You are on page 1of 40


Issues also available electronically.
(see inside front cover)

tomic pectroscopy
September/October 2003 Volume 24, No. 5

In This Issue:
Characterization of Environmental Samples in an Ophiolitic Area of Northern Italy Using ICP-OES, ICP-MS, and XRF
M. Bettinelli, G.M. Beone, C. Baffi, S. Spezia, and A. Nassisi ..................................... 159

Cloud Point Extraction Preconcentration and ICP-OES Determination of Trace Chromium and Copper in Water Samples
Jing Li, Pei Liang, Taqing Shi, and Hanbing Lu ......................................................... 169

A Rapid Method for the Determination of Selenium in Blood and Serum by ETAAS With Zeeman Background Correction
Arturo Montel, José Manuel de Pradena, José María Gervas, Guadalupe Álvarez Bustamante, and José Luis López-Colón .................................... 173

Flow Injection AAS Determination of Cd, Cu, and Pb at Trace Levels in Wine Using Solid Phase Extraction
Andréa Pires Fernandes, Mercedes de Moraes, and José Anchieta Gomes Neto .... 179

Tantalum Effects on the ICP-OES Determination of Trace Elements in Tantalum Powder
G. Anil, M.R.P. Reddy, Arbind Kumar, and T.L. Prakash .......................................... 185

Mixed Matrix Effect of Easily and Non-easily Ionizable Elements in the ICP-OES Analysis of Trace Impurities in Potassium Tantalum Fluoride
G. Anil, M.R.P. Reddy, Arbind Kumar, and T.L. Prakash .......................................... 190

ASPND7 24(5) 159–194 (2003) ISSN 0195-5373


tomic pectroscopy
Vol. 24(1), January/February 2003

Printed in the United States and published six times a year by: PerkinElmer Life and Analytical Sciences 710 Bridgeport Avenue Shelton, CT 06484-4794 USA

Editor Anneliese Lust E-mail: Technical Editors Glen R. Carnrick, AA Dennis Yates, ICP Kenneth R. Neubauer, ICP-MS SUBSCRIPTION INFORMATION Atomic Spectroscopy P.O. Box 3674 Barrington, IL 60011 USA Fax: +1 (847) 304-6865 2003 Subscription Rates • U.S. $60.00 includes third-class mail delivery worldwide; $20.00 extra for electronic file. • U.S. $80.00 includes airmail delivery; $20 extra for electronic file. • U.S. $60.00 for electronic file only. • Payment by check (drawn on U.S. bank in U.S. funds) made out to: “Atomic Spectroscopy” Electronic File • For electronic file, send request via e-mail to: Back Issues/Claims • Single back issues are available at $15.00 each. • Subscriber claims for missing back issues will be honored at no charge within 90 days of issue mailing date. Address Changes to: Atomic Spectroscopy P.O. Box 3674 Barrington, IL 60011 USA Copyright © 2003 PerkinElmer, Inc. All rights reserved. Microfilm Atomic Spectroscopy issues are available from: University Microfilms International 300 N. Zeeb Road Ann Arbor, MI 48106 USA Tel: (800) 521-0600 (within the U.S.) +1 (313) 761-4700 (internationally)

Guidelines for Authors
Atomic Spectroscopy serves as a medium for the dissemination of general information together with new applications and analytical data in atomic absorption spectrometry. The pages of Atomic Spectroscopy are open to all workers in the field of atomic spectroscopy. There is no charge for publication of a manuscript. The journal has around 3000 subscribers on a worldwide basis, and its success can be attributed to the excellent contributions of its authors as well as the technical guidance of its reviewers and the Technical Editors. The original of the manuscript should be submitted to the editor by mail plus electronic file on disk or e-mail in the following manner: 1. Mail original of text, doublespaced, plus graphics in black/white. 2. Provide text and tables in .doc file and figures in doc or tif files. 3. Number the references in the order they are cited in the text. 4. Submit original drawings or glossy photographs and figure captions. 5. Consult a current copy of Atomic Spectroscopy for format. 6. Or e-mail text and tables in doc file and graphics in doc or tif files to the editor: or All manuscripts are sent to two reviewers. If there is disagreement, a third reviewer is consulted. Minor changes in style are made in-house and submitted to the author for approval. A copyright transfer form is sent to the author for signature. If a revision of the manuscript is required before publication can be considered, the paper is returned to the author(s) with the reviewers’ comments. In the interest of speed of publication, a pdf file of the typeset text is e-mailed to the corresponding author before publication for final approval. Additional reprints can be purchased, but the request must be made at the time the manuscript is approved for publication. Anneliese Lust Editor, Atomic Spectroscopy PerkinElmer Life and Analytical Sciences 710 Bridgeport Avenue Shelton, CT 06484-4794 USA

PerkinElmer is a registered trademark of PerkinElmer, Inc. AAnalyst, Optima, GemTip, and WinLab are trademarks of PerkinElmer Life and Analytical Sciences. SCIEX and ELAN are registered trademarks of MDS Inc., a division of MDS Inc. Dell is a registered trademark of Dell Computer Corporation. Hewlett Packard and LaserJet are registered trademarks of Hewlett Packcard Corporation. Milli-Q is a trademark of Millipore Corporation. Ryton is a registered trademark of Phillips Petroleum Company. Suprapur is a registered trademark of Merck & Co. Triton is a registered trademark of Union Carbide Chemicals and Plastics Technology Corporation. Tygon is a trademark of Norton Co. Viton is a registered trademark of DuPont Dow Elastomers. Registered names and trademarks, etc. used in this publication even without specific indication therof are not to be considered unprotected by law.

Characterization of Environmental Samples in an Ophiolitic Area of Northern Italy Using ICP-OES, ICP-MS, and XRF
*M. Bettinellia, G.M. Beoneb, C. Baffib, S. Speziaa, and A. Nassisic Igiene Ambientale e Tossicologia Industriale, Fondazione "S. Maugeri," Clinica del Lavoro e della Riabilitazione I.R.C.C.S., via Ferrata, 8, 27100 Pavia, Italy b Istituto di Chimica Agraria e Ambientale, Facoltà di Agraria, Università Cattolica S. Cuore, via Emilia Parmense, 84, 29100 Piacenza, Italy c ARPA Sezione Provinciale di Piacenza, Area Analitica Specialistica Agropedologia, Loc. Gariga, 29027 Podenzano-Pc, Italy
a Laboratorio

INTRODUCTION Environmental studies require accurate and precise information on trace element distribution in the ecosystems. In this respect, partition and transfer coefficients can be extremely useful to represent the distribution and flow of elements among the various environmental compartments which are often characterized by extremely complex matrices. In environmental studies, the choice of the most suitable analytical methods and procedures by which good quality data can be obtained is of primary importance. Generally, the use of validated methods for which precision values are well known and measurement uncertainty has already been evaluated in compliance with the latest international guidelines is recommended (1,2). Validation of these methods requires accuracy tests with certified samples (3,4) or, when not available, comparison between different independent analytical techniques on real samples (5,6). Today, the characterization of environmental matrices is obtained using complete dissolution of the samples, followed by monoelement (FAAS or GFAAS) or multielement determination with spectroscopic

In Ponte Barberino, an ophiolitic environment in the Trebbia River Valley near the town of Piacenza, Northern Italy, rocks and soils are characterized by high concentrations of Cr and Ni. The determination of macro, micro, and rare earth elements in rock, soil, sediment, and plant samples was carried out by ICPOES, ICP-MS, and XRF after alkaline fusion or microwave acid mineralization. For sediment samples, microwave acid digestion (8 mL aqua regia + 2 mL HF), followed by ICP-OES, provided results comparable to those obtained by XRF. In the case of plants, microwave acid mineralization (7 mL HNO3 + 1 mL H2O2 + 0.2 mL HF) followed by ICP-OES analysis was particularly suitable for the determination of trace elements in leaves of the most common botanical species found in the investigated area. In addition, soil/plant transfer factors characterized the Alyssum spp. species as a hyperaccumulator of Ni. Alkaline fusion was found to be the best dissolution procedure for rock samples; recoveries ranged between 95 and 103% for Cr and Ni, respectively. Also, XRF and ICP-OES analysis proved to be very reliable and provided comparable results. In rock samples, ICP-MS allowed the determination of trace and rare earth elements not detectable by the other techniques.

techniques (ICP-OES or ICP-MS). Other techniques, such as X-ray fluorescence (XRF), can provide direct analysis of the sample. The current trend in analytical methods favors closed-vessel solubilization rather than the traditional openvessel systems, since the closedvessel method makes it possible to obtain quantitative recoveries for volatile elements of toxicological interest and improves method detection limits (MDL) by minimizing the required amount of reagents (7–10). Multielement techniques are gradually replacing monoelement techniques because they provide a complete elemental profile of the samples (11) and ensure a significant reduction in analysis time. In ophiolitic areas, environmental samples such as rocks, sediments, soils, and plants are characterized by high concentrations of Ni and Cr. However, the determination of these elements can pose serious problems due to the complexity of the matrices, particularly when siliceous materials are present which cannot easily be dissolved completely with common strong mineral acids (12). Ophiolitic environments are characterized by the presence of mafic or ultramafic rocks, partly or completely serpentinized. These rocks usually show high concentrations of Ni and Cr (sometimes up to a few thousands mg/kg), which decrease progressively as the rock acidity rises.

*Corresponding author. Tel: ++39 0382/ 592311 Fax: ++39 0382/ 592072 e-mail:

Atomic Spectroscopy Vol. 24(5), September/October 2003


Since ophiolitic soils were formed as deposits (colluvium or alluvium) due to the alteration of mafic rocks, Ni and Cr are commonly found co-precipitated with Fe and Mn oxides. Unlike Fe2+ and Mn2+ ions, which precipitate in situ, Ni2+ and Cr3+ ions are relatively more stable in water solutions and therefore can migrate over longer distances. In these soils, Ni2+ and Cr3+ ions are often found in aluminosilicate structures or as oxides mixed with Fe3+; their solubility being inversely proportional to soil acidity (pH). The determination of Ni and Cr concentration in soils and plants from an ophiolitic site is crucial for environmental monitoring, since both elements can be considered natural tracers of geological origin (13). In the present work, analyses were carried out in two different labs in order to evaluate the applicability of microwave dissolution procedures, in conjunction with multielement spectroscopic analyses (ICP-OES/ICP-MS), for the elemental analysis of rocks, sediments, and plants collected in an ophiolitic area. X-ray fluorescence was used to compare the different analytical methods, while alkaline fusion was employed as an alternative technique to closed-vessel acid digestion for the solubilization of samples with very complex matrices. EXPERIMENTAL Area of Study The analyzed area (Figure 1) is located upstream of the Trebbia River Valley. This area is part of an extensive formation of serpentine outcrops, covering several hundred square kilometers, with high concentrations of Cr and Ni ranging from 500 to 1220 µg g–1 and 900 to 1500 µg g–1, respectively. Samples of rocks, sediments, and some of the most common botani-

Fig. 1. Location of serpentinized ophiolitic outcrops in the Apennines of the Piacenza and Parma provinces, Commune of Bobbio, Piacenza, Italy. The investigated area is marked with a circle.

cal species (i.e., Alyssum spp., Helichrysum, and Euphorbia) were collected in this area in September 2001. Instrumentation Two different microwave digesters were used in each lab: MLS 1200 Milestone (FKV, Sorisole, Bergamo, Italy) and MDS-2000 (CEM, Matthews, NC, USA). The instrumental operating conditions reported in Table I are those described in a previous paper (14). Two inductively coupled plasma optical emission spectrometers were used: A PerkinElmer Optima™ 4300 DV simultaneous ICP-OES (PerkinElmer Life and Analytical Sciences, Shelton, CT, USA), equipped with an AS-90 autosampler, a cross-flow nebulizer, and a Scott-type Ryton® spray chamber, and a Jobin Yvon 24 sequential ICP-

OES (Jobin Yvon, France) with a cross-flow nebulizer. A Perkin Elmer SCIEX ELAN® 5000 inductively coupled plasma mass spectrometer was also used (PerkinElmer SCIEX Instruments, Concord, Ontario, Canada). This was equipped with a PerkinElmer GemTip™ nebulizer and a corrosion-resistant Scott-type spray chamber. The operating conditions were reported in previous papers (12,14) and are summarized in Tables II–IV. XRF analyses were performed using a Philips PW1400 spectrometer with a Rh-target x-ray tube and Philips x-14 software package (Milan, Italy). The instrument has a four-position internal sample turret with a 72-position external sampler (PW1500/15 autosampler) which allows analysis of the ratio of the standard and several unknowns in a batch.


a 250 mg of soil / sediment treated with 2 mL HF.552 – 206.754 – 327./Oct.647 220.68 MHz Incident power 1300 W 900 W Reflected power <5 W <5 W Ar gas flow rate: Outer gas 15 L min–1 13 L min–1 –1 Auxiliary gas 0.418 206.353 a – 261. Sept. 8 mL aqua regia.396 – 234. 2003 TABLE I Operating Conditions for Microwave Systems Plant Leaves a CEM 2000 / Milestone 1200 Step Power (W) Hold time (min) 1 250 2 2 400 2 3 500 15 TABLE III Wavelengths Used for the PerkinElmer Optima 4300 DV ICP-OES Element Wavelengths (nm) Cd 214. b The boric acid-satured solution was added before step #5.200 – 213.604 a – 221.Vol.0 mL min–1 1.5 L min <1 L min–1 –1 Aerosol gas 0. TABLE II ICP-OES Instrumental Operating Parameters PerkinElmer Jobin Yvon Optima 4300 DV Model 24 Simultaneous ICP Sequential ICP Frequency 40 MHz free-running 40.616 a 205.569 231.5 mL min–1 Sample pump Peristaltic pump Gilson Minipuls™ 2 Computer Dell® Optiplex GX150 Siemens Nixdorf PCD-4RsxA Printer HP LaserJet® 2200D Fujitsu DL1150 Autosampler Model AS90 – 161 .856 a a 250 mg of plant leaves treated with 7 mL HNO3.716 a 324.9 L min–1 Nebulizer Model GemTip™ cross-flow Scott-type cross-flow Sample flow rate 1. 1 mL H2O2. 24(5).149 – 267.610 a – 260.802a Co Cr Cu Mn Ni Pb 230. and 2 mL H3BO3 in each of six vessels.786 – 228. and 0.754 a 257.8 L min 0.2 mL HF in each of six vessels. Soils and Sediments a CEM 2000 / Milestone 1200 Step Power (W) Hold time (min) 1 250 8 2 400 4 3 600 6 4 0 2 5b 6 300 0 3 2 Zn a Wavelength used for the analysis.438 – 228.

following the program reported in a previous paper (12). 50.09 Cr + 0. Eighteen elements were determined in the sediment sample. The ICP mass spectrometer and optical emission spectrometers were calibrated using certified multielement solutions diluted with water that contained the same amount of acids as the samples.83 L(min 1 mL/min Cross-flow Peak-hop transient Normal 150 ms 50 ms 3 5 1’06" 50 s 60 s External calibration 10. The possibility of using two different wavelengths for every analyte in order to rapidly verify the presence of spectral interferences Interelemental corrections: 75 51 As = 75As – 3. An aliquot of the 0.TABLE IV Instrumental Parameters for ICP-MS Analysis RF power Plasma argon Nebulizer flow Auxiliary flow Sample flow rate Nebulizer Data Resolution Reading time Dwell time Sweeps / replicates Number of replicates Sample time Sample read delay Autosampler wash delay Calibration mode Calibration standard Curve fit Isotopes As Ba Be Cd Ce Co Cr Cu Cs Ga Gd Hf Ho La Mo Mass 75 138 9 114 140 59 52 63 133 71 158 180 165 139 98 1085 W 14. About 0. and reagent-grade boric acid. Merck.2 mm) was subjected to microwave acid digestion with HF/HNO3/HCl. Germany) were used: nitric acid (65% m/v).087 V = 51 77 53 Se + 2. hydrochloric acid (37% m/v). The solutions obtained from the sample mineralization were diluted to 50 mL and analyzed by ICP-OES and ICP-MS. USA). the solutions prepared by LAB A were analyzed by LAB B and vice versa. Bedford.353 Cr 162 .619 82 52 Se V – 3. 100 µg/L Linear through zero Isotopes Mn Nd Ni Pb Rb Sb Sm Sr Ta U V Y Yb Zn Zr Mass 55 142 60 208 85 121 152 88 181 238 51 89 174 66 90 Reagents and Standard Solutions The following Suprapur® reagents (E.2-mm fraction was also analyzed by XRF. Materials and Procedures Sediments A four-layer sediment sample (0–15 cm. MA. In order to evaluate the uncertainties associated with the instrumental readings and the preparation phase. hydrofluoric acid (40% m/v). The solutions derived from the HF/HNO3/HCl procedure were analyzed by two laboratories using two different instrumental configurations. An aliquot of each layer was then ground to 0.2 mm in a planetary mill with agate balls. The high-purity water used in this analysis was produced with a Milli-Q™ deionizing system (Millipore. 25–38 cm and 38–50 cm) was air-dried and sieved to obtain the fine earth [fractions with a diameter (Ø) of < 2 mm].1 L/min 0. Darmstadt. 30. 20.6 L/min 1.250 g of this fraction (Ø < 0. 15–25 cm.

