Analytica Chimica Acta 678 (2010) 18–25

Contents lists available at ScienceDirect

Analytica Chimica Acta

journal homepage: www.elsevier.com/locate/aca

Comparison of inductively coupled plasma spectrometry techniques for the

direct determination of rare earth elements in digests from geological samples
Francisco Ardini a , Francesco Soggia a , Francesco Rugi b , Roberto Udisti b , Marco Grotti a,∗
a
Department of Chemistry and Industrial Chemistry, University of Genoa, Via Dodecaneso 31, 16146 Genoa, Italy
b
Department of Chemistry, University of Florence, Via della Lastruccia 3, 50019 Sesto F.no (FI), Italy

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

Article history: Inductively coupled plasma quadrupole mass spectrometry (ICP-QMS), ICP sector field mass spectrome-
Received 19 April 2010 try (ICP-SFMS) and ICP atomic emission spectrometry (ICP-AES) were compared with regard to the direct
Received in revised form 19 July 2010 determination of rare earth elements (REEs) in geological samples. In order to reduce the polyatomic
Accepted 22 July 2010
interferences occurring in ICP-QMS, the use of a cooled spray chamber was optimized, obtaining a sig-
Available online 30 July 2010
nificant decrease of the oxide ions formation (about 50%) and a consequent mitigation of the interfering
effects. Precision and accuracy of the method were demonstrated by the analyses of sediment and soil cer-
Keywords:
tified reference materials. ICP-SFMS working in high-resolution mode also provided accurate results, with
Inductively coupled plasma
Atomic emission spectrometry
similar precision to ICP-QMS (RSD%: 3–8%) and comparable or better limits of detection. Quantification
Atomic mass spectrometry limits of the procedures were 18–52 ng g−1 and 10–780 ng g−1 for sector field- and quadrupole-ICP-MS,
Rare earth elements respectively. Accurate and precise determination of most REEs was also achieved by ICP-AES using both
Cooled spray chamber pneumatic and ultrasonic nebulization, after a careful selection of the emission lines and compensation
Ultrasonic nebulization for non-spectral interferences by internal standardization. The three techniques were finally applied to
glaciomarine sediment samples collected in Antarctica, providing comparable analytical data on REE
abundance and depth pattern.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction Established methods for REEs determination in geological mate-

The rare earth elements (REEs) include lanthanides (atomic
number between 57 and 71) and yttrium, which has a chemical
behaviour similar to the heavier lanthanides. REEs are generally
divided into light (from La to Sm) and heavy (from Eu to Lu) REEs.
They exhibit similar chemical and physical properties, including
low solubility and immobility in the terrestrial crust, and are gen-
erally in the +3 valence state over a wide range of oxygen fugacity
values. However, cerium can be partly or wholly in the +4 state
at the oxygen fugacity of the Earth’s surface and europium can be
partly in the +2 state at the low oxygen fugacity of the Earth’s inte-
rior. This gives rise to “anomalies” in the REEs pattern of abundance,
useful in studying the geochemical evolution of the Earth [1]. REEs
are also used as tracers to track emissions from petroleum refining
procedures [2] and their increasing use in industrial and techno-
logical applications is enhancing emissions into the environment
with accumulation in soil, sediments and, possibly, biota [3]. As a
result, the determination of REEs in geological samples is needed
for both geological and environmental sciences.
∗ Corresponding author. Tel.: +39 010 3536113; fax: +39 010 3536190.
E-mail address: grotti@chimica.unige.it (M. Grotti).
rials include instrumental neutron activation analysis (INAA) [4],
isotopic dilution mass spectrometry (IDMS) [5], inductively cou-
pled plasma atomic emission spectrometry (ICP-AES) [6] and
inductively coupled plasma mass spectrometry (ICP-MS) [7]. INAA
and IDMS show adequate sensitivity to detect most REEs at the
concentrations at which they occur in the geosphere; however,
these techniques are quite expensive and time-consuming. More-
over, IDMS cannot be applied to mono-isotopic elements, such
as Y, Pr, Tb, Ho and Tm. ICP-AES allows fast and multi-elemental
analysis of REEs with sufficient sensitivity, but its performance for
many REEs is weakened by the complexity of their emission spec-
tra, together with a number of severe interferences from major
elements. Hence, matrix separation by cation-exchange [6] and
a careful line selection [8] are usually required. On the other
hand, ICP-MS provides simpler spectra and lower detection lim-
its than ICP-AES, allowing the REEs determination in geological
samples, generally avoiding matrix separation [7,9,10]. However,
the formation of oxide and hydroxide ions in the ICP source is
a common feature of ICP mass spectra and can lead to rele-
vant spectral interferences, such as 135 BaO+ on 151 Eu+ , 144 NdO+
on 160 Gd+ , 144 SmO+ on 160 Gd+ and 143 NdO+ on 159 Tb+ . Hence,
several methods have been adopted to overcome these interfer-
ences, including measurement of doubly charged ions [11], use
of algebraic corrections [12–15], modifications of instrumental

0003-2670/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.aca.2010.07.036
 F. Ardini et al. / Analytica Chimica Acta 678 (2010) 18–25 19

Table 1 Table 2
ICP-QMS operating parameters. ICP-SFMS operating parameters.

Parameter Value Parameter Value

RF Power 1500 W RF Power 1250 W

Plasma gas flow rate 14.5 L min−1 Plasma gas flow rate 16.0 L min−1
Auxiliary gas flow rate 1.65 L min−1 Auxiliary gas flow rate 1.0 L min−1
Nebulizer gas flow rate 0.92 L min−1 Nebulizer gas flow rate 1.0 L min−1
Sample uptake rate 75 ␮L min−1 Sample uptake rate 0.9 mL min−1
Internal standard uptake rate 35 ␮L min−1 Washing time 3 min
Lens voltage Autolens® Take up time 50 s
Reaction cell rod offset −1 V Acquisition mode E-scan: electric scanning over small mass range
Quadrupole rod offset −8 V Resolution mode High (m/m ≈ 10,000)
RF amplitude 150 V Settling time 0.001–0.047 s
Axial field voltage 300 V Samples per peak 30
Cell path voltage −28 V Sample time 0.02 s
RPaa 0 Peak duration 0.720–1.080 s
RPqa 0.25 Run and passes 4–4
Dwell time 50 ms
Measured ions
Sweeps 40 89
Analytes Y+ ; 139 La+ ; 140 Ce+ ; 141 Pr+ ; 146 Nd+ ; 147 Sm+ ;
Replicates 7 151
Eu+ ; 157 Gd+ ; 159 Tb+ ; 163 Dy+ ; 165 Ho+ ; 166 Er+ ;
169
Measured ions Tm+ ; 172 Yb+ ; 175 Lu+
89
Analytes Y+ ; 139 La+ ; 140 Ce+ ; 141 Pr+ ; 146 Nd+ ; 147 Sm+ ; Internal standards 72
Ge+ ; 115 In+ ; 185 Re+
151
Eu+ ; 157 Gd+ ; 159 Tb+ ; 163 Dy+ ; 165 Ho+ ; 167 Er+ ;
169
Tm+ ; 174 Yb+ ; 175 Lu+
101
Internal standard Ru+ temperature-controlled condenser. The main operating conditions
a
Mathieu stability parameters of the cell’s quadrupole: a = 1.9 × RPa; are listed in Table 3.
q = 0.95 × RPq.

