water Res., 2002, 36,
115–122.
C. H. Chung, E. Iwamoto, M. Yamamoto and Y. Yamamoto,
Spectrochim. Acta, Part B, 1984, 39, 459–466.
T. Sadeh in The Chemistry of Organic Selenium and Tellurium
Compounds, ed. S. Patai, Wiley, Chichester, 1987, vol.2, pp. 367–
376.
R. S. T. Basnayake, J. H. Bius, O. M. Akpolat and T. G. Chasteen,
Appl. Organometal. Chem., 2001, 15, 499–510.
K. Tirez, W. Brusten, S. V. Roy, N. D. Bruker and L. Diels,
J. Anal. At. Spectrom., 2000, 15, 1087–1092.
T. Narukawa, J. Anal. At. Spectrom., 1999, 14, 75–80.
C. H. Yu, Q. T. Cai, Z. X. Guo, Z. G. Yang and S. B. Koo, Anal.
Bioanal. Chem., 2003, 376, 236–242.
C. F. Baes, Jr. and R. E. Mesmer, The Hydrolysis of Cations,
John Wiley & Sons, New York, 1976, pp. 384–396.
K. Ueno, T. Imamura and K. L. Cheng, Handbook of Organic
Analytical Reagents, CRC Press, Boca Raton, Florida, 2nd edn.,
1992, pp. 463–466.
J. Anal. At. Spectrom., 2004, 19, 410–413
413