163 . Mn.026 ND 21 45 Co Cr Cu Mn Ni Pb Zn a 17. 24(5). Pb and Zn observed in the CRM 320 River Sediment. Cr. and Euphorbia. XRF instrument detection limits (IDL) were several µg g–1. nondestructive technique. On the other hand.5 40.3 ± 1. the limits for ICP-OES are mostly due to problems arising during the dissolution steps which do not always permit a complete dissolution of the samples. Cu. The statistical methods used in this study are consistent with those reported in the literature (16).3 – 22. Cu. such as HNO3.250 g of the dried 0. The information given in Table VI shows that Al. Sept. therefore. Plants and Soils Trace elements were determined in plant leaves of the botanical species Alyssum spp.1 697 ± 44 72. and HNO3/HClO4. 22 samples by ICP-MS.4 42. and six trace elements were determined in the soil obtained from beneath these plants. Eleven elements were determined by ICP-OES. elements abundant in ophiolitic areas and difficult to solubilize with common acid mixtures. Using the wavelengths marked with an “a” in Table III. and 17 samples by XRF. in a planetary mill with agate balls in order to obtain 0. HNO3/ HCl.. The ICPOES/ICP-MS and XRF techniques provided comparable results as shown in Table VI. slightly higher than the values found for ICP-OES.2-mm particles. Cr. Particularly good were the recoveries for Cr and Ni. interferences were not a problem. (b) alkaline fusion with lithium tetraborate followed by ICP-OES and ICP-MS analysis. and K concentrations increase as the depth of the soil layers increases. Ni. RESULTS AND DISCUSSION Sediments Table V shows the concentrations of Cd.1a 138 ± 7.4 ± 2. which was present at a concentration of about 0. except Cd (uncertified). and Cr concentrations decrease. at these operating conditions.0 ± 2. The collected samples were dried in a thermostatic oven at 40°C and then ground. except the 0–15 cm depth (R Cr/Ni 1.2%. The mineralized solutions were then diluted to 50 mL and six elements in the botanical samples were determined by ICP-OES. The same procedure was applied for sediment samples. 2003 was evaluated using the PerkinElmer Optima 4300 DV (Table III).5 ± 0. (c) direct analysis by XRF on a pellet sample. Rocks A rock sample (serpentinized ophiolite) was analyzed as follows: (a) microwave acid digestion followed by ICP-OES or ICP-MS analysis.0 42.5 µg g–1. while ICPOES and ICP-MS provide a wider elemental profile. The accuracy of these analytical methods (expressed as % Recovery) was estimated by using the following certified samples: CRM 320 River Sediment for sediments.1 ± 0./Oct. The major limits of the XRF technique are related to accuracy. while the Ca. This confirms the common origin of both analytes from TABLE V CRM 320 River Sediment ICP-OES Determination of Cd. the short-term precision for 1 mg/L of different elements (%RSD for five replicates) was always less than 1. Helichrysum. NIM Dunite and NBS-688 Basalt for rocks. Approximately 0. NIST 1570a Spinach Leaves for plants. Intralaboratory (repeatability or %r) and interlaboratory (reproducibility or %R) precisions were estimated according to ISO standard 5725 (15).45). which depends on the types (matrices and interelement ratios) of the samples used to calibrate the instrument. These certified materials were treated according to the manufacturer’s instructions.3 138 ± 7 NP = Not Present. following the program reported in a previous paper (14) and also shown in Table I. This proves that XRF is particularly suitable for a general characterization of elements since it is a quick. Mg.2-mm fraction was submitted to microwave acid digestion (using a mixture of 7 mL HNO3 + 1 mL H2O2 + 0.80 122 ± 6.15). Ba.6 ± 10. ND = Not Determined. without washing. including for trace elements and rare earths. Ni.533 ± 0. The concentration difference measured for all the analytes at the two wavelengths was not larger than 4–5 %.2 mL HF). Acceptable accuracy (recoveries between 89 and 106%) and precision values (%r <10 and %R <26) were obtained for all of the elements determined.2 ± 1. Co.6 142 ± 3 20. Ni.1 619 – 786a 75. ND 88 95 ND 96 96 97 7 9 8 10 6 10 10 26 12 9 12 12 22 14 Indicative values. The Cr/Ni ratio remains constant in all layers (R Cr/Ni 1. Co. Pb. and Zn (µg g–1) After Digestion With HF / aqua regia Element Certified Values Found Values Recovery r R (µg g–1) (µg g–1) (%) (%) (%) Cd NP 0.Vol.0 44. Mn.

55 10.09 1.59 1.3 44. b Determined by ICP-MS.21 9.0 33.59 5. as shown in a previous paper (14).70 8.89 1.50 1. Plants and Soils ICP-OES determination of Al.6 26. except for Mn in the Euphorbia species and Ni in Alyssum spp.0 24.20 25. Ni.26).55 K (%) a 1.2 34.49 9.5 Ca (%) a 9. Cr. The Ca/Mg ratio observed in .52 0.7). This provides useful information on the cycling of elements between two different compartments.9 16. The Al/Ti ratio is constant at various depths (average value: 18.6 74.8 24.33 9. the TF values found in uncontaminated agricultural and forestry soils are less than one unit.38 1.75 a Mg (%) 2.7 28.12 30.53 0.29 0. This ensures the complete dissolution of materials particularly resis- tant to acids. Cu.54 0.4 32. with good precision values for Cr and Ni.9 86 –1) b 16. 0.81 1.7 23.0 ND 11.0 ND V (µg g–1) b –1) a Zn (µg g 90.52 1.20 9. The accuracy values found in both labs for plant reference material NIST 1570a Spinach Leaves were between 89 and 100% (Table VII).2 ± 1.39 2. a procedure similar to the one used for the sediment samples. is the ratio between the concentration of an element in a plant and its total concentration in the soil.88 26.16%).70 8.50 Fe (%) a 9.04 2. which means that both analytes can be used as normalization elements in order to calculate enrichment factors for trace elements in environmental biomonitoring studies (17).9 26.TABLE VI Determination of Major (%) and Minor (µg g–1) Elements in a Sediment Sample Collected in Ponte Barberino (Analysis by XRF or ICP-OES/ICP-MS After Sample Dissolution With HF/aqua regia in a Microwave Oven) Depth 0–15 cm 15–25 cm 25–38 cm 38–50 cm Element XRF ICP-OES/ XRF ICP-OES/ XRF ICP-OES/ XRF ICP-OES/ ICP-MS ICP-MS ICP-MS ICP-MS Al (%) a 5.59 a 0.0 Cu (µg g–1) b Mn(µg g–1) a 797 845 936 902 995 972 1024 942 Ni (µg g–1) b 123 171 131 116 98 103 90.22 0. the high Ni value (TF 8.01 2.14 0.27 9.0 42.15 Na (%) a Si (%) a 23.30 0.0 51.3 25. Cr. as already reported in the literature (18.51 Ti (%) Ba (µg g–1) a 231 250 299 320 319 330 420 411 11.8 ND 97. Mn. The values given in Table IX confirm this rule.40 1. Table VIII shows the concentration of some analytes in the botanical species collected in the area under investigation.35 2. 164 Table IX shows all the values found.17 2. Fe.4 9.55 2.8 13. e.44 2.1 101 113 123 115 149 129 123 a Determined by ICP-OES. ND = Not Determined.49 0.33 0.46 1. Mg. species suggests hyperaccumulation.31 0. Mn.19). Analytical accuracy and precision can be enhanced by adding HF to the solubilization mixture.51 2.0 90. Cu. Fe.20 0.9 44.g. while Euphorbia shows a significant concentration of Mn (approx. along with the soil/plant transfer factors. High Ni values were observed in the Alyssum spp. important and often used in environmental studies.29 8.1 %).5 26..65 1.36 1.20 0.18 9.5 26.55 0.02 8. ophiolitic rocks upstream of the investigated area.2 28. and Zn in botanical and soil samples was performed after closed-vessel microwave acid dissolution.1 10. This parameter (TF).28 0.0 43.87 25. compounds containing Cr and Ni.4 Pb (µg g Sr (µg g–1) a 417 437 242 224 229 220 251 181 62.3) in the Alyssum spp. As a rule. samples (approx.59 1.19 0.9 38. While a preferential accumulation of Mn is quite evident (TF 1.40 7.0 Co (µg g–1) b Cr (µg g–1) b ND 197 ND 169 ND 149 ND 125 31.5 43. and Zn. Ni.8 42. Additional analyses were carried out in the soil obtained from beneath these plants using the same procedures described for sediments in order to determine the concentration of Ca.26 8.

73 97 6 34 NP = Not Present.020 0.315 165 .41 11.13 4.3 ± 0. Helichrysum (µg g–1) (µg g–1) 23 ± 2.9 ± 8.15 102 ± 24.104 1.319 0. Cu.006 Cu 46 ± 1.05 344 ± 3. Cu.8 93 10 38 Ni 2.7 134 ± 1.15 ± 0. 24(5). Mn.018 Zn 114 ± 11 0.098 Fe 44252 ± 9575 0.27 89 19 35 Fe NP 260 ± 30 ND 8 36 Mn 75. Mn. Ni and Zn in Ophiolitic Soil Beneath Botanical Species Soil/Plant Transfer Factors (TF) for Alyssum spp.73 0. Helichrysum.10 2.0 ± 1.21 1626 ± 14.63 ± 0.99 0. Mn.81 TABLE IX Concentrations of Cr..050 0.9 ± 1.008 Mn 1289 ± 268 0. Ni and Zn (µg g–1) After Sample Digestion With HF/HNO3/H2O2 in a Microwave Oven Element Certified Values Found Values Recovery r R (µg g–1) (µg g–1) (%) (%) (%) Cr NP 1. Ni and Zn (µg g–1) After Microwave Digestion With HF/HNO3/H2O2 Determination of Ca/Mg Ratio Alyssum spp.324 0.22 36 ± 0.2 ± 8.45 ND 35 82 Cu 12..9 70. 2003 TABLE VII NIST 1570a.261 Ni 1260 ± 306 8.Spinach Leaves. Helichrysum.9 ± 0.415 0.89 ± 0.14 ± 0.14 ± 0./Oct.6 10. ND = Not Determined.087 0.83 0.52 ± 0.9 ± 1.248 0.72 10375 ± 1711 13.4 ± 0.83 1.36 100 46 47 Zn 82 ± 3 79.028 0. and Euphorbia Element Soil TF TF TF (µg g–1) Alyssum spp. Cr.23 4.82 Element Cr Cu Fe Mn Ni Zn Ca/Mg Euphorbia (µg g–1) 7. Fe.010 0. Sept.2 ± 8. TABLE VIII Alyssum spp. Fe.8 22.31 37 ± 16.3 79.004 0. Fe. Helichrysum Euphorbia Cr 1160 ± 203 0. Mg.20 1221 ± 210 197 ± 4. ICP-OES Determination of Cr.Vol. Cu.2 ± 0.002 0.079 0. and Euphorbia Leaves ICP-OES Determination of Ca.

00 8. CONCLUSION This study demonstrates that accurate analytical data can be obtained in very complex environments.7 DL = Detection Limit. comparable to those obtained by XRF for most of the elements.botanical species. the use of intraand interlaboratory comparisons are essential to verify the reliability of an analytical method in terms of accuracy and precision.2 51. most of all. for the last two analytes.01 0.26 0. the use of certified reference materials with matrices similar to the samples investigated and.28 Sr (µg g V (µg g–1) 40. due to lower instrumental detection limits (IDL).6 20. The ICP-MS technique. This is probably due to the XRF calibration which is strongly affected by the availability of certified reference materials with matrices similar to those of the samples used.28 1.82 9. 166 .9 Cu (µg g–1) –1 Mn(µg g )) 1704 1654 1032 799 839 Ni (µg g–1) 2050 2110 1971 1840 1941 Pb (µg g–1) 7.20 0. except for Ho. particularly for Cr TABLE X ICP-OES and XRF Determination of Major (%) and Minor (µg g-1) Elements in NIM Dunite and in an Ophiolitic Rock Sample Elements NIM Dunite Ophiolitic Rock Certified Found Values Found Values Values ICP-OES XRF ICP-OES XRF Al (%) 0. In this study. such as an ophiolitic site.0 90. Sr. Cr and Ni percentage recoveries ranged from 89 to 100%.66 0. with recoveries between 80 and 120% for NIM Dunite (Table X) and between 76 and 114% for NBS-688 Basalt (Table XI).21 0. allowed the determination of approximately 20 trace and rare earth elements which cannot be determined by ICP-OES at these concentration levels.1 14.70 <DL –1) 3.7 13. followed by ICP-OES and XRF analysis. Sm. ND = Not Determined. Ba. In the real ophiolite sample. the use of microwave digestion in conjunction with ICP-OES and ICP-MS provided very good analytical results. which is further evidence of the ophiolitic nature of the analyzed site.9 Cr (µg g–1) 2900 2760 ND 1170 ND 10.87 12.0 79. The use of multielement independent techniques with analytical sensitivities suitable for the determination of trace metal concentrations. except for Ca and Cu which were clearly underestimated by XRF. only indicative values are reported in the literature. Sb.1 28.016 0.90 4.. the agreement between ICP-OES and XRF results was on the whole satisfactory. Following the comparison between the employed techniques.5 47.8 < DL 55 <DL Co (µg g–1) 210 198 122 103 61.7 20.0 10.67 Fe (%) 11.01 0.1 Ti (%) 0. since the recovery of a large number of the analytes was incomplete. and ICP-MS.0 43.01 Mg (%) 26. Rocks Elemental determination in rocks was carried out by comparing the results obtained by three different analytical methods: XRF.012 0. Table X does not show the results of some preliminary tests performed by microwave acid digestion.43 < DL 0.04 0.75 Ca (%) 0.3 < DL 90.22 0. Table X shows the concentration values of both major and trace elements found in the NIM Dunite certified sample and in the ophiolitic rocks collected in Ponte Barberino. and in accordance with the results obtained by Pelosi (20) and Vergnano Gambi (21).0 20.1 20.08 K (%) 0.03 0.9 83.20 0. by using solubilization procedures based on the type and nature of the matrices.15 18. good accuracy was obtained for almost all analytes using ICP-MS.7 65.01 0. The use of HF in the acid mixture employed for sample dissolution provided quantitative recoveries for the determination of the total content of analytes. These samples were analyzed using alkaline fusion for dissolution. particularly in Alyssum spp.1 8.3 19. Pb and Cu concentrations were not detected by X-ray fluorescence.07 –1 Ba (µg g ) 10. and Yb.11 < DL 5.6 Na (%) 0. Zr.4 Zn (µg g–1) 90. ICP-OES.0 12.1 24.00 4.05 0.18 19.02 0. Alkaline fusion at about 1000°C ensured a complete solubilization of silicates.011 0. was less than one unit.01 0. the recovery of most of the analytes in the NIM Dunite material (Table X) was better with ICP-OES than XRF.18 0. For sediment samples.02 < DL Si (%) 18.

01 (0.79a (2.5 ± 0.4)b NA (9.05 ± 0.4a NA NA 13.19 3.0 ± 0.16 ± 0.01)b (5.86 7.Vol. ND = Not Determined.42 ± 0.12 ± 0.01 22.30 ± 0.6 ± 0.0 ± 1.9 1170 1. 2003 TABLE XI ICP-OES/ICP-MS Determination of Macroelements (%).01 10.02)b 250a (242 ± 8)b (19 ± 8)b 2.07 5.2 48 1.26 8.01 62.59 0.58 ± 0.20)b 58a (84 ± 10)b (60.31 ± 0.70 ± 0.55 ± 0. Sept.86 1.60 ± 0.4a (3.4)b 1.16 ± 0.30 ± 0.24 ± 0.24 ± 0.89 ± 0.6 0.78 1.95 0.44 ± 0.27 Recovery (%) 102 102 103 75 106 56 99 97 94 ND ND ND 81 101 110 ND 91 67 76 40 80 ND 79 102 83 45 68 77 114 94 84 76 107 77 Found Values ICP-OES/ICP-MS 0.6 ± 1.08)b (0.47 72.3 ± 0.86 1.10a 1.02 19.15)b 17.07)b 0.09 0.91 ± 0.2 0.2 < 0.47 ± 0.75 ± 0.20)b 2.02 ± 0.9 16.35 1.68 ± 0./Oct.09 7.0 ± 1.30 ± 0.05 0.01 14.2 < 0.7 0.2)b (0.2 < 0.05 55 0.07 0.3a (13 ± 2)b 49.10 ± 0.81 ± 0.27 ± 0.03 0.23 ± 0.31 ± 0.03 0. a Indicative value.04 189 ± 6 ND 1.47 8.2 154 2.58 ± 0.06 200a (197 ± 12)b 2.2 50 ± 4.88 ± 0.02 22.1)b (158 ± 30)b 1.26 236 ± 4.70a (8.41 46.19 0.08 1.09a (2.6a (1.90 NA = Not Available.58 ± 0.41 ± 0.36)b 7.95 1.05 8. Trace Elements and REEs (µg g–1) in NBS-688 Basalt and in an Ophiolitic Rock After Alkaline Fusion NBS-688 Basalt Ophiolitic Rock Element Al (%) Ca (%) Fe (%) K (%) Mg (%) Na (%) Si (%) Ti (%) Ba (µg g–1) As (µg g–1) Be (µg g–1) Cd (µg g–1) Ce (µg g–1) Co (µg g–1) Cr (µg g–1) Cs (µg g–1) Ga (µg g–1) Gd (µg g–1) Hf (µg g–1) Ho (µg g–1) La (µg g–1) Mo (µg g–1) Nd (µg g–1) Ni (µg g–1) Rb (µg g–1) Sb (µg g–1) Sm (µg g–1) Ta (µg g–1) U (µg g–1) V (µg g–1) Y (µg g–1) Yb (µg g–1) Zn (µg g–1) Zr (µg g–1) Certified Values 9.01 0.22 ± 0.47 1840 0.05 5.90 ± 0.18 ± 0. 24(5).14 < 0.03 0.35 ± 0.43 ± 0.56 < 0.37a (0.07 2.9)b Found Values ICP-OES/ICP-MS 9.02 8.36 < 0.2 ± 0.28 162 ± 3 1.01 0. 167 .66 1.0 ± 1. b Consensus value reported by Gladney (22).41 ± 0.01 2.09 4.04 0.14 ± 0.6 367 ± 1 ND 16.82 0.2 < 0.7a (49 ± 3)b 332 ± 9 (0.