parameters [16–20] and the application of desolvation techniques
[16,21–23].
In this study, we compared ICP quadrupole mass spectrome-
try (ICP-QMS), ICP sector field mass spectrometry (ICP-SFMS) and
ICP-AES (with both pneumatic and ultrasonic nebulization) for the
direct analysis of digests from geological samples, by considering
both certified reference materials (CRMs) and marine sediment
samples of various chemical compositions from Antarctica. In order
to reduce the polyatomic interferences occurring in ICP-QMS, the
use of a cooled spray chamber was explored.
2. Experimental
2.1. Instrumentation
2.2. Chemicals
Ultrapure water was supplied by the three-step ion-exchange
system Milli-Q, fed by the reverse osmosis system Elix 3, both from
Millipore (El Paso, TX, USA). 65% nitric acid, 37% hydrochloric acid,
70% perchloric acid and 99.9999% boric acid were of Suprapur®
grade quality from Merck (Darmstadt, Germany). Trace Select®
Ultra 49% hydrofluoric acid was purchased from Sigma–Aldrich (St.
Louis, MO, USA). 1000 mg L−1 single-element standard solutions for
the interference study were prepared using the following reagent-
grade quality oxides: Pr, Sm, Gd, Dy, Ho and Er from the American
Potash & Chemical Corporation (West Chicago, ILL, USA); Nd, La, Yb,
Pr and Ho from BDH (England); Eu and Tb from Sigma–Aldrich and
Tm from Janssen (Belgium). 365 mg L−1 Hf standard solution was
prepared from a pure (99.9%) hafnium bar and 1000 mg L−1 Ce stan-
dard solution was prepared from 0.1N CeSO4 (Carlo Erba, Milan,

The ICP-QMS instrument was a PerkinElmer-Sciex (Concord,

Ontario, Canada) Elan DRC II. The sample introduction sys- Table 3
ICP-AES operating parameters.
tem consisted of a PFA-ST microconcentric pneumatic nebulizer
(Elemental Scientific, Omaha, NE, USA), jointed to a programmable- Parameter Value
temperature teflon cyclonic spray chamber (Isomist by Glass RF Power 1200 W
Expansion, Melbourne, Australia), set at T = 0 ◦ C. An alumina 1.8- Plasma gas flow rate 15 L min−1
mm ID injector and Pt skimmer and sampler cones were used. The Auxiliary gas flow rate 1.5 L min−1
Nebulizer gas flow ratea 0.8 L min−1
main operating conditions are listed in Table 1.
Carrier gas flow rateb 0.75 L min−1
The ICP-SFMS system was a Thermo Finnigan MAT (Bremen, Sample uptake rate 0.75 mL min−1 a
Germany) Element-2. The sample was introduced by a quartz con- 1.0 mL min−1 b
centric pneumatic nebulizer (Thermo Fisher Scientific, Waltham, Internal standard uptake rate 0.25 mL min−1
MA, USA) jointed to a Scott-type double pass spray chamber Integration time 15 s
Replicates 7
(Thermo Fisher Scientific). A quartz 1.75-mm ID injector and Ni Heating temperatureb 140 ◦ C
skimmer and sampler cones were used. The main operating condi- Cooling temperatureb 2 ◦C
tions are listed in Table 2. The ion sampling depth and the ion lens
Selected wavelengths (nm)
setting were optimized daily for maximum sensitivity. The high- Analytes Ce 413.380; Ce 418.659; Ce 422.260b ; Dy
resolution mode (m/m ≈ 10,000) was applied in order to reduce 344.557b ; Dy 353.171; Er 369.265; Eu
spectral interferences. 420.504; Gd 343.999b ; Gd 376.840; La
The ICP-AES instrument was an axially viewed Varian (Spring- 333.749a ; La 379.082a ; La 398.852a ; La
408.671; Nd 378.425; Nd 406.108; Nd
vale, Australia) Vista PRO. Two sample introduction systems
430.357b ; Sm 388.528b ; Sm 432.902; Y
were employed: (i) a glass concentric K-style pneumatic neb- 360.074; Y 360.192; Y 361.104; Y 371.029;
ulizer (Varian) jointed to a glass cyclonic spray chamber and Y 488.368; Yb 328.937
(ii) an U-5000 AT (CETAC, Omaha, NB, USA) ultrasonic nebu- Internal standard Lu 291.139
lizer, built up with an air-cooled piezoelectric transducer, an a
For pneumatic nebulization only.
b
aerosol chamber, a temperature-controlled heated U-tube and a For ultrasonic nebulization only.
20 F. Ardini et al. / Analytica Chimica Acta 678 (2010) 18–25

Italy). In addition, the following 1000 mg L−1 single-element stan-

dard solutions were purchased: Lu, Ba and Sb from Sigma–Aldrich,
Y, Cs and Sn from Merck, Ge and Te from BDH. A 10,000 mg L−1 Si
standard solution was purchased from Sigma–Aldrich. Calibration
standards were prepared from a 10 mg L−1 multi-element REE stan-
dard solution supplied by PerkinElmer-Sciex (Concord, Ontario,
Canada). Daily standards were prepared by serial dilution with 1%
(v/v) nitric acid solution. Standard solutions for daily-performance
check and tuning for ICP-QMS were supplied by PerkinElmer-Sciex
and Merck.
Calibration standards for ICP-SFMS were prepared from a
50 mg L−1 multi-element REE standard solution provided by Spec-
troscan (Drøbak, Norway). Internal standards were prepared by
dilution of 1000 mg L−1 single-element standards of Ge (Merck),
In and Re (Sigma–Aldrich).