M. Fe. 187. Investigations on a nickel accumulating plant: Alyssum bertolonii. including for some trace and rare earth elements. Biol. Switzerland (1986). Mg. Spectrosc.M.E. and S. Publ. M. 9. D. Baffi. Cu. 821 (2002). calcium and magnesium content and distribution during growth. Radford.C. Spezia. VA. Desv. Genève. Ph. and G. G. C. Berman. M. Anal. REFERENCES 1. 17. while ICP-MS made it possible to obtain a wider range of chemical information. Baffi. Desv. 21. AOAC. the tests performed for this study suggest that for a first screening of samples. I. In: Spendley W. Proc. and C. 5. Quevauviller. 143 (1996). Introduction to microwave sample preparation Theory and practice. 19. (2000). P. Ph. At. Environ. C. Vernimont.which is often difficult to dissolve completely. London. Baffi. 60 (2000).E. Mn. (a cura di). and Zn in botanical samples. 53. G. R. 2003. M. H. Ni. 1641 (1976). and T. S. Beone. Spectrom. 12. 10. UNI CEI ENV 13005. XRF is more suitable than a multielement spectroscopic technique by virtue of its reduced analytical times. 16.C. NBS Spec. 3rd ed. 18. Bettinelli. At. 4. 2. EURACHEM/CITAC GUIDE: 2000. Galoppini.R.B. F. Bettinelli. I.. Bettinelli. 323 (1989). 233 (1993). and S. 3. Beone. 168 . 20. Webbia 33(2). Interlaboratory evaluation of an analytical process. C. M. 269 (1979). Second Edition (2000). and H. Fiorentini. II. 22. Beone. Morgan ed. On the other hand. editor. Vergnano Gambi. Convegno Nazionale Società Italiana della Scienza del Suolo. E. Chem. Use of statistics to develop and evaluate analytical methods 3rd ed. alkaline fusion seems to be the most suitable procedure to obtain complete recoveries for elements such as Cr and Ni. geological.C. Applicazioni dell’ ICP-MS nel laboratorio chimico e tossicologico. Soc. Quevauviller. Washington. Spezia. Cr. and environmental standard reference materials. van Renterghem. Bettinelli. Gills. Cantarini. On the nature of nickel compounds in alyssum bertolonii. Djingova and J. 50 (2000). A. 40(8).M. Muntau. and C. Nassisi. R. Griepink. Chemosphere 48. Precision of test methods – Determination of repeatability and reproducibility for a standard test method by interlaboratory tests. M.M. 11. Jassie. and L. Baffi.S. Talanta 57. USA. Kingston and L. Webbia 32(1). 175 (1977). Bettinelli. Parva Naturalia 20 (2002). Talanta 54. G. Roelandts. M. M. M. Int. Chem. Agr. Ivanova. American Chemical Society. J. Microchemical Journal 73. Spezia.R. Spezia. 21(2). C. The analysis of Alyssum spp. D. O. For rock samples. Bettinelli. S.T. Compilation of elemental concentration data for NBS clinical. Vergnano Gambi. At. L’accumulo di nichel nelle specie italiane del genere Alyssum. R. This is particularly important from a geological and environmental point of view. C. Herzig. H. Beone. Nickel accumulation by European species of the genus Alyssum. 14. 260–11 (1987). Brooks and C. ICP-OES and ICP-MS can perform better characterization of a wider profile of elemental concentrations. Generally. 217 (1978). Laboratory of Government Chemist. Spectrosc. O.T.S. 17. Bettinelli. Sci Tot. R. Hoenig. Tecniche. 299 (2002). S. Ronchi. Spezia. and S. 8. R. Sturgeon. P. In: ISO Standard Handbook 3. 183 pp (1990). L. Matusiewicz. Minoia. 1021 (2001). Spectrosc. Perotti. G. which includes both trace elements and rare earths. ISO (International Organization for Standardization) 5725. Suoli tipici in: Guida all’escursione pedologica. Bizzarri. Bergonzi. Pancaro. 133 (1996). Received June 10. Microwave acid digestion followed by ICP-OES analysis was particularly suitable for the determination of Ca. Pelosi. G. Brooks. Statistical Methods. 21(2). 4. B 200. and S. H. A. R. and B. biological. R. Formica. At. Spezia. leaves and the calculation of the soil/plant transfer factor showed that this species is a hyperaccumulator for Ni. and C. 131 (2002). and S. Quantifying Uncertainty in Analytical Measurement. Environ. Gladney. Alberici. XRF and ICP-OES provided comparable results for a large number of elements. B. Radford. 450 pp. J. 15.M. Nickel. O’Malley. 772 pp (1998). C. 13. 6. Mascetti. Arligton. 7. Muntau. Anal. Guida all’incertezza di misura (2002).

24(5). Preconcentration and separation can solve these problems and lead to a higher confidence level and easy determination of the trace elements.29 and 1. Any analyte solubilized in the hydrophobic core of the micelles will separate and become concentrated in the small volume of the surfactant-rich phase. speedy. The technique is based on the property of most non-ionic surfactants in aqueous solutions to form micelles and become turbid when heated to a temperature known as the cloud point temperature.Cloud Point Extraction Preconcentration and ICP-OES Determination of Trace Chromium and Copper in Water Samples Jing Li. September/October 2003 . and Gd (21) were determined by spectrophotometry. Separation and preconcentration based on cloud point extraction (CPE) is becoming an important and practical application in the use of surfactants in analytical chemistry (8. China INTRODUCTION Several analytical techniques such as atomic absorption spectrometry (AAS). in which the surfactant concentraABSTRACT A new method is proposed for the determination of trace chromium and copper in water samples by inductively coupled plasma optical emission spectrometry after cloud point extraction (CPE). and Hanbing Lu College of Chemistry. Cloud point methodology has been used to separate and preconcentrate organic compounds as a step prior to their determination in hydrodynamic analytical systems such as liquid chromatography (12. In the present work. and inductively coupled plasma mass spectrometry (ICP-MS) are available for the determination of trace metals with sufficient sensitivity for most applications.. There are many methods of preconcentration and separation such as liquid-liquid extraction (LLE) (1). the detection limits were 1. comparable to those obtained with other separation techniques. Pd (22) by room temperature phosphorescence. However.g. ion-exchange techniques (2). The proposed method was applied to the determination of trace chromium and copper in water samples with satisfactory results. any species that interacts with the micellar system. R. Er (20). Ni and Zn (25). P. we report the results obtained in a study of cloud point preconcentration of chromium and copper after the formation of a complex with 8-hydroxyquinoline (8-Ox). FAX: +86-27-67867955 Atomic Spectroscopy Vol. and activated carbon. metal ions after reaction with a suitable hydrophobic ligand) may be extracted from the initial solution and may also be preconcentrated. silica gel. either directly (generally hydrophobic organic compounds) or after a prerequisite derivatization reaction (e. Central China Normal University. the micellar solution separates into a surfactant-rich phase of a small volume and a diluted aqueous phase. etc. (4–7). concentration of chelating agent and surfactant. Cu (23). respectively. respectively (n=11). tion is close to the critical micellar concentration (CMC). The method is based on the complexation of metal ions with 8-hydroxyquinoline in the presence of non-ionic micelles of Triton X-100. Cr(III) and Cr(VI) (26) by FAAS after CPE using complexing agents. with relative standard deviations of 4. Accordingly. CPE also provides results 169 *Corresponding author.11) offers a convenient alternative to more conventional extraction systems. Cd (24). Wuhan 430079. *Pei Liang.3% and 4. highly efficient. Taqing Shi. The proposed method was applied to the determination of chromium and copper in water samples.8%. sorption on the various adsorbents such as chelate resin.13) and capillary electrophoresis (14). The use of preconcentration steps based on phase separation by cloud point extraction (10. inductively coupled plasma optical emission spectrometry (ICP-OES). the determination of trace metal ions in natural waters is difficult due to various factors. e-mail: Liangpei@etang. and of lower toxicity to the environment than those extractions that use organic solvents. Above the cloud point. co-precipitation (3). The effect of experimental conditions such as pH. and later analysis by inductively coupled plasma optical emission spectrometry using Triton® X-100 as the surfactant. The phase separation phenomenon has also been used for the extraction and preconcentration of metal ions after the formation of sparingly water-soluble complexes (15-18).9). The small volume of the surfactant-rich phase obtained with this methodology permits the design of extraction schemes that are simple. equilibration temperature and time on cloud point extraction was studied.31 ng mL-1 for chromium and copper. Under the optimum conditions. particularly their low concentrations and matrix effects. U (19).

R. USA) was used.. A Model 80-2 centrifuge (Changzhou Guohua Electric Appliance Co. 1. Figure 1 shows the effect of pH on the extraction of chromium and copper complexes. P.1 mol L–1) was prepared by dissolving appropriate amounts of Na2B4O7 10H2O in doubly distilled water.1 mol L–1 HNO3 was added to the surfactantrich phase. The samples were assayed as described in the previous section. a middle range of pH at 9.45-µm pore size membrane filter to remove suspended particulate matter and were then stored at 4oC in the dark.0 mL 0. Cu 324.2x10–2 mol L–1 solution of 8-Ox was prepared by dissolution in ethanol from the commercially available product. Shanghai. R. 1. TABLE I ICP-OES Instrumental Operating Conditions and Wavelengths of Emission Lines Examined Parameters RF power 1. Hence. Changzhou. All other reagents were of analytical reagent grade and all solutions were prepared in doubly distilled water.3 kW Plasma gas (Ar) flow rate 15 L min–1 Auxiliary gas (Ar) flow rate 1.0 mL of 0. The pH values were controlled with a Model 320-S pH meter (Mettler Toledo Instruments Co. Ltd. P. Effect of Buffer Concentration The influence of buffer amount was carried out in which the other experimental variables remained constant. The extraction yield depends on the pH at which complex formation occurs.0 mL of a solution containing 3. To a 10-mL water sample. Thus. The phase separation was accelerated by centrifuging at 3000 rpm for 5 min.EXPERIMENTAL Instrumentation An Optima™ 2000 DV inductively coupled plasma optical emission spectrometer (PerkinElmer Life and Analytical Sciences.8 nm mostatic bath maintained at 100oC for 25 min. Working standard solutions were obtained by appropriate dilution of the stock standard solutions. The final solution was introduced into the nebulizer of the spectrometer by conventional aspiration. China) was used to accelerate the phase separation. 1 mL 0. Since the surfactant density is 1. After phase separation.6 nm. The operating conditions and the analytical wavelengths are summarized in Table I. Stock standard solutions of chromium and copper at a concentration of 1000 µg mL–1 were obtained from the National Institute of Standards (Beijing. Analysis of Water Samples Prior to the above preconcentration procedure.2x10–2 mol L–1 8-Ox plus 1.1 mol L–1 HNO3 was added to the surfactant-rich phase. The aqueous phases can readily be discarded simply by inverting the tubes. Shelton.5 L min–1 Pump flow rate 0. added to 10 mL of solution.0 was chosen for these analytes.1 mol L–1 Na2B4O7 10H2O buffer solution (pH 9. After cooling in an ice-bath. Reagents and Standard Solutions The non-ionic surfactant Triton X-100 was obtained from Amresco and was used without further purification. no variation took place in the extraction yield. CT.37 g mL–1. RESULTS AND DISCUSSION Effect of pH Cloud point extraction of chromium and copper was performed in different pH buffer solutions. . Procedures For the CPE.0 mL buffer solution was added in all subsequent experiments.0 mL of buffer solution. P. the surfactant-rich phase became viscous and was retained at the bottom of the tube.0. Jiangsu.0–10. The separation of metal ions by the CPE method involves prior formation of a complex with sufficient hydrophobicity to be extracted into the small volume of surfactant-rich phase.0 g L–1 Triton X-100 and 1..R. The pipettes and vessels used for trace analysis were kept in 10% nitric acid for at least 24 h and subsequently washed four times with doubly distilled water.0) was added. In order to decrease the viscosity and facilitate sample handing prior to the ICP-OES assay. It was found that extraction was quantitative for chromium and copper in the pH range of 7. China). A stock buffer solution (0. Ltd. . aliquots of 10 mL of a solution containing the analytes.0 L min–1 Nebulizer gas (Ar) flow rate 0. Triton X-100 and 8-Ox buffered at a suitable pH were kept in the ther- 170 . The results showed that above 1.8 mL min–1 Observation height 15 mm Wavelength Cr 283. 1. China) supplied with a combined electrode. all water samples were filtered through a 0. the surfactant-rich phase can settle through the aqueous phase. thus obtaining the desired preconcentration. A 1.

Cd2+. Na+.0. 3. 1. 10 mL of a solution containing 0. Zn 10 171 .Vol. maximizing its concentrating factor. The variation in extraction efficiency of Cr and Cu within the Triton X-100 range of 0.1–6. Thus. 24(5). Effect of Triton X-100 concentration on the extraction recovery: 50 ng mL–1 Cr and Cu. The results showed that an equilibration temperature of 100oC and an equilibration time of 25 min was adequate to achieve quantitative extraction. Fig.5 µg of Cr and Cu in 3. The dependence of extraction efficiency upon equilibration temperature and time was studied with a range of 60–120oC and 5–30 min. Effect of 8-Ox concentration on the extraction recovery: 50 ng mL–1 Cr and Cu. 3. were subjected to the cloud point preconcentration process. 1x10–3 mol L–1 8-Ox.0. Effect of Triton X-100 Concentration A successful CPE would be that which maximizes the extraction efficiency through minimizing the phase volume ratio.0 g L–1 Triton X100 and at a medium buffer pH of 9. Pb2+ 2+ 2+ Mn . The yield increases up to an 8-Ox concentration of 8x10–4 mol L–1 for Cr and Cu and reaches near quantitative extraction efficiency. Effects of Equilibration Temperature and Time It was desirable to employ the shortest equilibration time and the lowest possible equilibration temperature as a compromise between completion of extraction and efficient separation of phases. The extraction yield as a function of the concentration of the complexing agent is shown in Figure 2. respectively. It can be seen that the presence of major cations and anions has no obvious influence on CPE under the selected conditions.Co2+. pH 9. Fig. Effect of 8-Ox Concentration For the two cations studied. pH 9. 2.0 g L–1 was examined and the results are shown in Figure 3. 1x10–3 mol L–1 8-Ox.0 g L–1 was chosen as the optimum surfactant concentration in order to achieve the highest possible extraction efficiency. Ca2+. the only interferences studied were those related to the preconcentration step. Quantitative extraction was observed when the Triton X100 concentration was above 3.0 g L–1 Triton X-100. Cations that may react with 8-Ox and extracted to the micelle phase were studied. Sept. a concentration of 3.0. 3. containing various amounts of 8-Ox.0 g L–1 Triton X-100./Oct. TABLE II Tolerance Limits of Coexisting Ions for the Studied Element Determinations Coexisting Tolerance Ions Limit of Ions K+.0 g L–1. Interferences In view of the high selectivity provided by ICP-OES. A concentration of 1x10–3 mol L–1 was chosen to account for other extractable species. thus. 2003 Fig. Effect of pH on the extraction recovery: 50 ng mL–1 Cr and Cu. Ni2+. Ba2+ 1000 100 Mg2+. The results are shown in Table II.

10 mL of each of the samples was preconcentrated with 3. and can be applied to the determination of trace metals in various water samples. Elçi. For this purpose.K. 83 (1992). Mikrochim. Spectrom. S. T. Makharadze.L. P. Acta 320.D. Chem. 24. Laespada. Talanta 44. 1803 (1998). and S. At. Fernández. 20.L. Triton X-100 was chosen for the formation of the surfactant-rich phase due to its excellent physicochemical characteristics: low CP temperature.M.M. J. and T. Watanabe.0 g L–1 Triton X-100 and 1x10–3 mol L–1 8-Ox following the proposed method. Fernandez-Sanchez. 24. Received June 12. J. Anal.F. M. Carabias-Martínez. Pérez Pavón. B. Kamidate.71 13. the detection limits (3σ) of this method were 1.Haraguchi.L. Gladis. Tanaka. Spectrom. Saitoh. 1101 (1996). 1017 (1997). Analyst 118. Sosa Ferrera.C. and D. 7. 14.F. E. 4. Tzouwara-Karayanni.E. C. Moreno Cordero. W. 71. Sahin. Miyajima. T. D. Kawamorita. O.I. G. At. 433 (1987). Pramauro. L. recovery experiments for different amounts of Cr and Cu were carried out. Anal. 15. Silva. Spectrom. and M. 8-Ox is a very stable and fairly selective complexing reagent. 409 (1999). and K.1 mol L–1 HNO3. and M. Perez Pavón. M. The surfactant-rich phase can be introduced into the nebulizer of an inductively coupled plasma optical emission spectrome- TABLE III Determination of Cr and Cu in Natural Water Samples (n = 3) Samples Added (ng mL–1) Measured (ng mL–1) Recovery (%) Cr Cu Cr Cu Cr Cu Tap water 0 0 6. A. J. Rev. Fernández. 3. J. K. 172 . The results are shown in Table III. inexpensive. S. 209 (1993).P. Environ. Sci. The proposed method gives very low LODs and good RSDs. At.H.M. Talanta 25. Analyst 124. 16. and M. 7.1 102 98 20 20 26.Chim. J. Revia and G. A. 5. Anal. Anal. J. Res.B. commercial availability and relatively low price. 17. Spectrochim. Igarashi and K. García Sánchez. 287 (2000). Y. M. 111 (1996). 133 (1993).Romero Beato. Stacchiola. Sci. Endo. C.J. Olsina. Malvankar and V. 2003. 2468 (1999). and G. A. Chim. L.69 18. 38. Wang. S. and L. 695 (1992). M. Martin. Anal. J. and K. Haraguchi. Shinde. Anal. At.M. J. Jimenes de Blas.39 10. Spectrom. 585 (1978). Anal. respectively. E. which facilitates phase separation easily by centrifugation. 6. Spectrom. Ind.D. García Pinto. Dominguez-Alvarez. J. rapid. 26. Olsina. high density of the surfactant-rich phase. 1177 (1998). 23. Silva.31 ng mL–1 for chromium and copper.6 20. 343 (2002).6 23. H. Chem.R. Watanabe. Haraguchi.L. Sanz-Medel. S. and non-pollution methodology for preconcentration and separation of trace metals in aqueous solutions. 11. Bae. E. 487 (1999). R. Cerrato Oliveros. Crit. 143 (2002). CONCLUSION Cloud point extraction offers a simple. 7. 251 (1999). and R. 12. Moreno Cordero.5 Lake water 0 10 20 0 10 20 8. Yaman. T. 11.H.5 30. Anal. Sahin.F. Moreno Cordero. REFERENCES 1. Hinze and E. with relative standard deviations (RSD) of 4. 18. 1081 (1991). Trends Anal.B. 179 (1992). Anal. Kamidate. Quina and W. Anal. and S.L. 4150 (1999). Anal. Sci. 15. Stalikas.A. F. Kimura. J. and B.F. and B. 21. Eiguren Fernández. Mohammad. Fernandez de la Campa. E. Chem. J. U. Kamidate.Y.5 28. and J. M. 37 (1996). S. Analyst 123. Watanabe and H. Analyst 116. U. H. C.8%. Determination of Chromium and Copper in Water Samples The proposed method was applied to the determination of Cr and Cu in tap water and lake water. L.9 100 ter by conventional aspiration after dissolving with 0.L.3 – – 10 10 16. 9. 577 (1989). Spectrom. H. Saitoh. 19. Acta 54B.1 101 – 98. Acta 106. 22. and low toxicity. Paleologos. J. 229 (1997). and B. B. 13. PérezPavón.29 and 1. Elçi. T. 21.Detection Limits and Precision According to the definition of IUPAC.M. Acta 342. 5. H. Karayannis. Z. Littlejohn. Santana Rodriguez. Spectrom.M. In addition. 2. Watanabe. 3. Chem. Eng. R.2 101 99. At. 23. Health A 31. Martin. At. Gonzalez.8 – 98.L. Rao. 25. Soylak. Pilidis. Zeng. Anal. R. The results given in Table III indicate that the proposed method can be reliably used for the determination of these metal ions in various waters. 13. the working standard solutions were subjected to the same preconcentration procedure as used for the analyte solutions. Hernández-Méndez. respectively (n=11).8 3.L. V.3% and 4. Stalikas. X. 547 (1998). RodríguezGonzalo. 13. 10. Güçer and M. Ure. For calibration purposes. Acta 322. Hinze. and D. C. Talanta 48. and J. T. 8. Anal. Biju. At. Murray. M. Chim.A. Öztas. 133 (1996).

and José Luis López-Colón Instituto de Medicina Preventiva del Ejército. With an atomic mass of 78. During the method development process. and biological samples is of clinical importance. respectively. serum. and the chemical industries (1–3). were obtained. conducted by Surrey University in Guilford. liver.96 Daltons. The main sources of exposure are industrial emission. and a linear range of 5.3%. September/October 2003 . even though in some areas it can exist in concentrations high enough to produce livestock poisoning. multifocal myocardium necrosis. Selenium is an essential element to human life. A sensitive method has been obtained. 28047 Madrid. resulting in a maximum value of 1. chronic deficiency has been associated to certain types of cancer. Surrey. For serum samples. paint and varnish production. The valence of 6+ is the most common in inorganic selenium salts. Therefore. Spain INTRODUCTION Selenium is an element that belongs to group VIA of the Periodic Table of the Elements. congestive cardiomyopathy and. Accuracy was calculated by participating in an interlaboratory proficiency study. Guadalupe Álvarez Bustamante. *Corresponding author. with serum and plasma concentrations approximately the same. cardiovascular system. José María Gervas. 24(5). It can be found in four states of oxidation: 0. Accuracy and precision determinations for the blood samples were carried out using Seronorm Trace Elements Whole Blood reference samples. but only 3% of plasma enzymatic activity (8–10). selenium deficiciency causes Kashin-Beck disease. Hospital Central de la Defensa. Epidemically. 4+ and 6+. chronic. selenium depletion in blood has been related to some neurological degenerative illnesses similar to Alzheimer’s (7).07 µmol/L. In blood. finally. most of the selenium is bound to hemoglobin and glutathione peroxidase.49 µ ABSTRACT Selenium is an essential element to human life. classified by their seriousness: acute.4 µmol/L. José Manuel de Pradena. metal treatment. eration. 2–. also called Keshan syndrome and. The activity of this enzyme represents a major part of total enzymatic blood activity. based on Electrothermal Atomic Absorption Spectroscopy with Zeeman Background Correction (ETAAS). Most published and latent. affecting mainly children and women during their child-bearing years. In erythrocytes. striated musculature degeneracy. the so-called Keshan syndrome. while carcinogenesis inhibitory action of the upper selenium levels has been found in some experimental models (6). There are four different types of this illness. selenium determination in blood.5% and 102.09 mg/kg. A deficiency in selenium causes an endemic cardiomyopathy. e-mail: amontel@eresmas. which can be a key contributing factor for the control of selenium ingestion. whose main symptoms are articulation deformation and hypertrophy. In children from five to thirteen years of age. In plasma it is much faster. Other problems related to selenium deficiency have been reported such as neutrofile bactericide capacity depletion. subacute. its concentration in the earth’s crust scarcely reaches 0. Deficiency in selenium causes an endemic cardiomyopathy. it shows intermediate properties between metals and non-metals. Once the illness has been diagnosed. and hyaline cartilage degen173 Atomic Spectroscopy Vol. selenium is distributed in the following way: between 40 to 45% in plasma and 55 to 60% in red blood cells. glass. Selenium determination in biological samples could be of clinical importance in cases of excessive or inappropriate amounts of ingestion of this element. and reduction of 5-diiodine-ase enzyme activity (responsible for triiodothyronine production from thyroxine) (4. Erythrocyte selenium regeneration is a relatively slow process that lasts more or less 120 days. Also. the Kashin-Beck disease.7%. and muscles. Glorieta del Ejército s/n. in children. the ingestion of selenium is of no therapeutic value. External Quality Assessment Scheme. with a detection limit in the samples of 0. It is not a very common element in nature. has been developed. the standard addition technique was used to evaluate precision. with symptoms such as muscular pain. Selenium was discovered by Berzelius in 1817. while Se2– is the most frequent for organic selenium compounds. as well as the mineralization and atomization temperatures.5). Selenium deficiency can produce malfunction in different organs such as the brain. A rapid method of analysis. United Kingdom. principally from electrical. the modifier concentration (palladium nitrate) has been optimized. An average precision and a recovery of 4. a characteristic mass of 0.A Rapid Method for the Determination of Selenium in Blood and Serum by ETAAS With Zeeman Background Correction *Arturo Montel.

for solution preparation. Darmstadt. Ramp Time Flow Gas (ºC) (s) (s) (mL/min) 110 1 20 250 Argon 130 1 30 250 Argon 1250 5 30 250 Argon 2000 0 3 0 Argon 2500 1 3 250 Argon Reading No No No Yes No 174 .studies use erythrocytes.7 nm 290 mA Peak area 3s 20 µL Spectrometer. and standard solutions were diluted 1/10 (v/v) with the matrix modifier solution. Shelton. whole blood or plasma as short-term indicators of selenium ingesta. current methods use Pd (13. The high volatility of selenium and its compounds in complex matrices makes the addition of a matrix modifier indispensable to fix selenium and to reach pyrolysis temperatures as high as possible. Each one of the temperature ramps was evaluated with the purpose of developing a sensitive and linear method for clinical ranges and applications (20). for atomic absorption spectroscopy (PerkinElmer). In ramp temperature optimization studies. avoiding bubble generation during this process. EDL type. also causing a characteristic ‘garlic’ odor in breath. the interlaboratory proficiency study of Surrey University. Serum samples. weakness of limbs. the method of choice for selenium determination is Electrothermal Atomic Absorption Spectrometry with Zeeman Background Correction (12-14). Triton® X-100 (Merck. blindness.16). Nitric acid. USA). "External Quality Assessment Scheme. Germany). 1998) (15. ascorbic acid.2. gastrointestinal problems. hepatosplenomegaly. blood samples. Mainly plasma or serum samples are used in today’s clinical monitoring. or hydroxylamine hydrochloride (J. and respiratory problems. while chronic poisoning causes anemia. Acute poisoning produces nervousness. perspiration. mucosae irritation. In both samples.6 µmol/L (1 g/L) in 2% HNO3. Suprapur® grade (Merck. enabling the removal of the matrix without selenium loss (15). Pd/Mg. paleness. and urine (11). PerkinElmer AA WinLab™ software.17–19). Pyrolytically coated graphite tubes with transversely heated integrated platforms and end caps. most methods used for the determination of selenium used nickel as the matrix modifier (11). while blood samples require an homogenization time of at least 20 minutes. elimination of the matrix effect caused by the presence of iron was taken into 1 2 3 4 5 TABLE II Graphite Furnace Temperature Program Temp. Darmstadt. with Longitudinal Zeeman Background Corrector and AS-800 Autosampler (PerkinElmer Life and Analytical Sciences. PerkinElmer selenium lamp. Germany). Analytical Procedure Experimental Conditions Table I shows the instrumental settings used for the selenium determination in blood and serum samples. Sample Pretreatment Serum samples should be centrifuged for 10 minutes at 2500 rpm. Version 4. and then homogenized. However. Germany). the theoretical values and acceptability ranges were indicated. Reagents Ultrapure water (resistivity 18 MΩ x cm) purified with a Milli-Q™ system (Millipore. dizziness and convulsions. Palladium modifier for graphite furnace. EXPERIMENTAL Instrumentation PerkinElmer AAnalyst™ 600 Electrothermal Atomic Absorption TABLE I Instrumental Operating Conditions Wavelength Slit width EDL lamp current intensity Reading Integration time Sample volume 196 nm 0. Selenium aqueous standard solution. Graphite Furnace Temperature Program Table II shows the temperature program used for selenium in graphite furnace analysis." was employed. Fischer. To assure measurement traceability and to assess method accuracy. CT USA). 10g/L of Pd(NO3)2 in 15% HNO3 (Merck. MA. Internal quality control samples: Nycomed (Seronorm Trace Elements Whole Blood) for blood and Nycomed (Seronorm Trace Elements Serum) for serum. This program has been studied in order to obtain a quick method that ensures the best accuracy and precision. Darmstadt. In the past. Currently. 12. Bedford.

175 . Fig. initially diluted with 1% HNO3 (v/v) in a volumetric flask until a concentration of 0. Optimization of the analytical conditions. Dependence of analyte absorbance (continuous line) and background absorbance (dotted line) vs. atomization temperature with a fixed temperature of pyrolysis at 1250ºC in a Seronorm Trace Elements Whole Blood "level 1" sample. 2. 1. modifier concentration. account.1 mL of Triton® X-100. This element causes lamp energy absorption at a wavelength close to that of selenium. 2. 24(5)./Oct. and the temperatures of the mineralization and atomization stages are described below.6 µmol/L (1g/L). Linearity Calibration standards were prepared from a concentrated standard solution of 12.27.53. An aqueous standard with a final concentration of 0. RESULTS AND DISCUSSION Optimization of the Matrix Modifier Solution The matrix modifier solution was prepared by adding the following amounts to a 100-mL volumetric flask: 1 mL of palladium modifier for graphite furnace analysis [10 g/L Pd(NO3)2 in 15% HNO3 solution] plus 0. 200. 400 and 600 µg/L. then bringing the solution to volume with ultrapure water. Figure 1 shows the dependence of absorbance with the concentration of palladium in the matrix modifier solution. Optimization of Mineralization Stage The pyrolysis temperature selected was based on the one that permitted the best organic matter destruction without loss of analyte. 3. 5. an optimum value of 90 mg/L (1. pyrolysis temperature with a fixed temperature of atomization at 2000ºC. where a stabilization can be observed. selection of the optimum concentration of the matrix modifier.Vol.44 µg of palladium in each injection of sample (20 µL). Dependence of analyte absorbance vs.8 µg / 20 µL sample) was selected. 2003 Fig. Fig.13 µmol/L (10 mg/L) was achieved. and at the same time obtained minimum matrix effect and higher signal-to-noise ratio (see Figure 2). Sept. The greater the palladium concentration. producing a high matrix interference that even Zeeman correction cannot totally compensate for.63. As a conservative approach.60 µmol/L (50. and 7. 100.125 µmol/L was used. Optimization of Atomization Stage The temperature allowing both a higher response and a greater relation between analyte and background signals (see Figure 3) was selected. Dependence of absorbance vs. Working standards of 0. the higher the absorbance until reaching a concentration of 72 mg/L corresponding to 1.07. 1. A Seronorm Trace Elements Whole Blood "level 1" sample was used. This solution is stable for at least three months in refrigerated conditions.

respectively.8%.1 pg). expressed in terms of percentage of the mean value. a blood sample. for a probability of 95%.51·10–4 ± 0.507 µmol/L (0 to 40 µg/L) was adopted. A slope of 1. To carry out a proportionality test. Confidence and prediction intervals of the calibration curve obtained with aqueous standards for a probability of 95%.9992. A value of 0.76 µmol/L (0 to 60 µg/L).34 µg/L) was obtained for standard solutions corresponding to 0.7%. the recoveries obtained were 95. Fig. 176 .2. the concentration range within 0 to 0. p=0. spiking selenium amounts to a blood sample) and the calibration line. As reagent blank solution. 60. The slope of the regression line obtained after representing the measured absorbances against the added amounts was calculated.507 µmol/L (see Figure 4). confidence limits of Y-axis intercept of the line obtained with standard solutions were calculated (95% probability). Quantification and Detection Limits The instrumental detection limit was calculated from the analysis of 30 reagent blank samples. In the addition tests.05). and 106% for concentrations of 0. The correlation coefficient was 0. The Matrix Effect To prove that the measured absorbance is free of interferences from the matrix. As a working curve.4 µg/L) for blood and serum samples when the dilution proposed in this procedure was used.76.05) was obtained. The experimental characteristic mass was 495 µmol (39.respectively) were prepared by diluting this solution with ultrapure water. 0.48. 4. was analyzed. At each point. The slope of the regression line obtained in the measurement of aqueous selenium standards (1. Thus.04 µmol/L (0.. Slope variance.01·10–4. and 1. of the line obtained using the standard addition method and spiked blood samples (1. which is lower than the characteristic mass granted in the instrument manufacturer’s (PerkinElmer) specifications. including zero. The value obtained was 7·10–4± 6.14·10–4.26 µmol/L (40.2. 103. was 0. the sensitivity of the method does not seem to be affected by the presence of the remaining blood constituents. A linear response was observed up to a concentration of 0. p=0.63·10–4 ± 0.e. i. These working standards were diluted 1/10 (v/v) with the matrix modifier solution in order to handle the standards in the same way as the samples. this curve showed a maximum relative standard deviation of 3. Comparison between the slope of the regression line (obtained with the standard addition method.51·10–4 ± 0. The parallel line to that obtained with the standard addition method falls inside this interval. and 100 µg/L). ultrapure water diluted 1/10 (v/v) with the matrix modifier solution was used.05) falls inside the slope confidence interval Fig. 5.7·10–4 (n=15. The dotted lines represent the confidence interval of the calibration line (95% probability).01·10–4 (p=0.4 µmol/L (3. spiked with increasing amounts of selenium. p=0. 570 µmol or 45 pg (21).05) (see Figure 5). The linearity of the calibration curve was evaluated analyzing the former working standard solutions diluted as samples in order to obtain a final range within 0 and 0.

02 2.6% 1.10 1.10 1.01 1.03 5% 102.262 1.567 2.4 µg/L).8% 101. all values obtained fell within the range determined by the mean plus/minus one standard deviation of the participating laboratories.262 Total 1.07 1.262 1.39.07 µmol/L (400 µg/L) which fulfils the TABLE III Determination of Se in a Sample With a Known Se Concentration: Seronorm Whole Blood "level 1" (Selenium concentration: µmol/L) 1st day 2nd day 3rd day 1.5% 1.1.90 µmol/L.5% 102.00 1. the target value was 1.07 1.06 1. as the regression line passes through zero. named Standard Deviation Index (SDI) (see Figure 6). 2003. with a 95% significance level (Table III).03 1.88. The tests for assessing method accuracy for the serum samples were carried out on 12 serum samples within the interlaboratory proficiency study.7%. with a precision value of 4.10 1. 6.05) 1.03 5.1. It turned out to be useful in detecting selenium deficiencies. 2002.03 1. The experimental detection limit of the method for serum and blood samples was 0. Seronorm.01 µmol/L (80 µg/L) of Se.95 0.89 and 2. .6% 1. and in the interlaboratory proficiency study for serum samples indicate that the method shows suitable precision and accuracy values. The mean values of each day fulfil the requirements of the Student’s statistical t-test for mean comparison. 24(5).94% and 1. The obtained standard deviation values varied between 0.92 Mean RSD (%) Recovery Practical "t" Theoretical "t" (p=0. conducted by Surrey University. Thus.07 1.02 0.10 1.512 2.06 1.05 1.95 1./Oct.807 2.01 1.5% 102.4 µmol/L (3. 110.01 1.03 4. Method precision for the serum samples was studied from the analysis of serum samples to which increasing quantities of selenium were added in order to obtain final concentrations in the samples of 0. For this sample. as a result of the considerable difference among the detection limit and the reference selenium values in the above-mentioned fluids." Lot 404107) were employed for precision and accuracy tests.06 1." For the samples received.07 1.Vol.03 1. no matrix effect was observed which makes it possible to calibrate the method with aqueous standards. Revised text received August 12. The tests were carried out on three different days (10 times each day). CONCLUSION No statistically significant difference between the slopes of the lines obtained by measuring aqueous standards and by the standard addition method was observed.00 1. Standard Deviation Index (SDI) for the serum samples of the External Quality Assessment Scheme of the University of Surrey. an average recovery of 102.00 1. The calibration line corresponding to aqueous standards was shown to be linear in the interval between 0 and 5. 2003 Precision and Accuracy Reference samples (Seronorm Trace Elements Whole Blood "level 1. 177 The tests carried out for the reference blood sample. "External Quality Assessment Scheme. requirements of the proportionality test.03 0.30% Fig.01 1. Sept.02 1. 150 and 230 µg/L). both in serum and blood. Received August 19.5%.05 0.98 1.3% was obtained. (70.