2.3. Real samples and certified reference materials

Marine sediment samples were collected at MIS (McMurdo Ice

Shelf), near the actual margin of the Ross Ice Shelf (Ross Sea, Antarc-
tica), in the framework of the international project ANDRILL [24,25].
The CRMs were: BCR-667 (estuarine sediment) and ERM-CC690
(calcareous soil) from the Institute for Reference Materials and
Measurements (Belgium), MURST-ISS-A1 (Antarctic marine sedi-
ment) from the Istituto Superiore di Sanità (Italy) and GBW 07313
(marine sediment) from the National Research Centre for Certified
Reference Materials (China).
Fig. 1. REEs determination in GBW 07313 by ICP-SFMS using two sample dissolution
procedures. (a) Light REEs; (b) heavy REEs.
2.4. Analytical procedure

Antarctic sediment samples were freeze-dried, sieved (clasts

greater than 2 mm were manually removed) and milled (agate ball
mill, from 15 to 60 min, depending on the sample’s hardness). A por-
tion (∼150 mg, weighed with a precision of ±0.1 mg) of the sample
was transferred to a microwave digestion vessel and then 3 mL of
37% hydrochloric acid, 1 mL of 65% nitric acid and 0.15 mL of 49%
hydrofluoric acid were added. The samples were digested using the
microwave digestion system MDS-2000 (CEM, Matthews, NC, USA)
by heating for 50 min at 175 ◦ C at a maximum pressure of 150 psi.
Then 1.05 mL of H3 BO3 saturated solution (corresponding to seven
times the HF volume) was added to neutralize the hydrofluoric
acid and the samples were digested again with the microwave
system by heating for 20 min at 175 ◦ C at a maximum pressure
of 100 psi. The digests were transferred to graduated flasks and
diluted to 15 mL with Milli-Q water. A similar procedure (except
sieving and milling) was also applied for CRMs and procedural
blanks. The so-obtained sample solutions were splitted in three
aliquots: (i) 4 mL of solution was analysed by ICP-AES, using on-line
internal standardization (4 ␮g mL−1 Lu standard solution, deliv-
ered at 0.25 mL min−1 ); (ii) 1 mL of solution was further diluted
to 10 mL with 1% HNO3 and analysed by ICP-QMS, using on-line
internal standardization (50 ng mL−1 Ru standard solution, deliv-
ered at 35 ␮L min−1 ) and (iii) 1.3 mL of solution was diluted to 13 mL
with 1% HNO3 and analysed by ICP-SFMS, adding 100 ␮L of internal
standard solution (60 ␮g L−1 Ge for Y, 12 ␮g L−1 In for La–Gd and
30 ␮g L−1 Re for Tb–Lu).
For the evaluation of the sample dissolution procedure, the cer-
tified reference sample GBW 07313 was also treated according to
the following procedure: an aliquot of ∼100 mg of the material was
placed into a clean 15-mL PFA Savillex vial; 2 mL of HNO3 , 3 mL of
HF and 3 mL of HClO4 were added. The sealed vials were heated
at 140 ◦ C for 24 h. Therefore, samples were evaporated to incipient
dryness. In order to remove HF, 1–3 mL of HNO3 was added and
evaporated three times. The residue was made up to 100 mL and
sonicated for 30 min to facilitate solubilization.
3. Results and discussion
3.1. Preliminary evaluation of the sample dissolution procedure
For sample dissolution, a microwave-assisted acid digestion
procedure was applied, using a HNO3 , HCl and HF mixture. In order
to provide confidence in this procedure, it was first applied to
the certified reference sediment GBW 07313 and the results were
compared with those previously obtained by another sample disso-
lution method (procedure 2), which uses HClO4 , a greater amount
of HF and much longer dissolution times. ICP-SFMS was used for
the determinations of REEs. The results are shown in Fig. 1. It can
be seen that the procedures provided comparable results. In addi-
tion, the found values were in good agreement with the certified
ones.
3.2. Development of the instrumental methods
3.2.1. ICP-QMS
3.2.1.1. Optimization of the spray chamber temperature. Direct
determination of REEs in digests from soil and sediment samples
is affected by severe polyatomic interferences, mainly due to the
oxide and hydroxide ions of light REEs (La, Ce, Pr and Nd) in the
determination of heavy REEs (Gd, Tb and Dy) and to BaO and BaOH
ions, which overlap both the Eu isotopes [26]. In order to miti-
gate these spectral interferences, the use of a cooled spray chamber
was investigated. In fact, cooling the spray chamber was reported
to reduce the oxide formation [21,23,27–29], thus providing bet-
ter signal-to-noise ratios [28,30], stability [30] and detection limits
[28] compared to standard spray chamber systems. However, the
efficiency of this approach for REEs determination in geological
samples has never been demonstrated.
Dependence of the oxide ratio on the temperature of the spray
chamber was studied for La, Ce, Pr, Nd and Ba, between 20 ◦ C and
 F. Ardini et al. / Analytica Chimica Acta 678 (2010) 18–25 21