" Aspects sanitaires et nutritionnels des oligoelements et des elements en traces. 88 (1998). Moya.S. 4. 13. J.G. Gauvreau. E. and J. Halls. Larsen. Bugel. "Elementos traza esenciales en la practica clinica.J. Rubio.M. Littlejohn. Clin. Sci. Ficher and C. Biol. Trace Elements Med. J." In: Toxicología Médica Clínica y Laboral. 19. "Toxicologia. Biol. Emard.J. Chim. 17. 9. Trace Elements Med. Adamowicz. Fell. 17.L. Bermejo. and T. PerkinElmer Life and Analytical Sciences. Trace Elements Med." In: Toxicologia industrial e intoxicaciones profesionales. 121–135 (2000). Trace Elements Med. Lauwerys. Ed. Acta 53. S. pp." In: Bioquímica Clínica. Spectrosc. GarciaBeltran. Patching and P. and L. D. M. Acta 53. 847–58 (1995). B. Quím. Zachara. M. 2. and E.E. 1st Ed. Gonzalez. S. Vartiainent. J. 6. A. J. Rev. Acta 398. G. and J. Bermejo. Trafikowska. At. 74–81 (1995). Alvarez. OMS. Cocho. Thouez. Spectrochim. 198–204 (2001). Rubio. Kantola and T. 5. Rademeyer. 36. 227–257 (1995). Trace Elements Med Biol 15.L. 15. M. J. Montel.M. J.H. D.. C. 16. J. R. 14. 161-166 (2001). Ann. Sabe.J. 1994). CT USA (1992). and L. Anal. GarciaBeltran.R. A. J. 12. Biol. J. 7. Rademeyer. Halls. Fraga. M. Chem. Med. 15. McGraw-HillInteramericana (1995). R. 279-287 (1999).T.REFERENCES 1.M.P. Gardiner. Biochem. 20. 9. T.H. Barcanova (1994). 11–17 (2001). A. 8. Clín. "Selenio. Sandtrom. Ed. J. 537-548 (1998). E. 149–156 (1999).F. Anal. 20(5). 11. 301–315 (1999). Sheehanl and D. Gardiner. and D. Soc. "Selenium. Forensic Science International 107.E. Schlemmer. R. U. Queralto. De Pradena. Manitius. A. 40(6). Sabe. Hoya. Feuerstein and G.H. The THGA graphite furnace: Techniques and recommended conditions. Masson (1994). 21. 13. R. Spectrochim. Ficher and C. 193–214 (1999). Nartowicz. J. 178 . 18. J. 14. Acta 419. 3.M. J. and G. Lopez Colon.A. 105– 222 (1997). "Selenio y sus compuestos. Biol. Diaz. 180–1855 (1999). Lorenzo. J. P. Chim." Rev. P. J. Quimica Clinica (April 9–11. Ladron de Guevara and V.H. 10. 549–567 (1998). Shelton. Sanz and C.J.

and Cd. On- line separations by complexation with this ligand for ultra-trace multielement determination in waters and biological samples by inductively coupled plasma mass spectrometry (ICP-MS) have also been proposed (20).br ABSTRACT A comparative study is reported between C18 bonded silica gel and powdered polyethylene (PE) as sorbent for Cd. and a sorbent-packed minicolumn. Zn.8) samples. These shortcomings can be circumvented with online sample pretreatment (2.Odiethyldithiophosphoric acid (DDTP) was first studied in the 1960s by Bode and Arnswald (13). and Pb using solid reversed phase sorbent C18 (octadecyl groups bonded silica gel) and diethylammonium-N. Cu. Although solid sorbents such as carbon (14. The best elution flow rate was 6.5 mL min–1.0. V. <2%. depending on the nature and concentration of the acid. a manual injector-commutator.19) and knotted polyethylene tubing (17) have been proposed as solid support for Cd. Cu and Pb determination are lacking in the literature. Pb.0 mL min–1 for Cu. and Pb added to wine samples varied from 96–99%. loss of analyte. Cu. Nb. using C18 or PE.14 mol L–1 HNO3 and processed in a simple flow system comprising a peristaltic pump. 24(5).6. C18 (5–12. in some cases of direct determination of trace metals.14. but does not react with alkali and alkaline earth metal elements and others. 97–102%.6 µg L–1 Cu. or Pb. and Cd. The coupling of FIA-FAAS for preconcentration and the determination of Cd. The optimum concentration of the complexing agent was 0. Ga.15. and biological (6) and environmental (7.25% (m/v) ADTP for Cd determination using PE. Pt. However. Cr. such as Al. and 11 times (C18) or 18. Rh. Universidade Estadual Paulista P. the sampling rate was 40 h–1 and the enrichment factors of 33. amongst others. the possibility of sample contamination. The limits of detection (LOD) were 1. 11 µg L–1 Pb. Cu. there is insufficient sensitivity and limit of detection (LOD) offered by FAAS unless preconcentration steps are used. or ammonium diethyldithiophosphate (ADTP) (7–12) as the complexing agents can be efficiently applied to the analysis of water (5). and 2. and José Anchieta Gomes Neto Instituto de Química. and Pb at Trace Levels in Wine Using Solid Phase Extraction Andréa Pires Fernandes. Cu. and semi-metals in an acidic medium. the performance of C18 bonded silica and powdered polyethylene-packed mini-columns Atomic Spectroscopy Vol. Cu. more systematic studies on the use of C18 and polyethylene for Cd. and 3. Mo. respectively. the main difficulties associated with manual sample preparation are time and reagent consumption. The results revealed that C18 was more efficient for Cu and Pb determination.18). The sample loading flow rates were 5. Cu. *Mercedes de Moraes. Mn.5–7). Accuracy was checked for Cd. 3. and 0. a method for the on-line determination of these species in waters by ETAAS was also proposed (19). Fe(II).Ndiethyldithiocarbamate (DDTC) (4. Ethanol was selected as the eluent and analytes in the eluate were determined by flame atomic absorption spectrometry. 22. In chemical analysis. The flow injection technique has made on-line sample preparation procedures even more attractive because it is an excellent tool for solution management (2).O. e-mail: anchieta@iq. Ir. and <6% for Cd. and Ag in several samples for FAAS (7. 36. respectively. and 90–99%. or risk to the analyst. In this study. For a 60-sec preconcentration time. This reagent has been used for the separation and preconcentration of different analytes such as Pb. It was demonstrated that this ligand complexes several transition metals *Corresponding author. W.unesp.0 µg L–1 Cd using PE. The ammonium salt of O. Cu. and 23 times (PE) were obtained for Cu. Cd. and Pb determination using ammonium diethyldithiophosphate (ADTP) as the complexing agent in a flow injection system. respectively. Th. Ru. Box 355. The relative standard deviations (n = 12) were typically <2%. Cu. 19 µg L–1 Pb. and Zr. Brazil INTRODUCTION Flame atomic absorption spectrometry (FAAS) has for many years been used in the analysis of various samples (1).16–18).9 µg L–1 Cu.11. and Pb determination in six wine samples digested by block digestor and open-vessel microwave-assisted digestion systems. Se(VI).3).16.14.1% (m/v) ADTP for Cu and Pb determination using either C18 or PE. Bi.Flow Injection AAS Determination of Cd. and 1. The recoveries of Cd. September/October 2003 179 .15) or electrothermal atomic absorption spectrometry (ETAAS) detection (8. Ge(IV). and Pb. respectively. Ti. Since the ADTP complexes only As(III) and not As(V). while PE was the best sorbent for Cd. respectively. The complexes were formed in 0. 14801-970 Araraquara-SP. Ta. Pb.0 µg L–1 Cd using C18 or 2.

and 283.8 mm i.0. USA). which was transferred to the TE05/50 digestion block (Tecnal. Six Cd or Cu and five Pb calibration solutions (Cd: 5. Pb: 25.) with about 30 mg C18 or 60 mg of powdered polyethylene. WI.0. As analytical demonstration. MA. At the same time. and Pb using ADTP as the complexing agent was evaluated. Reference Solutions. 15 mA. Milwaukee.0. A 1. Merck). This solution was purified by passing it through a C18 bonded. and Pb determination. 324. A flow injection system coupled to a flame atomic absorption spectrometer was used to perform the detection automatically.8 nm (Cd). Titrisol®. 50. The air-acetylene flame was adjusted and the operating conditions were those recommended by the manufacturer. The wet ashing digestion procedure in an open vessel system employed the Star System 6 microwave oven (CEM. Cu (P/N N305-0157). FIA System Figure 1 shows the FIA system designed for Cd. 150 µg L–1) were prepared daily in 0. The performance of the FIA procedures with C18 and PEpacked mini columns was checked after analyzing digested wine samples and the accuracy was also checked using spike recovery tests. 25. Shelton. Mini columns were prepared by filling a Tygon tube (50 mm x 2. and Samples High-purity deionized water using a Milli-Q™ water purification system (Millipore.0.0. Brazilian red and white wine samples were purchased at a local supermarket in Araraquara City. respectively. 150. Cu. The other metals and species that do not form complexes with ADTP pass through PC and are discharged (W). A 100-mL sample. The FIA system comprised a home-made manual injector-commutador. ethanol (E) continuously flows through the nebulizer of the spectrometer. Milford. 75.0. unless specified otherwise. operating at 4 mA.00 g of the ammonium salt (Aldrich. and 10 mA for Cd (P/N N3050115). silicapacked column. and Pb (P/N N305-0121). a sorbent-packed minicolumn. 50. USA) in about 5 mL of ethanol and diluting up to 100-mL volume with water. Bedford. then stored in a refrigerator.25% (m/v) ADTP solutions were prepared by dilution of the ADTP stock solution in water. Metal complexes formed inside the reaction coil (RC) are retained by the C18 or PE-packed mini-column (PC).7 mm i. USA). an Ismatec IPC–8 peristaltic pump equipped with Tygon® pumping tubes used to manage the complexing agent. EXPERIMENTAL Instrumentation A Model AAnalyst™ 100 flame atomic absorption spectrometer was used (PerkinElmer Life and Analytical Sciences. Octadecyl bonded silica (C18) from commercially available SepPak cartridges (Waters. Cu: 10.d) was used as the reaction coil and transmission lines. 100. 75. Cu. Brazil. Ethanol passes through the mini-column in reverse direction and the eluate containing the analyte flows towards the detector. Germany).d. A poly-tetrafluoroethylene tubing (0. The temperature was then increased to 120oC and kept steady for 8 h. A Viton® organic solvent-resistant pumping tube was used to aspirate the ethanol (eluent).14 mol L–1 with Suprapur® grade nitric acid (Merck. The passage of the eluate zone through the flame of the spectrometer produces a transient absorbance signal with the peak height proportional to the analyte content in the sample. equipped with glass vessels. PerkinElmer Lumina™ hollow cathode lamps. Piracicaba. 50. the mini-column (PC) is inserted into the eluent channel. CT. This minicolumn was initially flushed with ethanol for cleaning.3 nm (Pb). USA) was used to prepare the minicolumns. sample. The ends of the packed mini-column were filled with glass wool to avoid sorbent losses. the sample (S) merges with the ADTP reagent stream (R) at the confluent point x. It consists of two steps: preconcentration and elution/ detection. The 0. were used. Preparation of Samples The samples were prepared in triplicate by two different procedures involving acid mineralization in open systems. Ethanol was used as the eluent without previous treatment. A 50-mL wine sample was 180 . All digests were clear solutions and the final volume was made up to 100 mL with MilliQ™ water. 150 µg L–1. CT. Reagents.0. 100. Brazil).for on-line preconcentration of Cd.0% (m/v) ADTP solution was prepared weekly by dissolving 1.8 nm (Cu). the determination of analytes in wine samples will be presented. 100. equipped with a deuterium arc background corrector for absorbance measurements at the wavelengths of 228. 20 mL of concentrated HNO3. and analytical solutions.14 mol L–1 HNO3 by diluting the individual stock solutions (1000 mg L–1. When the injector was changed to the other position (Figure 1B). Darmstadt. and 600 µg V2O5 were sequentially added to a digestion tube (21). the injector-commutator is switched back to the initial position and another cycle is started.1% and 0. In the position of injector-commutator (IC) as specified in Figure 1A. NC. After peak maximum measurement. 250 µg L–1. 200. All reference solutions and samples were acidified to 0.0. USA) was used throughout.00. Matthews.

The effect of sample flow rate on absorbance was studied in the 1–10 mL min–1 range for all elements. The influence of variation of the eluent flow rate was studied from 2 to 8 mL min–1. 24(5).1 15 – – HNO3 10 – – Optimization The influence of variation in flow rates of the sample (S). Sept. The influence of variation of concentration of reagent R on absorbance was evaluated in the 0.03 A) for C18 and PE. respec- 1 2 3 5 0. Pb: 3. In these situations. respectively.12 A) and 7 mL min–1 (0. x: confluent point.).8 nm. TABLE I Optimized Heating Program of Wine Digestion Using Microwave Oven Step Ramp Temp.d.05–1% (m/v) ADTP range.6 mL min–1 . The signals increased steeply with flow rates up to 10 mL min–1 (0. Cu: 1. respectively. Cu: 324./Oct.8 mm i. above which they became constant. (min) (oC) Hold (min) Rea. Pb: 1.0 mL min–1 . and ethanol (E) on the absorbance was carried out with an additional peristaltic pump.8 nm. PC: preconcentration column (50 mm x 2. Cu: 5. 181 . S: sample or reference solutions (Cd: 3. All digests were clear solutions and the final volume was made up to 50-mL volume with Milli-Q water. W: waste. 1. Cu. The faster the rotation speed of the peristaltic pump. IC: injector-commutator.2–2 mL min–1.3 nm). FAAS: spectrometer (Cd: 228.Vol. As the system was designed on a time-based injection mode. RC: reaction coil (200 mm x 0. one channel was varied while the others were unaltered. 2003 placed into acid-cleaned vessels and 10 mL of concentrated nitric acid added. the influence of loading time on absorbance was studied within the 1–10 minute range. These flow rates were simultaneously changed in order to eliminate dilution effects.5 mL min–1). The flasks were placed in the microwave oven cavity and the heating program as listed in Table I was run.4 mL min–1. the height of the analytical signals for Cu increased about 10 times for C18 and 5 times for PE when the sample flow rate was varied from 1 to 10 mL min–1 and from 1 to 7 mL min–1. With this strategy. No significant measured peak heights for 50 µg L–1 Cu were observed for flow rates < 1 mL min–1 using C18 or PE. E: ethanol (6. and Pb determinations. ADTP solution (R). the higher the peak height.6 mL min–1). R: ADTP solution (Cd: 0.0 mL min–1) .1 5 70 70 95 5 0. Flow diagram of the system at the sampling (A) and elution (B) positions for Cd.0 mL min–1.).Volgent ume (mL) RESULTS AND DISCUSSION FIA System Optimization The FIA system as depicted in Figure 1 was studied in relation to chemical and flow variables in order to obtain optimum analytical conditions.7 mm i. The rotation speed of the peristaltic pump was varied to produce different flow rates of S and R within the range of 1–10 and 0. Fig. Pb: 283.d.