3.2.1.2. Spectral interferences. The REEs inter-crossed interferences

were evaluated by measuring the emission intensities due to
5 ␮g L−1 single standard solutions of the analyte and of the inter-
fering element, at the analyte m/z ratio. Possible interferences due
to other elements present in extracts from sediment samples were
also considered (e.g. Ba, which is present in sediments as a major
element). In addition, we evaluated the overall contribution of the
elements with mass between 102 and 133 (Ru, Rh, Pd, Ag, Cd, In,
Sn, Sb, Te, Cs and Ba), at a concentration similar to their abundance
in soils and sediments, and in the presence of inorganic acids (1%
HNO3 , 2% HCl, 0.1% HF). The results are summarized in Table 5.
For each analyte (A), the potential interfering species (X), the per-
centage contribution of the interfering signal to the analytical one
(%IX /IA ) and the resulting interference error for the analysis of the
certified reference sediment BCR-667 (E%, calculated as above) are
reported.
Fig. 2. 144 Nd16 O+ /144 Nd+ percentage ratio as a function of the spray chamber tem- The contribution of oxide ions on the elements at m/z +16 was
perature. generally low (<3%). However, there were few but important excep-
tions. Most isotopes of gadolinium were considerably interfered
−2 ◦ C. A representative result is shown in Fig. 2 for 144 Nd. It can by oxide ions of Ba, La and Ce, with errors up to 37%. Besides,
be seen that the oxide ratio decreased by decreasing the tempera- other important spectral overlaps were due to 144 Nd16 O+ on 160 Dy+ ,
148 Nd16 O+ on 164 Er+ and 160 Gd16 O+ on 176 Lu+ . Isobaric interfer-
ture until about 0 ◦ C, then it remained almost unchanged down to
−2 ◦ C. The pattern for the oxides of other REEs was quite similar. ences between elements having the same mass (i.e. 142 Ce+ /142 Nd+
In order to minimize the risk of condensation (freezing temper- or 144 Nd+ /144 Sm+ ) can be easily compensated for by applying alge-
ature for a 10% (v/v) nitric acid solution is about −3 ◦ C), T = 0 ◦ C braic correction based on the relative abundance of the elements.
was selected as the optimal value. In this condition, the oxide ratio Finally, no significant interferences from the elements with mass
nearly halved compared to T = 20 ◦ C. A comparison between the between 102 and 133 were observed.
oxide ions formation and the resulting interference errors observed Data of Table 5 also allowed an easy selection of the analytical
at 20 ◦ C and 0 ◦ C is shown in Table 4. For each interfering oxide ions for which a low interfering error (<5%) was expected. It can be
(XO), the interfered analyte (A), the percentage ratio of oxide for- seen that each of the REEs has at least one isotope free from poly-
mation (%XO/X) and the resulting interference error for the analysis atomic interference, allowing their direct determination in digests
of the certified reference sediment BCR-667 (E%) are reported. from sediment samples without the use of correction equations.
The latter was calculated by multiplying the interferent-to-analyte Moreover, for most multi-isotopic elements different isotopes can
signal ratio (%IXO /IA ) (evaluated for CX = CA ) with the actual be selected, making the isotope dilution method possible.
interferent-to-analyte concentration ratio (CX /CA ) in the consid-
ered certified sediment sample (CX /CA for the other CRMs are quite 3.2.1.3. Accuracy assessment. The accuracy and the precision of the
similar). analytical procedure were evaluated by replicated analyses of the
It can be seen that the decrease in the oxide ratios due to the certified reference materials BCR-667 (estuarine sediment), GBW
cooled spray chamber caused a significant mitigation of the inter- 07313 (marine sediment) and ERM-CC690 (calcareous soil). The
ference errors, which resulted lower than 5% except for 139 La16 O+ certified reference sediment MURST-ISS A1, with indicative values
on 155 Gd+ and 140 Ce16 O+ on 156 Gd+ . However, gadolinium can be of REEs concentration, was also considered. The results are sum-
determined at m/z 157 (15.65% natural abundance), where the con- marized in Table 6. A satisfactory agreement between the found
tribution of 141 Pr16 O+ at T = 0 ◦ C is low (<5%). values and the certified ones may be generally observed. The only
relevant bias was observed for some heavy REEs (Tb–Lu) in BCR-
667. Since the found values were not significantly different from
Table 4 those obtained by the other instrumental techniques (see below),
Effect of the spray chamber temperature on the oxide ion formation and the resulting this result could be related to incomplete dissolution of this sample.
interference errors for the analysis of the certified reference sediment BCR-667.a
For all the other determinations, the found values were not signif-
XO A 20 ◦ C 0 ◦C icantly different from the certified ones (t-test at 95% confidence
XO/X (%) E% XO/X (%) E%
level), thereby confirming the efficiency of the developed method
134
to remove the spectral interferences and provide accurate results.
Ba16 O+ 150
Sm+ 0.06 2 0.04 1
135 Precision of the procedure generally ranged from 2% to 10% (n = 4),
Ba16 O+ 151
Eu+ 0.05 3 0.02 1
136
Ba16 O+ 152
Sm+ 0.06 1 0.03 1 depending on the analyte and without any difference among the
137
Ba16 O+ 153
Eu+ 0.06 5 0.02 2 four matrices analysed.
138
Ba16 O+ 154
Sm+ 0.04 9 0.02 4
138
Ba16 O1 H+ 155
Gd+ 0.02 8 0.01 5
139
3.2.2. ICP-SFMS
La16 O+ 155
Gd+ 1.48 47 0.81 24
140
Ce16 O+ 156
Gd+ 1.41 76 0.80 38 Although the resolving power (R) required for the separation of
142
Ce16 O+ 158
Gd+ 1.49 8 0.84 4 the interfering oxide ions from lanthanides is typically between
141
Pr16 O+ 157
Gd+ 1.09 9 0.62 4 7000 and 10,000 [31], many authors [14,15,32,33] performed
142
Nd16 O+ 158
Gd+ 0.98 6 0.55 3 measurements of REEs in geological materials at low resolution
143
Nd16 O+ 159
Tb+ 1.01 4 0.56 3
144 (R = 300), in order to achieve maximum sensitivity. This implied,
Nd16 O+ 160
Gd+ 0.99 6 0.53 3
145
Nd16 O+ 161
Dy+ 1.00 2 0.58 1 however, that for some isotopes mathematical correction had to
146
Nd16 O+ 162
Dy+ 1.01 4 0.53 2 be applied. In this study, the high-resolution mode (R = 10,000) was
a
A, analyte; X, potential interfering element; XO, potential interfering oxide; selected for ICP-SFMS analyses, since it provided adequate sensi-
XO/X (%), percentage oxide formation ratio of the interferent; E%, interference error tivity for the considered analytical task and avoided the need for
for the analysis of the certified reference sediment BCR-667. algebraic correction. The precision and the accuracy were verified
22 F. Ardini et al. / Analytica Chimica Acta 678 (2010) 18–25