000 µg L–1 Pb. For Cu and Pb determination with PE. In general. the higher the peak height up to 4 mL min–1 for both C18 (0.09 10–4 + 5.64 10–4 µg L–1 (Pb) With direct aspiration of the analytical solutions containing 10–2000 µg L–1 Cd. It should be pointed out that the flow rate of sample loading could be adjusted to 10 or 7 mL min–1 if a higher sensitivity were to be required.30 10–5 + 3. and Pb were obtained by submitting the calibration solutions to the proposed flow system shown in Figure 1. both obtained with reference solutions prepared in 0.12 10–3 + 1. and system stability. and Pb determination with the PE and C18 mini-column. The enrichment factor (EF) was obtained as the ratio of the slopes of the calibration curves with and without preconcentration.26 10–6 µg L–1 . ADTP Concentration The influence of ADTP concentration on the absorbance of Cd. and 50–20.05 10–-3 µg L–1 (Cd) A = 3. The influence of sorbent on sensitivity was evaluated by means of the analytical curves in the 5–150 µg L–1 Cd. With the main variables of the procedure optimized for each sorbent material. When the nebulizer uptake rate (5 mL min–1) exceeded the elution rate.03 A) and PE (0. 0. The flow rate of sample loading is an important parameter for the Cd determination in the proposed flow system with a C18 or PE minicolumn.47 10–3 µg L–1 (Cd) A = 6. the ratio of measured absorbance from C18 and PE extractions in the same flow rate were ~ 5.05 10–3 µg L–1 (Cu) A = -6. A 20-cm coil length was selected for further experiments. Flow rates higher than 10 mL min–1 caused high hydrodynamic impedance of the flow system. the signals increased with ADTP up to 0. the selected flow rate of sample loading for Cu determination using C18 and PE was 5 mL min–1. Since higher flow rates did not significantly alter the measured signals.17 A) for PE and C18. the correlations between absorbance (A) and analyte concentrations were as follows: For Cd: A = 1. it was found that the faster the rotation speed of the peristaltic pump. the loading flow rate selected for Pb determination using C18 and PE was 3.05% (m/v) (0. and Pb is shown in Figure 2. When a 100-µg L–1 Pb sample solution was processed in the flow system with a C18 or PE minicolumn and a sample carrier stream of < 1 mL min–1. calibration curves for Cd.34 10–4 + 7. The length of the reaction coil (RC) was investigated in the 10–50 cm range.48 10–3 + 1. Analytical Features Table II shows the analytical characteristics of the flow system for Cd. the higher the peak height up to 3 mL min–1 using C18 (0.28 10–4 + 1. which resulted in solution leaking through the connecting points or confluences. As a compromise between sensitivity and system stability.26 10–3 + 1. A 0.22 A) and 0. In general. respectively.006 A for PE). vacuum segments or gas bubbles formed inside the capillary that connects the FIA injector with the nebulizer and no repetitive transient signals were recorded.017 A).5 mL min–1 was selected for C18 and PE. Cu. It was found that the faster the rotation speed of the peristaltic pump. resulting in an absorbance similar to that of the blank solution (0.74 10–5 µg L–1 For Pb: A = 1. the ratio of measured absorbance from C18 and PE extractions in the same flow rate was ~ 2. The signals increased steeply with coil lengths up to 20 cm. 50–5000 µg L–1 Cu.34 10–4 µg L–1 For Cu: A = 2. For a 50-µg L–1 Cd solution. For Cd determinations. above which they became constant. As a compromise with regard to sensitivity. resulting in absorbance similar to 1 mL min–1 (0.60 10–4 µg L–1 (Pb) Using C18: A = 4. The correlations between absorbance (A) and analyte concentrations were as follows: Using PE: A = 3. The signals increased by ~ 40% when the flow rates increased from 5 to 7 mL min–1 for C18 and PE.tively. then the signals decreased steeply up to 10 mL min–1. Flow rates of < 5 mL min–1 resulted in a very unstable system.45 10–4 + 1.6 mL min–1.05% (m/v) ADTP solution was sufficient for Pb and Cu determination with C18. Another parameter studied was the flow rate of the ethanol carrier stream. Cu. Measurements in the 1–4 mL min–1 range revealed that the absorbance of the Cd solutions extracted using PE was ~ 2 times that obtained with C18. and 10–250 µg L–1 Cu ranges.25% (m/v) (0. The study showed that diluted ethanol cannot be used as the eluent because unsatisfactory signals were observed due to poor elution.97 10–3 + 2. sample consumption. Cu. 25–150 µg L–1 Pb. The flow rate of the carrier stream selected for the Cd determinations with C18 and PE mini-column was 3 mL min–1. above which they became constant. Higher ligand concentrations did not 182 alter the absorbance.075 A) and PE (0.1% (m/v).003 A). the eluent (E) carrier stream of 6. no significant absorbance was observed. the signals increased with an ADTP concentration up to 0. Ethanol concentration is another important parameter. above which they became practically constant. then the signals decreased steeply up to 10 mL min–1. The flow rate of the eluent was varied from 3 to 8 mL min–1.14 mo1 L–1 HNO3.91 10–3 µg L–1 (Cu) A = 3.16 A).018 A for C18.

Fig.0 0.4 1. and 100 µg L–1 Pb using C18 and PE. 22. Additionally. The limit of detection (LOD) defined as the concentration of metal with a response equivalent to three times the standard deviation (s) of the blank (n=12) was found to be 1.5 6. and Pb. Cu. 2%.9 19 PE LOD (µg L–1) EF 23 18 22 C0 = characteristic concentration.0 2.0 µg L–1 Cd using C18. 2. and 1. Cu.0 1.0 2. the recoveries for the spiked samples were within 96–99% for Cd. The samples were mineralized in triplicate together with blanks which allowed for the contribution of analytes present in the reagents in the digestion procedures. the sampling rate was ~ 40 h–1.6 Sample flow rate (mL min–1) Ethanol flow rate (mL min–1) 6.6 11 Mini-column EF 11 33 36 C0 (µg L–1) 1.5 ADTP flow rate (mL min–1) 0. 11 µg L–1 Pb. TABLE II Analytical Characteristics of the Proposed FIA System Using PE or C18 Cd Cu Pb ADTP % (m/v) 1.0 µg L–1 Cd using PE. Determination of Cd. and 23 times for C18 and PE. 183 . and Pb determination with the flow system consisting of the PE or C18 mini-column. respectively.1 3. and the results were compared with those obtained with microwave-assisted digestion.4 4. LOD= limit of detection. 97–102% for Cu. respectively. Six table wine samples were analyzed in order to investigate the performance of C18 and PE extractors.3 17 C18 LOD (µg L–1) 2.6 1. C18–Cd (B). Sept. Cu. C18–Pb (C). In addition. 2003 For a 60-sec preconcentration time. According to a paired t-test. C18–Cu (D). 24(5).5 6./Oct.0 5. all results were in agreement at the 95% confidence level. Curves correspond to the following solid extractor–analyte pairs: PE–Cd (A).2 27 1. EF= enrichment factor.0 3. Cu. and 6% for Cd.9 µg L–1 Cu.1 0. and 90–99% for Pb using C18 or PE. PE–Cu (E). and 2. and 2. Pb. and Pb in Wine Table III shows the results of the Cd. and Cd calculated from absorbance ratios using curves with and without preconcentration were 33.6 µg L–1 Cu. 19 µg L–1 Pb.0 5–150 10–250 25–150 Calibration (µg L–1) Preconcentration time (s) 60 60 60 Sampling frequency (h–1) 42 40 40 –1) C0 (µg L 3. the preconcentration data revealed that PE was the best sorbent for Cd while C18 was best for Cu and Pb. The relative standard deviations for most samples (n=12) were lower than 2%. 36. Effect of the ADTP concentration on absorbances of 50 µg L–1 Cd. The enrichment factors for Cu. and 11 times and 18.Vol. and PE–Pb (F).

0 <2. P. and M. Bode. 5. 201 (1996). Analysis of the recovery data for the analytes added to the original wine samples revealed that there were no significant analyte losses in both digestion procedures. W. O. Chem. Anal. Adams. Spectrom. and R. 215 (1995).0 <1. M. 1249 (1999).P.H. Arnswald. Anal. 14. J. J. and A. and 0. respectively. Acta 285.0 PE Cu C18 244(2. 32. Anal. Anal. 8. Curtius. Curtius. Cu. 343. Chim. K.7)* 226(1)* 217(2. Gomes Neto. J. B. and Pb in wines since the limits of quantification (LOQ) obtained for Cd. C.2)* 26. R. 4. Hansen.0 <2. At. R. Curtius. Talanta 38. Spectrom.6(0)* <11 <11 <11 22. W. 3rd edition.7)* 220(0)* 241(1)* 247(1)* PE 243(0)* 225(1)* 216(0)* 219(1)* 240(1)* 244(1)* Pb C18 20.0 <19 <19 <19 <19 23(0)* <11 <11 <11 22(0)* 26(0)* <1. 613 (1991). M. and F. J. 17. Z-L. R. Van Mol. M.8(1. 11. and F.5)* 21. W. M. 13. Monte and A. F. 243 (1989). Van Mol. 181.F. A.0 <1.0(1. D. V. Anal. The recovery data of the analytes added to the sample digests showed that there are no significant matrix effects caused by the original matrix components or the reagents. Ma and F. 12. Fresenius’ J. 185. 1917 (1996). Chem. Spectrom. Van Mol. Pozebon. Analyst 124. López. X. 179 (1962). Chim. Welz. Ruzicka and E. Chim. 12. Total Environ. Cu. Dressler.0 <2. Cu. At. Anal. W. Lett. Yin.0 <1. R. Acta 317. Fang. J. 941 (1999). V. 498 (1994). Anal. Adams. Acta 309. UK (1995). while the PE extractor was more appropriate for Cd-DTP determination. and F.Giacomelli. J. A.0 <2. Welz. 13. Cabrera. ACKNOWLEDGMENTS The authors thank the FAPESP (Projects 99/10486-9 and 00/148877) for financially supporting this work. Welz and M.0 PE 245(1)* 230(1)* 216(0)* 222(1)* 242(1)* 245(1)* Cu C18 244(1)* 228(0)* 216(0)* 221(0)* 242(1)* 248(1)* PE Pb C18 PE 19. 9. and the CNPq for fellowships to J. and Pb exceeded the requirements of the Brazilian Food Regulation (Decree No. At. Ma.0 <1. C Campos.1)* 17. 176 (1996).0 <1. Anal. Ruzicka and A. Wiley & Sons: New York. 21. 2091. L. 7. 363 (1998). CONCLUSION This work presents a simple. Z. pg.A. Weinheim.0 <2. Flow Injection Atomic.3)* *S. C. A.50 mg L–1. Ma. J. and A. J. Curtius. Acta Part B 51. L. Adams. J. Germany. 184 .Sample PE 1 2 3 4 5 6 <1. J. M. Adams. 18. V. and A.. and B. Arnswal.3(1. Sella. and Pb Determination in Wine Samples Analyzed by the Proposed FIA System Using PE and C18 With Digestion in Open Systems FIA-FAAS Microwave Oven Digestion (µg L–1) Digestion Block (µg L–1) Cd C18 <2. 754 (1992). Ma.0 Cd C18 <2. Adams.0)* <1. B.0 <1. Anal. 5. Mena. Acta 216. Van Mol. Adams. of Anal. da Silva. Flow Injection Analysis. Spectrom. John Wiley & Sons. 21 (1990). 395 (1995). The on-line solid phase extraction using C18 or PE can be used to determine Cd. Ma.6(1. 3. Acta 293. Adams.8(1. Chim. 33 (1994). J. 2. A.N.0 <1. 6. and F. Sci. Spectrosc. 10. K. REFERENCES 1. R. Yan and F. 459 (1997).50 mg L–1. D.0 <2. Ávila. T. M. Fresenius’ Z. Cu. and Pb in table wines by FAAS using on-line preconcentration with C18 or PE. Atomic Absorption Spectrometry. C.G. Avila and A. L.0 TABLE III Results (n=3) for Cd. Fang. and accurate method for the flow injection determination of Cd. and Pb at 0. 9.3)* <19 <19 <19 <19 22. Curtius. 15. H. 251 (1994). Received November 17. A. and B. fast. Dressler. These studies demonstrated that the C18 extractor was more efficient to preconcentrate Cu-DTP and PbDTP complexes and determine both elements in addition to providing the highest enrichment factors.00 mg L–1. Sperling. Lorenzo. Pozebon. Chim. the fellowship to A. 543 (1994). At. Chichester. X. 55871 from the Ministry of Health) which established the maximum permissible levels for Cd. R. 5. Guo. 19. S.9(1. 20.D. 2002.0 <2.0 <2. 16.0 <2.0 <1. W. (1999). Wiley-VCH. 1167 (1998). 17. Sperling. At. Anal. Spectrochim. Anal. Cu. Ma and F.0 <1. Talanta 45. J.0 <2. L. pg.

Cu. and W determination in a tantalum matrix by using X-ray fluorescence spectrometry. IDA. Prakash Centre for Materials for Electronics Technology (C-MET). Calcium and magnesium are suspected to increase the DC leakage of solid tantalum capacitors. B.0-m monochromator ICP optical emission spectrometer (13). the Zr. Nb. using a 0. Mo. This analytical work is very vital in regard to tantalum capacitor performance and applications. and W were determined in the tantalum matrix by anion exchange coupled with ICP-MS (12). Cu. Co. Addition of molybdenum also has a very pronounced effect in decreasing the susceptibility of pure tantalum metal towards hydrogen embitterment in severe corrosion conditions. *M. Transitional elements are known to increase high field leakage of anodic tantalum oxide films. and during recycling of tantalum products. The effects of tantalum in the determination of trace level impurities were studied by determining these elements in solutions containing tantalum from 1000 to 20. we reported the matrix effects of tantalum in the determination of the trace impuri185 *Corresponding author.11). and isotope dilution analysis with a multi-collector coupled with ICP-MS (6-8). The presence of boron up to 100 mg/L is acceptable in overcoming the DC leakage problem. Co. Anil.64m monochromator inductively coupled plasma spectrometer. ICP-OES analysis of tantalum powder is routinely carried out and many matrix interferences are found during trace element determination. sodium reduction. during micro alloying. B. e-mail: mrpreddy@yahoo. Earlier works by other researchers describe the determination of Zr in a tantalum matrix using trace matrix separation by anion exchange. September/October 2003 . Cu.2). Mo. As the concentration of tantalum is increased. and the hydriding and dehydriding process. W. (9) reported Zr. Mg. and Ti present in ore as well added during the reduction stage (5). and Al are added during the recycling process of solid tantalum capacitor scrap (3. there was no change in intensity. Mo. Micro alloying of zirconium and tungsten with a maximum of up to 1000 mg/L with ties of Cr. Mo. while Ca and Mg are added during chemical processing such as deoxidation.000 mg/L tantalum solutions. Co. and T. extraction chromatography. Mn.000 mg/L. Mg. Hyderabad – 500 051. Spectral line profiles of the trace elements and spectral overlaps of tantalum at 1000 mg/L for interference-free line selection is presented and discussed. Al. W. etc. Ni. The elements Zr. Co. the trace elements Zr. Shuji (14) determined boron using on-line anion exchange combined with ICP-MS. The dopant usually involves compounds containing boron. Mo. The presence of even trace amounts of certain elements in tantalum powder and wires can adversely affect the electrical properties of solid tantalum capacitors.L.4) Atomic Spectroscopy Vol. The presence of just a few atomic percent of oxygen in tantalum increases hardness. (15) describe the determination of tungsten in niobium tantalite present in tin slag using derivative spectropho- ABSTRACT In this study. Mn. is actively involved in the preparation of low and high voltage capacitor-grade tantalum powder and lead wires using solvent extraction. For these reasons. Phase-II. Mo. Hyderabad. 24(5). and B are doped in order to increase the capacitance and life of tantalum products.P. and Al are contaminants added during recycling of solid tantalum capacitors. In a previous work. Cu. and Al decreases in intensity counts while Cu and Mn showed an increase in intensity counts. HCL Post Cherlapalli. tensile strength. +1 states showed an increase in intensity and +2 states showed a decrease. boron amounts above 100 mg/L cause tantalum lead wires to harden and become brittle during the sintering process (1.Tantalum Effects on the ICP-OES Determination of Trace Elements in Tantalum Powder G. For Ca and Mg. Mn. Knote et al. while other studies report the direct determination of Cu and Mg by ICP-OES in potassium tantalum lithium niobates and by electrothermal atomic absorbance spectrometry in tantalum powder (10. Co. and Al). Mn. and elasticity but decreases ductility and corrosion resistance. W. Arbind Kumar. Molybdenum is added to tantalum foil to modify the etch characteristics in the manufacture of etched foil capacitors. India INTRODUCTION C-MET. The relative standard intensities of trace elements in the tantalum matrix were calculated from 1000 to 20. Most high-charge tantalum powders contain one or more dopants which are added to retard sintering. Ca. In this study we extended to include 10 more elements (Zr. Padma Subashini et al. W. and Al are determined in tantalum powder. In the case of boron. ConteRosa (13) reported the direct determination of several elements present in high concentrations in tantalum products using a 1. tantalum can minimize the deleterious effects of oxygen added during processing and extend the useful life of tantalum products. Reddy. These traces are added to tantalum powder during the addition of dopants. Ca. Fe.

959-nm line showed no change in intensity.270 nm and 279.847nm line was selected as it had very high sensitivity. All of the above elements were of +2 state (29). The 280.106-nm manganese line was selected and showed a 3% decrease. However. Reagents and Standard Solutions All chemicals used were of Suprapur® grade (Merck. Germany).700nm copper line showed a 12% decrease. Line profiling (28. Further studies (16–25) of tantalum matrix effects on trace elements have been performed and appropriate correction factors were calculated. which tend to overlap lines of the analytes or trace elements. tantalum.29) was done for each trace element. The zirconium line at 339. the 201. prepared with a Nanopure system (Barnstead.tometry and ICP-OES. which are present at high concentrations. In case of molybdenum. The instrumental parameters are given in Table I. The cobalt line at 228. deionized water was used. For all analytical work. Dubuque.199-nm line could not be resolved at this concentration of tantalum. Boron at the 208. Longjumeau Cedex.875nm line showed a 4% decrease.64-m monochromator ICP spectrometer without trace matrix separation by choosing a spectral interferencefree tantalum line for trace element determination.5 L min–1 12 L min–1 est sensitivity and showed only a 7% increase in intensity with respect to 1 mg/L.000 mg/L. The best lines for each element were selected by taking into account the matrix effects and the sensitivity of the line at 20.616 nm was selected as it had the highest sensitivity and showed a 4% increase in intensity. calcium at the line of 422. The singleelement standard solutions were also procured from Merck. adjustable Dynamic range 2.198 nm was selected as it had the high- 186 .673-nm +1 state (29) showed a 4% increase in intensity.5 x 1010 Measures complete emission spectrum in 2 min (230. TABLE I ICP-OES Operating Conditions Sequential spectrometer Grating Spectral range Slit width HDD Detector Image Jobin Yvon JY-24R. For calcium.64-m Czerny Turner Holographic. Experimental Procedure Tantalum powder was dissolved in a HF + HNO3 mixture and appropriate concentrations (ranging from 1000 to 20.511-nm line was selected as it showed a 6% increase in intensity. 0. In the present work. Profiles of a 1-mg/L trace element standard were plotted and overlapped with 1000 mg/L tantalum.553 nm +2 state (29) showed an 8% decrease. Trace element solutions of 1 mg/L were added to the tantalum solution. To overcome analytical bias due to possible contamination from matrices. the 396.27) in the selection of analyte peaks are coincident line overlaps. IA. Complex spectra are most troublesome in the analysis when produced by the major constituent. i. The detailed results of trace elements added to different tantalum concentrations are given in Table III. the results are listed in Table II. a wing overlap from a much more intense line nearby.000 mg/L of Ta.826 nm +1 state (29) showed a 46% increase and the lines at 280.270 nm Plasma torch assembly Nebulizer and spray chamber Argon coolant gas flow rate Argon plasma gas flow rate RESULTS AND DISCUSSION Matrix-free Line Selection The spectral interferences commonly found (26.000 mg/L) were prepared. USA).84-nm +2 state (29) showed a 6% decrease in intensity. The aluminum line at 394.. the trace elements were directly determined using a 0. while the 344. 3600 grooves mm–1 160–500 nm 20 nm for entrance and exits. This is due to the spectral lines from tantalum. France). The 224. The magnesium line at 383.e. while the line at 396. The line at 280. the matrix solution was analyzed as a blank for all trace element studies. Darmstadt. The interference-free lines were checked for matrix effects of tantalum by doping 1 mg/L standard into concentrations of tantalum ranging from 1000 mg/L to 20. EXPERIMENTAL Instrumentation The instrument used for the determination of trace elements in tantalum powder was a Model JY24R inductively coupled plasma optical emission spectrometer (Jobin Yvon.000 points) Fused quartz with capillary injection Cross-flow type nebulizer and Teflon® dual-tube spray chamber 1. A cross-flow type nebulizer was used for pneumatic nebulization. tungsten at the 224.401 nm +1 state (29) showed a 17% increase in intensity.