Table 5
Spectral interferences in the determination of REEs by ICP-QMS.a

A X IX /IA (%) E% A X IX /IA (%) E%

89 + 73 16 + 159 + 143 16 +
Y Ge O 0.01 <0.01 Tb Nd O 0.06 2.1
139
La+ 123
Sb16 O+ 0.01 <0.01 160
Dy+ 160
Gd+ 860 940
140
Ce+ 124
Sn16 O+ 0.01 <0.01 144
Nd16 O+ 3.8 24
124
Te16 O+ 0.01 <0.01 144
Sm16 O+ 0.47 0.55
142
Ce+ 142
Nd+ 240 110 161
Dy+ 145
Nd16 O+ 0.20 1.3
126
Te16 O+ 0.04 <0.01 162
Dy+ 146
Nd16 O+ 0.27 1.7
141
Pr+ 125
Te16 O+ <0.01 <0.01 163
Dy+ 147
Sm16 O+ 0.05 0.06
142
Nd+ 142
Ce+ 41 94 164
Dy+ 164
Er+ 4.8 2.8
126
Te16 O+ 0.02 <0.01 148
Nd16 O+ 0.09 0.56
143
Nd+ – – – 148
Sm16 O+ 0.04 0.05
144
Nd+ 144
Sm+ 13 2.5 165
Ho+ 149
Sm16 O+ 0.01 0.07
128
Te16 O+ 0.01 <0.01 164
Er+ 164
Dy+ 2100 3600
145
Nd+ – – – 148
Nd16 O+ 1.9 20
146
Nd+ 130
Te16 O+ 0.02 <0.01 148
Sm16 O+ 0.90 1.8
148
Nd+ 148
Sm+ 200 38 166
Er+ 150
Nd16 O+ 0.09 0.92
150
Nd+ 150
Sm+ 140 25 150
Sm16 O+ 0.02 0.04
134
Ba16 O+ <0.01 0.12 167
Er+ 151
Eu16 O+ 0.03 0.01
144
Sm+ 144
Nd 740 4000 168
Er+ 152
Sm16 O+ 0.09 0.18
128
Te16 O+ 0.11 0.02 170
Er+ 170
Yb+ 21 20
147
Sm+ – – – 154
Gd16 O+ 0.05 0.09
148
Sm+ 148
Nd+ 49 260 154
Sm16 O+ 0.13 0.25
149
Sm+ 133
Cs16 O+ 0.02 0.04 169
Tm+ 153
Eu16 O+ <0.01 0.02
150
Sm+ 150
Nd+ 73 400 170
Yb+ 170
Er+ 480 510
134
Ba16 O+ <0.01 0.46 154
Gd16 O+ 0.23 0.46
152
Sm+ 136
Ba16 O+ <0.01 0.54 154
Sm16 O+ 0.60 1.3
154
Sm+ 154
Gd+ 8.8 8.3 171
Yb+ 155
Gd16 O+ 0.24 0.48
138
Ba16 O+ 0.03 3.1 172
Yb+ 156
Gd16 O+ 0.23 0.46
151
Eu+ 135
Ba16 O+ <0.01 0.85 173
Yb+ 157
Gd16 O+ 0.23 0.47
153
Eu+ 137
Ba16 O+ <0.01 1.7 174
Yb+ 158
Gd16 O+ 0.18 0.36
154
Gd+ 154
Sm+ 1140 1200 176
Yb+ 176
Lu+ 20 3.0
138
Ba16 O+ 0.33 37 176
Hf+ 38 100
155
Gd+ 139
La16 O+ 3.1 19 160
Gd16 O+ 0.40 0.80
138
Ba16 O1 H+ 0.04 4.7 160
Dy16 O+ 0.03 0.05
156
Gd+ 140
Ce16 O+ 2.6 34 175
Lu+ 159
Tb16 O+ 0.21 0.44
157
Gd+ 141
Pr16 O+ 3.2 4.4 176
Lu+ 176
Yb+ 490 3300
158
Gd+ 142
Ce16 O+ 0.28 3.6 176
Hf+ 190 3400
142
Nd16 O+ 0.45 2.5 160
Gd16 O+ 2.0 27
160
Gd+ 160
Dy+ 12 11 160
Dy16 O+ 0.14 1.7
144
Nd16 O+ 0.44 2.5
144
Sm16 O+ 0.05 0.06
a
A, analyte; X, potential interfering species; IX /IA (%), percentage contribution of the interfering signal to the analytical one; E%, interference error for the analysis of the
certified reference sediment BCR-667.

by using the four certified reference materials previously analysed
by ICP-QMS. Also in this case a satisfactory agreement between the
found values and the certified ones was generally observed.
3.2.3. ICP-AES
Because of the complexity of the REEs emission spectra and of
severe interferences from major elements, the determination of
REEs in geological samples by ICP-AES usually requires their separa-
tion from the matrix prior to the instrumental analysis. In contrast,
we explored the possibility of directly measuring the REEs concen-
tration in digests from sediment samples by a careful selection of
the emission line wavelengths (to minimize spectral interferences)
and the application of on-line internal standardization (to com-
pensate for non-spectral matrix effects). Moreover, both pneumatic
(PN) and ultrasonic (USN) nebulizations were considered.
It was found that REEs except Pr, Tb, Ho, Tm and Lu can be accu-
rately measured in digests from sediment samples, using both PN
and USN, as demonstrated by the analysis of the already mentioned
CRMs. As expected USN provided 6–7-fold higher sensitivity than
PN.
3.3. Comparison of the analytical figures
3.3.1. Precision
Precision of the procedures was estimated by considering the
results from duplicate analyses of 15 marine sediment samples
collected in Antarctica and by calculating the pooled standard
deviation values (Table 7). In order to establish the statistically
significant differences in precision, F-values were also computed.
RSD% values, which comprise precision of the sample preparation
method and of the instrumental determination, were 4–8%, 3–6%
and 4–11% for ICP-QMS, ICP-SFMS and ICP-AES, respectively. F-test
showed that, with few exceptions (highlighted in bold), precisions
of the considered techniques were not significantly different, at 95%
probability level.
3.3.2. Limits of detection and quantification
The instrumental detection limits and the quantification lim-
its of the procedures are reported in Table 8, together with the
concentration ranges of REEs in the analysed samples. The instru-
mental detection limits were calculated as three times the standard
deviation of ten procedural blanks. The quantification limits were
computed considering ten times the standard deviation of the
blanks and by taking into account the dilution due to the sam-
ple preparation procedure (100-fold for ICP-AES and 1000-fold for
ICP-MS, see Section 2).
ICP-QMS and ICP-SFMS detection and quantification limits were
generally comparable, except for some light REEs (Y, La, Ce and
Nd), whose ICP-SFMS values were about one order of magnitude
lower than those obtained by ICP-QMS. As expected, ICP-AES detec-
tion limits were 1–2 orders of magnitude higher than ICP-MS,
with ultrasonic nebulization providing 3–8-fold better values than
 F. Ardini et al. / Analytica Chimica Acta 678 (2010) 18–25 23

Table 6
Analysis of certified reference materials by ICP-QMS (values in ␮g g−1 ).

GBW 07313 (marine sediment) MURST-ISS A1 (marine sediment)

a b
Analyte Found Certified Analyte Founda Certifiedb , c

Y 103 ± 3 104 ± 19 Y 16.9 ± 0.3 19.1 ± 2.5

La 64.6 ± 2.0 67.8 ± 11.5 La 35.1 ± 1.1 30.2–40.4
Ce 90 ± 3 92 ± 29 Ce 73.9 ± 2.9 70.76–86.2
Pr 19.5 ± 1.1 20.1 ± 3.6 Pr 8.4 ± 0.4 –
Nd 82.3 ± 4.9 91.8 ± 11.0 Nd 31.1 ± 1.6 33.1 ± 1.1
Sm 20.5 ± 1.4 21.5 ± 4.1 Sm 6.04 ± 0.33 4.61–5.9
Eu 5.0 ± 0.3 5.3 ± 1.0 Eu 1.33 ± 0.06 1.0–1.94
Gd 21 ± 1 22 ± 3 Gd 5.08 ± 0.23 –
Tb 3.2 ± 0.2 3.4 ± 1.0 Tb 0.70 ± 0.03 0.72 ± 0.06
Dy 19.5 ± 1.4 19.9 ± 5.2 Dy 3.83 ± 0.22 2.79–3.8
Ho 3.8 ± 0.3 4.3 ± 0.4 Ho 0.67 ± 0.03 –
Er 11 ± 1 11 ± 2 Er 1.79 ± 0.10 –
Tm 1.45 ± 0.11 1.54 ± 0.35 Tm 0.223 ± 0.010 –
Yb 10.1 ± 0.3 9.8 ± 3.2 Yb 1.47 ± 0.03 2.2–2.86
Lu 1.39 ± 0.11 1.46 ± 0.57 Lu 0.188 ± 0.008 –

BCR-667 (estuarine sediment) ERM-CC690 (calcareous soil)

Analyte Founda Certifiedc , d Analyte Founda Certifiedd

Y 16.3 ± 0.8 16.7–25.3 Y 14.4 ± 0.6 –

La 25.4 ± 1.9 27.8 ± 1.0 La 24.0 ± 1.2 24.4 ± 1.7
Ce 52.9 ± 3.9 56.7 ± 2.5 Ce 49.7 ± 3.1 49.1 ± 2.5
Pr 6.1 ± 0.4 6.1 ± 0.5 Pr 5.1 ± 0.3 –
Nd 23.2 ± 1.6 25.0 ± 1.4 Nd 18.9 ± 1.1 19.1 ± 2.2
Sm 4.71 ± 0.35 4.66 ± 0.20 Sm 3.7 ± 0.2 3.5 ± 0.4
Eu 0.95 ± 0.06 1.00 ± 0.05 Eu 0.71 ± 0.06 –
Gd 4.15 ± 0.31 4.41 ± 0.12 Gd 3.2 ± 0.2 3.2 ± 0.4
Tb 0.578 ± 0.037 0.682 ± 0.017 Tb 0.44 ± 0.04 0.50 ± 0.07
Dy 3.28 ± 0.21 4.01 ± 0.14 Dy 2.60 ± 0.19 2.90 ± 0.28
Ho 0.60 ± 0.03 0.80 ± 0.06 Ho 0.48 ± 0.05 –
Er 1.65 ± 0.09 2.35 ± 0.15 Er 1.43 ± 0.12 –
Tm 0.205 ± 0.014 0.326 ± 0.025 Tm 0.175 ± 0.029 0.232 ± 0.026
Yb 1.41 ± 0.06 2.20 ± 0.09 Yb 1.39 ± 0.07 1.57 ± 0.19
Lu 0.177 ± 0.014 0.325 ± 0.020 Lu 0.161 ± 0.029 –
a
Mean value ± 95% confidence interval (n = 4).
b
Mean value ± standard deviation.
c
Indicative values in italic.
d
Mean value ± 95% confidence interval.

pneumatic nebulization. Concerning the quantification limits, it is
important to note that the sample dilution applied for ICP-AES anal-
yses was 10-fold lower than that required for ICP-MS. Hence, the
difference in performances among the ICP-AES and ICP-MS tech-
niques with regard to LOQs was lower than for LODs.
The achieved ICP-QMS detection and quantification limits are
comparable to those reported for the direct analysis of geologi-
cal samples using the same instrumental technique [7]. Similarly,
ICP-SFMS quantification limits are comparable to literature data
[15,32], despite the lower sensitivity of the high-resolution mode.
Finally, ICP-AES detection and quantification limits are in good
agreement (with the exception of cerium) with results obtained
by cation-exchange separation procedures [6,34].
In addition, the achieved quantification limits were significantly
lower than the REEs concentrations in sediment samples, thus mak-
ing all the developed methods adequate for the analytical task.

Table 7
Precision of the procedures.

Analyte RSD% F-valuesa

ICP-QMS ICP-SFMS ICP-AES ICP-QMS/ICP-SFMS ICP-QMS/ICP-AES ICP-SFMS/ICP-AES

Y 3.9 3.1 4.5 1.851 1.274 2.358

La 4.9 3.6 4.6 2.061 1.161 1.776
Ce 4.9 4.1 4.8 1.504 1.008 1.492
Pr 5.0 4.6 – 1.375 – –
Nd 5.0 4.9 4.9 1.174 1.207 1.029
Sm 5.0 4.1 4.8 1.705 1.485 1.148
Eu 5.2 4.1 4.5 2.034 1.424 1.429
Gd 4.8 4.1 5.6 1.989 1.077 2.143
Tb 5.5 4.2 – 2.207 – –
Dy 5.3 4.6 4.5 1.707 1.676 1.019
Ho 5.6 5.7 – 1.300 – –
Er 5.7 4.8 10.8 1.927 2.033 3.917
Tm 7.2 4.1 – 3.469 – –
Yb 6.5 5.3 4.2 2.119 1.717 1.234
Lu 7.5 5.0 – 2.451 – –
a
Values greater than the tabulated F-value at the 95% probability level (F = 2.40, n = 15) are highlighted in bold.
24 F. Ardini et al. / Analytica Chimica Acta 678 (2010) 18–25

Table 8
Instrumental limits of detection and limits of quantification of the procedures.

Analyte Instrumental LOD (␮g L−1 )a LOQ (␮g g−1 )b Conc range in sedimentsc (␮g g−1 )

ICP-QMS ICP-SFMS ICP-AES ICP-QMS ICP-SFMS ICP-AES

USN PN USN PN

Y 0.068 0.007 0.04 0.35 0.23 0.02 0.01 0.12 13.7–103

La 0.080 0.010 0.65 2.14 0.27 0.04 0.22 0.71 22.9–65.3
Ce 0.234 0.008 3.03 10.4 0.78 0.03 1.01 3.48 45.7–90
Pr 0.017 0.006 – – 0.06 0.02 – – 4.9–19.5
Nd 0.066 0.014 1.87 6.76 0.22 0.05 0.62 2.25 17.0–82.3
Sm 0.013 0.014 0.73 4.96 0.04 0.05 0.24 1.65 4.29–20.5
Eu 0.003 0.011 0.17 0.59 0.01 0.04 0.06 0.20 0.65–5.0
Gd 0.013 0.016 1.96 5.65 0.04 0.05 0.65 1.88 2.9–21
Tb 0.002 0.006 – – 0.01 0.02 – – 0.41–3.2
Dy 0.012 0.010 0.26 1.73 0.04 0.03 0.09 0.58 2.19–20.2
Ho 0.002 0.007 – – 0.01 0.02 – – 0.45–3.8
Er 0.008 0.010 0.24 1.30 0.03 0.03 0.08 0.43 1.18–11
Tm 0.002 0.007 – – 0.01 0.02 – – 0.160–1.45
Yb 0.022 0.009 0.07 0.27 0.07 0.03 0.02 0.09 1.18–10.1
Lu 0.001 0.005 – – 0.01 0.02 – – 0.147–1.39
a
3 Blank ; n = 10 (ICP-MS and PN-ICP-AES), n = 6 (USN-ICP-AES).
b
10 Blank , sample dilution: 100 for ICP-AES (0.15 g in 15 ml) and 1000 for ICP-MS (further 10-fold dilution).
c
Concentration ranges in the analysed certified reference materials and Antarctic marine sediment samples.