the solution was made up to 100mL volume (i. 2003 TABLE II Spectral Interference of Tantalum on Trace Element Wavelengths (nm) (1 mg/L of trace element line overlap with 1000 mg/L of tantalum) Wavelengths for: (Zr ) 256.006c 226.826b ---------– a (Co) 228.807 202.013d 272.106b 293.270b 285.5 mg/L trace element solution.G.346c 238. From these experiments it can be seen that the experimental values.261b 273.778c 223.139c 257.503c 205. Left wing overlap of tantalum line.198b 343.366b 396.115b 349.875e 232.887d 257. Germany) dissolved by using 3 mL HF.486b 274.983c 279.773c – – – – – – – – – --– – – – c (Al) 221.628e 339.705c 237. Mg.660b 357. The results are presented in Table IV.379a 237.582 279.569c 279.887 317.923c --– c (Ca) 315. was selected for magnesium as it had high sensitivity.475c 209.339c -------– b (Cu) 213.823d 348. B./Oct.827c 280.993b 393. K.Vol. equipped with a 0. a 2-g/100-mL tantalum powder was spiked with pure standard solutions of 0. Mn.597b 327..754c 327.213b 293.553b 280.764 256.609c 239.414c 277.678e 249.935c 220.673b -------------– e (B) 208.199b 403.160c 231.307c c (Mg) 202.448c 222.844c 204.894c 219.739c 235.482c 279.256b 316.215c 309.396c --------– a (Mn) 257. No tantalum line overlap present in the vicinity of ±0.958c 221.312e 237.458c 221.876c 268.306c 293. Comparison of Results With a Certified Powder and Spike Recovery The method developed was verified by analyzing 2 g of certified tantalum powder STA-18KT (H.998c 203.615b 284. when multiplied using the matrix factor. 24(5).930c 294.305c ----– d (W) 200.862e 238. the % recovery was calculated with and without a matrix factor.893 208. showed very close proximity to the certified values as well as to the recovery results of the spiked analytes.426e 234. Ca.25 mg/L and 0.911a 208. This method and the matrix factors can be applied for the ICP-OES determination of trace elements in tantalum powder.709c 248.567c 355.008c 224. Co.C.e.373c 260.636c 238.511 202.284e 394. 20.405c 231.892c 327.498e 234.05 nm.823c 287. This factor was applied to the analysis of certified materials and was found to be very suitable. 1 mL HNO3.271e 309. Right wing overlap of tantalum line. W.151c 292. Cu.000 mg/L of tantalum).959b 249. The interference-free lines selected can be used for the determination of these traces in other tantalum- 187 .610 259.616 230.401b 396.482c 277.247c a b c d e a (Mo) 201.922c 236.863c 270. Direct tantalum line overlap.926c 330. Starck GmbH & Co.860c 216..910c 226. Similarly..076c 403.000 mg/L tantalum.654d 383.786a 231.64-m monochromator.589c 224.598c 263.030c 203.621c 350. was successfully applied for the direct determination of the trace elements Zr.335e 257.510c 308. and Al in tantalum powder without the need for trace matrix separation by selecting tantalum spectral interference-free lines and calculating the matrix effects factor of tantalum at different concentrations of tantalum ranging from 1000 mg/L to 20.346c 226.152c -----– – = = = = = Right and left wing overlap of tantalum line. Mo.819c 209.231c 383.342d 236.847b 422. CONCLUSION The ICP optical emission spectrometer.632c 218. The results are presented in Table V.598 217.700d 324. and 5 mL of water.468c 207.540c 281. The selected lines of trace elements were used for the analysis by calculating the matrix factor as described in a previous work (5).920c 344.305c 327.810c 222. Sept.

875 317.0088 CR 0.0± 0.0± 0.0076 1.198 201.0065 0.0± 0.46± 0. 5.01 ± .000 mg/L (S.96± 0. 5.0092 1. 4.0073 1.01 ±0.078 0.0072 1.No 1.0075 CR 1.93 = 1.0099 1.95± 0.0081 1.0± 0.0093 1.0089 0.700 280. 16.98± 0.04± 0.D) 10.03± 0. 17.0065 1.0085 CR 1.091 1.0± 0.772 1.0098 1.0074 CR: Cannot be resolved.0029 1.46± 0.915 1.0 ± 0.959 394.426 235.087 = 9.0± 0.826 228.0± 0. Element Wavelength 1.0079 1.68 0.98± 0.0082 1.90 x 0.0096 1.97± 0.98± 0. Elements Zr Mo Cu Mn Mg Co W Ca B Al TABLE IV.0055 1.401 <2 <2 <1 <2 9 N. 22.0081 0. 25.06 = 2.0068 0.36 x 1.0± 0.616 231.0095 1. 2.80 1.0± 0.0065 1.99± 0.0± 0.213 383.0085 CR 1.0± 0.342 237.35 x 0.TABLE III.0081 1.0075 1.0092 1.0084 0.0± 0.19 x 0.336 5.115 355.0075 1.0± 0.498 234.0077 1.0078 1.256 316.0094 1.90 2.615 224.68 x 1. 19. S. S.12± 0.01± 0. 28.0087 1.0078 1.0± 0.02± 0.02± 0.0084 1.0± 0.0± 0.08 0.0053 0.0± 0.46± 0. 4.261 273.0096 1.0± 0.0± 0.061 1.0055 1.0098 1.847 208.0± 0.0018 1.0± 0.02± 0.010 1.0060 1.0± 0.0077 0.0± 0. Analysis of Certified Powder (Results in mg/L) Wavelength Certified Values Experimental Results x Matrix factor (STA-18KT) (STA-18KT) 339.0± 0.0072 1.R: Values not reported.0084 1.31± 0. 9.0± 0.04± 0.0071 1.95± 0.0054 CR 0.98± 0.0066 CR 1.0086 0.993 393.0075 1.0066 CR 1. 188 .000 mg/L (S.10 x 0.97± 0.94± 0.0071 CR CR 0.99± 0.0071 1.0066 1.0065 1.0095 1.198 348.0± 0.0± 0.0± 0.0077 0.0± 0.13 = 0.0± 0.0± 0.0087 1.07± 0.D) 20.0077 0.0093 1.0063 0.07± 0.67 8.0± 0.0072 CR 0.511 224.700 280. 14.96± 0.0096 1.0± 0.0± 0.0079 1.97± 0.0± 0.0087 1.0083 1.673 208.0± 0. 8.98± 0.0± 0.0064 1.0078 1.000 mg/L (S. 27.0018 1.03 = 1.0068 0.0094 0.0095 1. 9.0091 1.0075 0.0088 1.0± 0.D: Standard deviation of five readings.R <5 <3 <1 1 2.959 394.0± 0.94± 0.0095 1.18± 0.0058 1.98± 0.366 396.04± 0.0074 CR 1.0095 0.01± 0.786 0.0065 0.0076 1. 3.847 422.270 285.0± 0.0066 1.0068 1.0091 1.0061 CR 1.0082 1.0± 0.0044 0.0076 CR 1. 10.0075 1.0± 0.No.0± 0.0± 0.0071 1.98± 0.98± 0.D) 1.01± 0.05 x 0.0062 1.97± 0.0± 0.01± 0.52 x 1.0069 1.553 280.04 ±0.0065 CR CR 0.0078 1.97± 0. 3.0± 0.660 201.0087 1.46± 0.0082 1.D) 20.0078 0.0083 0.D) 5.94± 0. 6.90± 0.0095 1.0± 0. 26.96 = 0.85 = 1.0064 1. 10. 6.0078 1.0± 0.92± 0.0094 1. 18.0081 CR 0.0± 0.106 344.106 280.0± 0.616 224.80 x 1 = 0.92± 0. 20.511 281.01 N.486 274.09± 0.0084 0.17± 0.0± 0.98± 0.0075 0.000 mg/L (S.0072 1.62 x 1.0067 1.0076 0.0085 0. 7.04± 0.597 339.270 228. 24. 21.99± 0. 11. 8. Zr Zr Zr Zr Zr Zr Zr Mo Mo Cu Mn Mn Mg Mg Mg Mg Co Co Co Co Co W Ca Ca Ca Ca B Al 272. 23.06± 0.22± 0. 12.0± 0.0088 1.88± 0. Trace Elements (1 mg/L) Added to Different Ta Concentrations S.0086 1.04± 0.0035 1. 7.000 mg/L (S.06± 0.98± 0.0093 1.0065 1.02± 0.96± 0.06± 0.0093 1.92± 0.97± 0.99± 0.0082 1. 13.875 396. 15.0 ± 0. 2.0± 0.05± 0.199 279.0059 0.0061 0.862 224.46± 0.94 = 1.401 0.0088 1.02± 0.0075 1.0± 0.96 = 4.

3.08 SD = Standard Deviation of five readings.28±0. Proceedings of International Symposium of Tantalum and Niobium. 5. Yang and C. Lilli Paama. Inc. IICT.15 106. 5.06 100. Takahashi Shigeto. Mare Must. 8. 20.K.16 100.18±0.02±0. Part B.08 99. ACKNOWLEDGMENTS The authors would like to thank the Department of Information Technology (DIT).R. 1858 (1982). Analyst 127.17±0. J. Director. Joaquim Nobrega. Senior Director. and L. 271 (1988).19 96.12 100. 14. Padma Subashini.Wuensch. O.K. Kozono. Harald Knote and Viliam Krivan. (1992). Sethi.28± 0. Jean-Michel Mermet. (2000).05±0. Anal. Daskalova.97±0. Edivaldo Garcia. Ana Rita. Anil. At. Acta. V. Bezrukavnikov. Anal.29±0.50 mg/L Trace Elements Spiked in 2 g/100 mL Tantalum Powder 0.09 94. 21. C-MET. Moltelica-Heino. N. 1069 (1998).19±0.18 94.07 100. V. Arbind Kumar and T.511 224. 1339 (2002). 12. At. and Juan Mora. 57. J. R.J. Klaus-Christian Friese and Viliam Krivan. 21.03 100. and A. 28. Opilz. Mermet.19 88. 26. 47.17 106. Jose De Anchieta Rodrigues. Received September 20. Acta.07±0. Gentsheva.20±0.95±0.E. 69 (2000). E. At. Proceedings of International Symposium of Tantalum and Niobium. Prakt.06 100. 44 (1997). G. Huber. Sciences 16.19±0. Chem. 281 (1997). Reddy.25 99. Acta Chim. Spectrom. E.Vol. 825 ( 2001).13 100. I. The authors are also thankful to Dr. Eric Salin. 3. 12.270 228. 16. Akbar Montaser and D. 57. 1215 (1997). Acta. Spectrom. VCH Publishers. B.15±0. Charles Pokross. M. 119 (2002). O. This method of analysis is very fast and accurate and can be applied for routine analysis. J. Element Wavelength Factor calculation (SD) Factor calculation (SD) 1. for their helpful discussions. and S. 1333 (2001).5 mg/L Spiked trace elements Recovery % Recovery % Recovery % Recovery % without matrix with matrix without matrtix with matrix No. Anal. 16. R. 23. Das. 10. 153 (2000). and H./Oct. Talanta 50. 14. and Paavo Peramaki. Spectrochim.847 208. Miguel Catasus. 29. 9. Executive Director. Spectrom. 13. 23. Spectrosc. Emmanuelle Poussel. Aleksieva.98±0. 930 (2002).06±0.04 117.10 97. Chim. Part B. Optical Emission Lines of the Elements.21± 0. 2.Velichkov. Vicente Hernandis. Daskalova. Agranat. X. Chandrasekar. Margarita Villaneuva. and J. 13. Sai. and Roberto Frache. 255 (1988). Chemical Geology. 89 (2000).07 100. J. Bojan Budic. Acta 458. 297 (1988).17 88. Anal.C.700 280.08±0.29±0. 18. 419 (1998). K. Anal.M.K.875 396. and Eva Curdova. At. 27.07 107.04 100.19 100. 2002. B. J.616 224. 675 (1999).96±0. Anal.11 99. L. ConteRosa. Jose Todli. Meinhard Aits. 339. 1351 (2002). Segal. 57 (2002). A. Spectrom. 6. 54. Anal. 53.25 mg/L Spiked trace elements 0.H. Anal. 46 (2000). Anna Krejcova.W. Ganguly. Nougueira. Tripp Terrance.04 99.22±0. 2. Spectrosc. 19. 7.15 96. M. 16.S.88±0. Spectrom. Haraguchi. Anal. Malhotra.198 201. S. At. Recovery of 0. Mark Rehkamper. and Jean-Louis Dimartino. REFERENCES 1.P. New Delhi.25±0. Sept.03±0. John Wiley and Sons Ltd.82±0.02 100. At. and S. K. At. Stefan Weyer. Dr. 375 (2002). and Dr. 25. Spectrom.89±0. At. Kozono Shuji. Anal. J. At. 295 (2002). M. Prakash. Spectrom.15 92. 4. Satyanarayana.M.02±0. G.18 92. K. 24(5). 17.34±0.38± 0. 189 . J.Yoffeo.02±0. 15. M. 11. 877 (2000).97±0. and Mario Pomares. Dr. Emanuele Magi. Jean-Michel Mermet. DIT. Spectrom. related compounds using ICP-OES.97±0. Spectrochim. 9. and Klaus Metzger.06 99. Carsten Munker.33± 0.05 99.18±0. Spectrochim. At. At. and R.26 99. J. Second Edition. 16. Bergman. Jesus Cano. Pinc.12±0. and Haraguchi Hiroki. Anal.X.06 99.15 104. Cendrine Dubuisson. 669 (1999). Inductively Coupled Plasmas in Analytical Spectrometry.33±0. Proceedings of International Symposium of Tantalum and Niobium. J. 22. Velichkov. 2003 TABLE V.106 280.25 mg/L and 0.13 97. M. Tomas Cernohorsky.07 99.Donard.08 99. 6. A. 12. At.F. 1002 (2001).34±0. A. Part B. Chem.401 107. Proceedings of International Symposium of Tantalum and Niobium. 17.06 100. Marco Grotti. Dorfman. 10. J. Spectrom. J. 15. J.K. 333 (1988).26±0. 24. Spectrom. Chen. for sponsoring the Tantalum Technology Mission Project. G. I.97±0. Eed Parnoja. 7. Anal. 15. N.14 104. Slov. and A. Anal. E.J. 4. and Jean Michel Mermet.L. Zirconium Molybdenum Copper Manganese Magnesium Cobalt Tungsten Calcium Boron Aluminum 339. 8. L. Richard Payling and Peter Larkins. Takashi. 187.12±0. Golightly. Scientist.07 117.959 394. Government of India.

It was found that NEIE showed an increase and decrease in intensity in the range of 10% only. where matrix interferences are noted during trace element determinations. England.R. so that any shortcomings existing in atomic emission spectrometric analysis can be lowered or eliminated. the effect of trace elements by NEIE. The deionized water used was prepared using the Nanopure system (Barnstead. IA. A cross-flow nebulizer was used for nebulization. Hyderabad – 500 051. 24(5). France ). Co. a grasp of the broad aspects is meaningful because it bears on the conduct of practical analysis. Studies on the matrix interference effects have been conducted by several research groups (1–9).P. Germany). There is a *Corresponding author. Cu. e-mail: mrpreddy@yahoo. India INTRODUCTION Despite the intensive work for the development of inductively coupled plasma optical emission spectrometry (ICP-OES) techniques. particularly from the point of their control. whereas EIE showed a decrease up to 40%. and the mixed effects of both elements on the analytical intensity of trace elements. The easily ionizable element potassium causes larger effects than tantalum. the matrix solution was analyzed as a blank for all trace element studies. the fundamental processes that control analyte emission intensity and spatial dependence are not fully understood. which has a higher ionization potential. Fe. observation heights of 14 to 18 mm above the load coil. and due to potassium. 0. The study of the processes occurring within the plasma (such as desolvation. Analysis of K2TaF7 is routinely carried out using a 0. Phase-II. September/October 2003 190 . The purity of K2TaF7 is extremely critical for the preparation of tantalum powder. the intensity of all traces started decreasing. and also discussed the spectral interference of Ta and Nb. Middlesex.Mixed Matrix Effect of Easily and Non-easily Ionizable Elements in the ICP-OES Analysis of Trace Impurities in Potassium Tantalum Fluoride G. (11) determined traces of Cu. EIE. atom/ion formation and excitation) is important. 0. The effects are rather complicated and far from fully characterized.1994 g of potassium and 0. HCL Post Cherlapalli. which showed an increase. Presently our laboratory is working on the different process parameters for the preparation of K2TaF7 by acid washings and the recrystallization process. Li. and the mixed effect of NEIE and EIE. Reddy.25. Anil. However. and V in potassium lithium tantalate niobate crystals using ICP-OES. Dubuque. Nb. and the mixed effects (10) of NEIE and EIE. Almost any matrix element can cause a change in the sensitivity of a susceptible spectral line. but the matrix elements studied have been mostly restricted to alkali metals (EIE).75. except for Ca and Mg. Ca. The single-element standard solutions were obtained from Glen Spectra Reference Materials. Ti. the mixed matrix effect showed an initial increase for all trace elements. hydriding and dehydriding routes. at least at Atomic Spectroscopy Vol. Darmstadt. Nevertheless. Prakash Center for Materials for Electronics Technology (C-MET). The aim of this work was to investigate the effects of potassium. a non-easily ionizable element (NEIE).com ABSTRACT The determination of trace impurities (B. In this paper. an easily ionizable element (EIE). vaporization. Cr. *M. C-MET in Hyderabad is actively involved in the preparation of low and high voltage capacitor-grade tantalum powder through solvent extraction. The instrument parameters are given in Table I. sodium reduction. Ni.64-m monochromator ICP-OES.4614 g of tantalum by weight. EIE. Reagents and Samples The acids used were of Suprapur“ reagent grade (Merck. As the concentration of the mixed matrix increased. and 1 mg/L. but not their matrix effects. and Mg) in potassium tantalum fluoride (K2TaF7) by inductively coupled plasma optical emission spectrometry is complicated due to the matrix interference of tantalum on trace impurities. and existing methods for atomic spectrometric analysis optimized. Ti. except for Cu.5. 1 g of K2TaF7 contains 0. Fe. Arbind Kumar. and T. while at the same time estimating their trace element content. was studied with doped trace elements at concentrations of 0. To overcome analytical bias due to possible contamination from the matrices (acids) present at high concentration levels. ICP remains an important field of study. tantalum. Irina Segal et al. Experimental Procedure 2 g of K2TaF7 was dissolved in 10 mL HF. IDA. Longjumeau Cedex. EXPERIMENTAL Instrumentation The instrument used for trace element analysis in K2TaF7 was a Model JY-24R ICP-OES (Jobin Yvon. need to establish a method to determine the effects of NEIE.L. USA).