3.3.3. Analytical results
The developed procedures were applied to 15 marine sediment
samples collected in Antarctica in the framework of the ANDRILL
project [24,25]. The project aims to recover a 1200 m glacioma-
rine sediment core in the northwest corner of the Ross Ice Shelf,
the McMurdo Ice Shelf (MIS) sector, east of Hut Point Peninsula,
Ross Island. This site is located at a critical juncture between glacial
fluxes coming from the Transantarctic Mountains (TAM) and the
McMurdo Volcanic Group (MVG). Sedimentary records will be used
in reconstructing climatic history of the East-Antarctic Sheet by
evaluating retreats and advances of the Ross Ice Shelf in the differ-
ent climatic periods along the last 13 Ma [35].
Samples were analysed in duplicate and pneumatic nebulization
was applied for ICP-AES analyses. Results for three representative
samples are shown in Table 9. These samples were chosen in order
to enlighten the reliability of the ICP methods in analysing different
sedimentary materials: sample at 40.375 m depth is a diamictite
with large content in volcanic and other terrigenous materials
originated in the MVG area; sample at 59.015 m depth contain
terrigenous materials from TAM area; sample at 83.275 m depth
is a diamictite characterized by large inclusions of volcanic ashes
and biogenic deposits (terrigenous-marine mixed sediments). The
chemical composition of the three samples changes according to
their different origins.
Analytical results obtained by applying the three ICP techniques
to all the samples gave a satisfactory agreement, as confirmed by
the application of the Student t-test at 95% confidence level. The
only relevant differences were related to Er and Gd determinations
by ICP-QMS, which provided 15–25% higher concentration values
than ICP-SFMS, and to Yb determination by ICP-AES, which fur-
nished 20–60% higher values than ICP-MS techniques. However,
these discrepancies were not observed for the analyses of CRMs.
In order to show an example of the climatic implication of REEs
determination in glaciomarine sedimentary records, we report the
lanthanide pattern for the samples 40.375 m and 59.015 m, charac-
terized by terrigenous content of different origins (MVG and TAM
areas, respectively). REEs patterns (Fig. 3) were built by plotting the
logarithm of their relative abundances (normalized to chondritic
values [36]) as a function of atomic number. It can be seen that
the Eu dip (named “Eu anomaly”) is more pronounced for the sam-
ple with a larger content of terrigenous materials from TAM area.
This difference was observed for the other analysed core samples,
too. So, the study of the depth profile of the europium anomaly can
contribute to reconstruct the changes in transport and depositional