NEIE. when doped at a concentration of 1 mg/L into tantalum solutions [tantalum interference-free lines were selected (15-16)] with respect to K2TaF7 weight. 24(5). using interference-free lines. The trace elements B. and Mg showed an increase in intensity. the intensity of the traces further decreased. the "V" will move downward. This type of correction should be considered when a proper reference material for such an intermediate compound is not available. a matrix factor (15) was calculated and applied to the determination of trace elements doped at concentrations of 0.5 g. CONCLUSION This study demonstrated that trace element determinations in K2TaF7 by ICP-OES. Hobbs and Olesik (12–14) stated that the extent of EIE matrix effects on ionization and ion excitation was more severe near an incompletely desolvated droplet or vaporizing particle than away from a droplet or particle. Ti. The results are given in Table II. 2002. Das. Potassium. and 2. Executive Director. Doping traces at different concentrations in K2TaF7 showed that. it will markedly affect the rate or even the mechanism of solute particle vaporization. the more rapid vaporization of potassium salt.0 mg/L into K2TaF7 2 g/100 mL. its volume increases by diffusion. Dr. the error caused by the matrix was rectified. by applying the matrix correction factor. 191 .50. Using these results. 1. 2003 TABLE I ICP-OES Operating Conditions Sequential spectrometer Grating Spectral range Slit width HDD detector Plasma torch assembly Nebulizer and spray chamber Argon coolant gas flow rate Argon plasma gas flow rate RESULTS AND DISCUSSION The matrix effect. Sai. would yield an overall vaporization process that begins lower in the plasma. 0. ACKNOWLEDGMENTS The authors would like to thank the Department of Information Technology (DIT). The results showed a decrease in intensity for all trace elements. 0. The results are given in Table III. V. except for Ca and Mg. As the concentration of a mixed matrix increased. 0. except Cu. This is consistent with lower temperatures and electron number densities as well as higher analyte and matrix species concentration near an incompletely desolvated drop or a vaporizing particle.C. the intensity of traces started decreasing.5 g. B. of India. Director.75. New Delhi. as the concentration of EIE increased. is defined as the percentage difference in the net line signals between solutions of the analyte with and without the matrix.S. Trace elements of 1 mg/L were doped into potassium solutions containing equivalent weight with respect to 0. Iron and Co showed a decrease in intensity while Ni.64-m Czerny-Turner Holographic. DIT. The addition of 1 mg/L trace elements into mixed solutions of potassium and tantalum showed an increase in intensity for all elements. 3600 grooves mm–1 160–500 nm 20 nm for entrance and exits. is not the complete analysis but the matrix effects of EIE. Cr.K./Oct. Sethi. an EIE. and their mixed effects should be considered thoroughly in order to get accurate and repeatable results. often constitutes the bulk of the solute mass. Received September 20.0 g. M. In turn. Senior Director. adjustable Dynamic range 2. Depressions in analyte ion number densities due to the presence of EIE were. Govt. by a Model JY-24R. The results obtained (see Table V) showed that the matrix factor calculated was very suitable and can be applied for trace determinations in K2TaF7. and Nb.Vol. showed no change in intensity even with an increase of tantalum concentration. Ca. As the vapor cloud appears larger and more dilute. K. which showed an increase (see Table IV). and 1. The authors are also thankful to Dr. or the consequent breakup of solute particles.5 x 1010 Fused quartz with capillary injection Cross-flow nebulizer and Teflon® dual-tube spray chamber 1. Sept. for sponsoring the Tantalum Technology Mission Project.5 L min–1 12 L min–1 factor of two to three. which showed a decrease of 14%. CMET and Dr. Cu.K. The temporally integrated emission will therefore appear to have a "V" shape. As the resulting analytecontaining vapor cloud travels upward through the plasma. 1.0 g/100 mL of K2TaF7.25. more severe near an incompletely desolvated drop or a vaporizing particle than when no droplet or particle was in the observation volume.

and 2.54 87.98±0.61 92.29±0.5 g (SD) 208.54±0.959 100.36±0.59±0.61±0.68±0.28 109.25 113.73 224.90±0. 1.08±0.23±0.24 108.82 79.270 92.76 82.24±0.25 100.24 96.700 89.58±0.50 280.33 77.89±0.74±0.15±0.81 396.37±0. 192 .58±0.121 105.16±0.31±0.69 77.5 g (SD) 2.42±0.50±0.93±0.847 105.19 99. and 2.716 112.68±0.65±0.31 103.63 85.33 100.21 112.32 105.21±0.23 267.58±0.088 91.33 101.58±0.37 101.021 98.700 112.82 60.62 108.71±0.39±0.19±0.959 94.17 89.088 107.45±0.97±0.5 g.44 101.36 76.088 100.616 96.31±0.39 104.5 g (SD) 1.0 g (SD) 1.43 101.35±0.35±0.45 110.29 227.23±0.270 104.39±0. 1.56 66.36±0.38±0.5 g (SD) 1.34±0.47 91.48 96.65 81.21±0.36±0. Elements B Cr Fe Nb Ni Ti Cu Co Ca Mg TABLE III Recovery (%) of 1 mg/L of Trace Elements Doped Into Tantalum Solution (Equivalent to 0.45±0.53±0.54±0.79 227.3 127.77 91.35 62.5 g.97±0. 1 g.40 77.270 101.27 104.56 98.49±0.16 99.89±0.0 g (SD) B 208.27 91.18 103.121 93.716 80. 1 g.20 Cu 224.42 84.97±0.35 99.25±0.021 106.46±0.68 94.87±0.58±0.52 74.5 g.57 280.38±0.37 Cr 267.50±0.35±0.20 267.36 99.47±0.932 121.48±0.49±0.25±0.0 g K2TaF7) Wavelength (nm) 0.84±0.48±0.71 58.26±0.43 ± SD of five readings.42 112.75 295.49±0.28±0.19 336.79 74.21 111.121 103.58 ± SD of five readings.53 274.27 96.39 Nb 295.22±0.616 95.27 107.28±0. and 2.06±0.89±0. TABLE IV Recovery (%) of 1 mg/L of Trace Element Doped Into Mixture of Potassium and Tantalum Solution (Equivalent to 0.78±0.48±0.45±0.49±0.49 88.65±0.58 99.0 g (SD) 1.37 94.85±0.18 99.71±0.64±0.0 g K2TaF7) Wavelength (nm) 0.73 78.0 g (SD) 208.82 Ca 396.37±0.58 113.44 Mg 280.58 82.66 80.95±0.58±0.Elements B Cr Fe Nb Ni Ti Cu Co Ca Mg TABLE II Recovery (%) of 1 mg/L of Trace Elements Doped Into Potassium Solution (Equivalent to 0.20 99.29 103.85 91.23 396.93±0.15 108.0 g (SD) 1.5 g.53 82.61±0.85 336.41±0.64 74.021 108.38 105.52 228. 1.20 99.56 92.63 102.616 113.66±0.87±0.29 92.68 79.69±0.5 g.932 93.19 99.847 90.700 88.41 97.79±0.48 113.31 295.27±0.24±0.15 114.0 g K2TaF7) Elements Wavelength (nm) 0.42±0.32 224.37 68.65 ± SD of five readings.45 123.90 81.36±0.51±0.48 Fe 274.39±0.5 g (SD) 1.23±0.26 Co 228.4 106.5 g.54 74.19 274.27±0.23±0.61±0.21 84.63±0.81±0.5 g (SD) 2.45±0.0 g (SD) 99.45 79.716 100.05±0.42 106.61±0.19 101. 1 g.45±0.65±0.86±0.18±0.56±0.15±0.48±0.13±1.25 Ti 336.46±0.68 228.35±0.64±0.255±0.34 2.69 Ni 227.58 108.73 72.05±0.44 97.847 106.10±0.41±0.959 103.29 85.32±0.932 82.73±0.39 101.23±0.

73 127. and G.R.121 224. Ana Rita. Emmanuelle Poussel. 8.32±0.P. and A. 12.87±0.46±0.40 100. M.80 100. Mermet. E. 6. and Eva Curdova.W.Galley. 21. Spectrom.50±0.60 88.35 88. G.31±0.46±0.89±0.88 113.F.28 99. Jean-Michel Mermet. 14 X.65 100.24 98. Vicente Hernandis. Spectrosc. Joaquim Nobrega. 23.27 99. Factor Factor Corr.31 96.49±0.0 mg/L Recovery % (SD) Recovery % (SD) Recovery % (SD) Recovery % (SD) Without Applying Without Applying Without Applying Without Applying Factor Corr. At. Jesus Cano. and Roberto Frache. J.45 99.26 77. Spectrom. 193 . At.34 96. Acta. Arbind Kumar. Slov. 89 (2000).10 100. 12. Spectrochim. Spectrosc. M. Marco Grotti.75 100.33±0. 2.23 99. Spectrom. 9. 1333 (2001).Olesik. At. 14. L.48±0.31 100. and J.88 128. 16.932 295. and A.32 77.Nougueira.25±0. Spectrom. 877 (2000). At.35±0. Anal. Factor = Correction Factor. 419 (1998).37 99. Acta. 153 (2000).35±0.91 ± SD of five readings.616 396. REFERENCES 1.39 101. A. 1002 (2001).65±0.021 336.43 100.959 267. Anal. Anal. J.716 274. Ethel Dorfman. E.21 91.45±0.95 99. Irina Segal.78 113. and Juan Mora. 3. Spectrom.20 99. 15.P. Spectrosc. Edivaldo Garcia. Anal. J. Anna Krejcova. Anal.25 98. 16. 15. Anal.Glick.61±0. 119 (2002).Hieftje.36 99. 4.270 TABLE V Different Concentration of Traces Doped Into K2TaF7 (2 g/100 mL) 0.38±0. and J. 353 (1997).65 99. Mermet.32 96.54±0.78±0.76 100. J.M. to be published in 24(5) (2003).18 100. J. M. 495 (1997). Lilli Paama. Donard. Jose Todli. J.36 98. 13 S. Olga Yoffe. 10. Moltelica-Heino. 46 (2000).22 100.75 mg/L 1.30 101. Eric Salin.50 99.J . Margarita Villaneuva.23 96. Spectrom.29 91. P.23 99.32 100. At.79±0.33±0.49±0. 15.M. Emanuele Magi.32±0.25 mg/L 0. 47.18±0. Bojan Budic. Arbind Kumar.28 100. Spectrom. A.65 99.31±0.J.14±0. Spectrochim.85 99./Oct. At. and T.65±0. 16. At. 675 (1999). Acta. 48B.45 101. 57 (2002). Sept. 769 (1993). Spectrochim.18 99. Anil.30 100. At.847 280. and Mario Pomares.05±0.R.49 100. Anal.57±0. 13. Cendrine Dubuisson. Agranat. Tomas Cernohorsky. Reddy.70±0. At.41 77.53 100.80 100. Miguel Catasus.89±0.74 100.E.79 100. Factor 91.36 100.52 99.41 101.88±0. E. Factor Factor Corr. 16.L.50 mg/L 0.23 100. 825 (2001).77±0.78 127.34 88.29±0. 11.35 99. J.75 99. 281 (1997). 17. Spectrom.92±0. 24(5). At. 2003 Wavelength Elements B Cr Fe Nb Ni Ti Cu Co Ca Mg 208.44 99. Mare Must.088 227. Eed Parnoja. Anal.75 99. Prakash.62±0.45 108. and Jean Michel Mermet.15±0.31 98. Reddy.76 127.28 99. 5.31 88. 52B. and Paavo Peramaki. Spectrom.Vol. M.55±0. Irina Bezrukavnikov. G. Poussel.51 108. M. O. At.29 100.700 228. At. Acta Chim.65±0.39 113.38 100.39 77. Romeo. 7. J.43 99.48 108.L. Corr. Anil.M.43±0. 52B. and T.43 108.75 113.98 91. Factor Factor Corr.42 100. J. Hobbs and J. Prakash.X. Anal.

com/bonaventure) in Fort Lauderdale.wyndham. and software innovations • Sample introduction and transport phenomena • Sample preparation. and transportation details. Hadley. databases • Stable isotope analyses and applications January 2-4. and software • Spectroscopic standards and reference materials. Or contact: Ramon Barnes. treatment. Tue-Thu January 4. and sample preparation • Excitation mechanisms and plasma phenomena • Flow injection and flow processing spectrochemical analysis • Glow discharge atomic and mass spectrometry • Inductively coupled plasma atomic and mass spectrometry • Laser ablation and breakdown spectrometry • Microwave atomic and mass spectrometry • Plasma chromatographic detectors • Plasma instrumentation.O. high-purity materials • Spectrochemical chemometrics.2004 Winter Conference on Plasma Spectrochemistry Fort expert systems. P.unix. Florida (www. Sunday 62 Continuing Education Short Courses Manufacturer’s Seminars 5th Annual Golf Tournament Six Provocative Panel Discussions Workshop on New Plasma Instrumentation USP Roundtable Discussion Information For program. January 5-10. Fri-Sun January 2-4. visit our website at: http:>// www. hotel. USA. speciation sampling. Fax: (+1) 413-256-3746. e-mail: wc2004@chemistry. 2004 The 13th biennial international Winter Conference will be held at the Wyndham Bonaventure Resort and Spa (www. Sunday Daily January 6-8. USA Tel: (+1) 413-256-8942. and automation. microplasmas.umass. ICP Information Newsletter. More than 600 scientists are expected and approximately 300 papers on modern plasma spectromchemistry will be presented. 194 ..umass. MA 01003-0666. Inc. Fri-Sun January 4.oit.sunny. Box 666. Symposium Features • Elemental speciation. Florida.

com . 800-762-4000 © 2003 PerkinElmer. Talk to a PerkinElmer inorganic analysis specialist today. Even troubleshooting and repairs are easier. our newest AAnalyst is a better way to do AA. As rugged and reliable as ever.S. U. No service visit. that’s all the training you need.Atomic Absorption Just touch and go.perkinelmer. no down time. Experience it for yourself. It practically tells you what to do — and in your own language. available at your fingertips. There. Inc. (+1) 203-925-4600 www. with quick-change parts you simply snap out and snap in. Walk up to the AAnalyst 200 and let the touch screen guide you through everything from setup to analysis. All instrument controls are right there on the screen.

providing uncompromised sensitivity and performance in all matrices for even the toughest applications. This ion-molecule chemistry uses a gas to “chemically scrub” polyatomic or isobaric species from the ion beam before they enter the analyzer. 006872_03 . Unlike more simplistic collision cells. Eliminates interferences COMPLETELY When your applications extend beyond the capabilities of conventional ICP-MS. Mg. As. maximum uptime PlasmaLok® Easy optimization and extended cone life Dynamic Bandpass Tuning Efficiently screens out interferences while maximizing analyte transmission AutoLens™ Optimizes voltage for each element All-Quartz Sample Introduction Minimizes contamination SimulScan™ Simultaneous dual-stage detector with 9 orders of dynamic range High-Speed Quadrupole Fast transient signal analysis AutoRes™ Custom resolution minimizes spectral interferences and improves detection limits Dynamic Reaction Cell™ (DRC™) with Axial Field Technology provides superior interference reduction and performance for all applications. The DRC II combines the power of patented Dynamic Reaction Cell (DRC) technology with performance-enhancing Axial Field Technology. these uncontrolled reactions increase spectral complexity and create unexpected interferences. Unlike collision cell. or cold plasma systems. resulting in improved detection limits for elements such as Fe. K. high-resolution. you need the power of the innovative ELAN® DRC II. it eliminates sequential side reactions that create new interferences. Unless kept in check by DRC technology. patented DRC technology not only reduces the primary interference.Axial Field™Technology Optimizes performance and speed in all matrices Platinum Quick-Change Interface Cones Easy maintenance. Ca. Se. Cr. and V. The DRC II uses chemical resolution to eliminate plasma-based polyatomic species before they reach the quadrupole mass spectrometer. the DRC II completely eliminates polyatomic interferences providing ultratrace-level detection limits.