Table 9
REEs determination in marine sediment samples from Antarctica.a

Analyte ANDRILL-MIS 40.375 m ANDRILL-MIS 59.015 m ANDRILL-MIS 83.275 m

ICP-QMS ICP-SFMS ICP-AES ICP-QMS ICP-SFMS ICP-AES ICP-QMS ICP-SFMS ICP-AES

Y 24.8 ± 1.0 23.2 ± 0.8 23.8 ± 1.2 26.4 ± 1.0 23.5 ± 0.8 25.6 ± 1.2 30.6 ± 1.0 29.6 ± 0.8 30.6 ± 1.2
La 44.5 ± 2.3 44.7 ± 1.6 44.1 ± 2.2 47.2 ± 2.3 43.9 ± 1.6 46.1 ± 2.2 65.3 ± 2.3 65.2 ± 1.6 64.7 ± 2.2
Ce 92 ± 5 92 ± 4 93 ± 5 98 ± 5 92 ± 4 98 ± 5 132 ± 5 126 ± 4 136 ± 5
Pr 10.4 ± 0.6 10.4 ± 0.5 – 11.0 ± 0.6 10.0 ± 0.5 – 14.6 ± 0.6 14.3 ± 0.5 –
Nd 39.4 ± 2.1 39.3 ± 1.9 37.5 ± 1.9 40.8 ± 2.1 37.2 ± 1.9 38.0 ± 1.9 52.7 ± 2.1 52.5 ± 1.9 49.6 ± 1.9
Sm 7.5 ± 0.4 7.3 ± 0.3 6.6 ± 0.3 7.8 ± 0.4 6.9 ± 0.3 6.8 ± 0.3 9.6 ± 0.4 9.2 ± 0.3 8.4 ± 0.3
Eu 1.9 ± 0.1 1.9 ± 0.1 2.0 ± 0.1 1.8 ± 0.1 1.6 ± 0.1 1.7 ± 0.1 2.2 ± 0.1 2.0 ± 0.1 2.1 ± 0.1
Gd 6.6 ± 0.3 5.7 ± 0.2 6.1 ± 0.3 6.8 ± 0.3 5.8 ± 0.2 6.0 ± 0.3 8.2 ± 0.3 7.0 ± 0.2 7.1 ± 0.3
Tb 0.90 ± 0.05 0.85 ± 0.04 – 0.96 ± 0.05 0.86 ± 0.04 – 1.14 ± 0.05 1.09 ± 0.04 –
Dy 5.2 ± 0.3 4.8 ± 0.2 4.8 ± 0.2 5.6 ± 0.3 4.9 ± 0.2 5.0 ± 0.2 6.6 ± 0.3 6.3 ± 0.2 6.5 ± 0.2
Ho 0.94 ± 0.06 0.87 ± 0.05 – 1.04 ± 0.06 0.92 ± 0.05 – 1.22 ± 0.06 1.10 ± 0.05 –
Er 2.6 ± 0.2 2.3 ± 0.1 1.9 ± 0.2 2.9 ± 0.2 2.5 ± 0.1 2.2 ± 0.2 3.5 ± 0.2 3.0 ± 0.1 2.6 ± 0.2
Tm 0.34 ± 0.03 0.34 ± 0.02 – 0.39 ± 0.03 0.36 ± 0.02 – 0.46 ± 0.03 0.45 ± 0.02 –
Yb 2.4 ± 0.2 2.0 ± 0.1 2.9 ± 0.1 2.6 ± 0.2 2.2 ± 0.1 3.0 ± 0.1 2.8 ± 0.2 2.6 ± 0.1 3.7 ± 0.1
Lu 0.30 ± 0.03 0.30 ± 0.02 – 0.35 ± 0.03 0.33 ± 0.02 – 0.42 ± 0.03 0.43 ± 0.02 –
a
Samples were obtained from the marine sediment core collected at MIS (McMurdo Ice Shelf, Ross Sea, Antarctica) in the framework of the international project ANDRILL
[24,25].
Values are expressed in ␮g g−1 as mean ± pooled standard deviation (n = 15).
 F. Ardini et al. / Analytica Chimica Acta 678 (2010) 18–25
Fig. 3. Chondrite normalized REEs concentration for the samples 40.375 m (dashed
line) and 59.015 m (unbroken line), characterized by terrigenous content of different
origins (MVG and TAM areas, respectively).
mechanisms of terrigenous material stratified in the MIS site due
to variations in ice sheet extension, grounding line position and
main ice-stream flow directions owing to changes of the climatic
conditions in the McMurdo Sound region.
4. Conclusions
The use of a cooled spray chamber proved to significantly
decrease the oxide ions formation (about 50%), with consequent
mitigation of the spectral interferences that affect REEs determina-
tion by ICP-QMS. The developed method was precise and accurate,
as demonstrated by replicated analysis of both CRMs and real sam-
ples. ICP-SFMS working in high-resolution mode also provided
accurate results, with similar precision to ICP-QMS and compara-
ble or better (for light REEs) limits of quantification (18–52 ng g−1
and 10–780 ng g−1 for sector field- and quadrupole-ICP-MS, respec-
tively). Direct determination of most REEs in digests from sediment
samples was also achieved by ICP-AES equipped with a pneumatic
or ultrasonic nebulization system, after a careful selection of the
emission lines and compensation for non-spectral interferences by
internal standardization. On the other hand, Pr, Tb, Ho, Tm and Lu
could not be measured by ICP-AES, due to strong spectral interfer-
ences and/or insufficient sensitivity, even using USN. The reliability
of the developed analytical methods was tested by the analysis
of sedimentary materials having different chemical compositions,
25
according to their origin. The good agreement among the ICP tech-
niques observed for all the samples and the obtaining of consistent
REEs patterns of abundance make the proposed analytical proce-
dures suitable for geological and environmental studies.
References
[1] S.M. McLennan, in: B.R. Lipin, G.A. McKay (Eds.), Geochemistry and Mineralogy
of the Rare Earths, Mineral. Soc. Am., Washington, 1989, pp. 169–200.
[2] I. Olmez, E.R. Sholkovitz, D. Hermann, R.P. Eganhouse, Environ. Sci. Technol. 25
(1991) 310–316.
[3] A.M. Rusu, P.D.J. Chimonides, G.C. Jones, R. Garcia-Sanchez, O.W. Purvis, Envi-
ron. Sci. Technol. 40 (2006) 4599–4604.
[4] E. Orvini, M. Speziali, A. Salvini, C. Herborg, Microchem. J. 67 (2000) 97–104.
[5] N. Nakamura, K. Misawa, A. Masuda, M. Kitamura, S. Watanabe, Earth Planet.
Sci. Lett. 99 (1990) 290–302.
[6] R. Djingova, J. Ivanova, Talanta 57 (2002) 821–829.
[7] K.E. Jarvis, J. Anal. Atom. Spectrom. 4 (1989) 563–570.
[8] J.N. Walsh, F. Buckley, J. Barker, Chem. Geol. 33 (1981) 141–153.
[9] F.E. Lichte, A.L. Meier, J.G. Crock, Anal. Chem. 59 (1987) 1150–1157.
[10] A.R. Date, D. Hutchison, J. Anal. Atom. Spectrom. 2 (1987) 269–276.
[11] K.E. Jarvis, A.L. Gray, E.J. McCurdy, J. Anal. Atom. Spectrom. 4 (1989) 743–747.
[12] N.M. Raut, L.S. Huang, K.C. Lin, S.K. Aggarwal, Anal. Chim. Acta 530 (2005)
91–103.
[13] M.A. Vaughan, G. Horlick, Appl. Spectrosc. 44 (1990) 587–593.
[14] J. Riondato, F. Vanhaecke, L. Moens, R. Dams, Fresenius J. Anal. Chem. 370 (2001)
44–552.
[15] I.V. Nikolaeva, S.V. Palesskii, O.A. Koz’menko, G.N. Anoshin, Geochem. Int. 46
(2008) 1016–1022.
[16] H. Tao, A. Miyazaki, J. Anal. Atom. Spectrom. 10 (1995) 1–5.
[17] H. Louie, S.Y.P. Soo, J. Anal. Atom. Spectrom. 7 (1992) 557–564.
[18] M.A. Vaughan, G. Horlick, Spectrochim. Acta, Part B 45 (1990) 1289–1299.
[19] D. Günther, H.P. Longerich, S.E. Jackson, L. Forsythe, Fresenius J. Anal. Chem.
355 (1996) 771–773.
[20] J.W.H. Lam, G. Horlick, Spectrochim. Acta, Part B 45 (1990) 1327–1338.
[21] H.P. Longerich, W. Diegor, J. Anal. Atom. Spectrom. 16 (2001) 1196–1201.
[22] R. Tsukahara, M. Kubota, Spectrochim. Acta, Part B 45 (1990) 581–589.
[23] Y. Hu, F. Vanhaecke, L. Moens, R. Dams, P. del Castilho, J. Japenga, Anal. Chim.
Acta 373 (1998) 95–105.
[24] D. Harwood, F. Florindo, F. Talarico, R. Levy, G. Kuhn, T. Naish, F. Niessen, R.
Powell, A. Pyne, G. Wilson, Eos 90 (2009) 90–91.
[25] G. Giorgetti, F. Talarico, S. Sandroni, A. Zeoli, Global Planet. Change 69 (2009)
94–102.
[26] P. Dulski, Fresenius J. Anal. Chem. 350 (1994) 194–203.
[27] R.C. Hutton, A.N. Eaton, J. Anal. Atom. Spectrom. 2 (1987) 595–598.
[28] C. Rivas, L. Ebdon, S.J. Hill, J. Anal. Atom. Spectrom. 11 (1996) 1147–1150.
[29] G. Zhu, R.F. Browner, J. Anal. Atom. Spectrom. 3 (1988) 781–789.
[30] D. Pollmann, C. Pilger, R. Hergenröder, F. Leis, P. Tschöpel, J.A.C. Broekaert,
Spectrochim. Acta, Part B 49 (1994) 683–690.
[31] J.S. Becker, H.-J. Dietze, J. Anal. Atom. Spectrom. 12 (1997) 881–889.
[32] Z. Varga, R. Katona, Z. Stefánka, M. Wallenius, K. Mayer, A. Nicholl, Talanta 80
(2010) 1744–1749.
[33] G. Bayon, J.A. Barrat, J. Etoubleau, M. Benoit, C. Bollinger, S. Révillon, Geostand.
Geoanal. Res. 33 (2009) 51–62.
[34] K.E. Jarvis, I. Jarvis, Geostand. Newslett. 12 (1988) 1–12.
[35] S.R. Taylor, S.M. McLennan, R.L. Armstrong, J. Tarney, Phil. Trans. Roy. Soc. Lond.
A 301 (1981) 381–399.
[36] S.R. Taylor, S.M. McClennan, The Continental Crust: Its Composition and Evo-
lution, Blackwell, Oxford, 1985.