Double-focusing Sector Field Inductively Coupled Plasma Mass Spectrometry for Highly Sensitive Multi-element and Isotopic Analysis† Invited Lecture J. SABINE BECKER* AND HANS-JOACHIM DIETZE Zentralabteilung fu� r Chemische Analysen, Forschungszentrum Ju�lich GmbH, D-52425 Ju�lich, Germany The dierent areas of application in double-focusing sector charged atomic ions (e.g., 92Zr+ and 92Mo+: m/Dm#52 000) requires a mass spectrometer with high mass resolution (e.g., field ICP-MS are described, such as the determination of trace and ultratrace elements in environmental and materials Fourier transform or ion trap mass spectrometers,10–12 interferences of singly charged with doubly charged atomic ions research and for the characterisation of long-lived radionuclides in environmental and radioactive waste samples.(e.g., 92Mo+ and 184W2+: m/Dm#1300) can be separated in many cases at the required mass resolution by using a double- Analytical methods using double-focusing sector field ICP-MS allow the determination at low mass resolution of ultratrace focusing sector field mass spectrometer. The apparent interferences of atomic ions of analyte and disturbing molecular elements (e.g., rare earth elements) and some radionuclides in the mg l-1 and pg l-1 concentration ranges (e.g., for U, Th, Pu ions at the same nominal mass in the mass spectra (e.g., 41K+ and 40ArH+: m/Dm#5000) in all mass spectrometric methods and 129I ) in aqueous solutions and at the ng g-1 level and lower in solid samples after a digestion step.Some examples are of the greatest importance, as demonstrated by the numerous investigations that have been reported.13–20 Therefore, of trace analysis of solid samples after matrix separation are discussed. For the determination of spallation nuclides and overcoming these interference problems of atomic ions of the analyte with molecular ions is important for accurate and impurities in irradiated tantalum (with an 800 MeV proton beam) from a spallation neutron source, double-focusing sector precise trace, ultratrace and isotopic analysis of inorganic materials by ICP-MS. The rate of formation of molecular ions field ICP-MS was used after liquid–liquid extraction of the tantalum matrix.Further, the results of the analysis of high- (such as MOn+, MOnH+, CnOm+, NnOm+; MN+, MCl+, MAr+ with M=matrix element; ArnXm+ with X=H, O, N, Cl purity GaAs by double-focusing ICP-MS after dissolution with and without matrix separation are compared with those of and others) in an inductively coupled plasma is dependent on the stability of the ionic species (bond dissociation energy) and quadrupole ICP-MS.varies as a function of rf power, nebulizer gas flow rate and Keywords: Double-focusing sector field mass spectrometry; sampling depth, as demonstrated in many papers.14–16,18,20 A environmental samples; high-purity GaAs; inductively coupled mass spectrometric separation of such interferences applying plasma; long-lived radionuclides; molecular ions; tantalum; a double-focusing sector field mass spectrometer at higher ultratrace analysis mass resolution in comparison with low-resolution quadrupole ICP-MS (ICP-QMS) is often useful for the determination of elements abundant in nature in the mass range up to 75 u.21–26 Among the dierent methods for the determination of trace and ultratrace elements and isotopic composition in inorganic Reed et al.19 reported these possible interferences of molecular ions and atomic ions from 24Mg+ to 76Ge+ and discussed the materials, ICP-MS is well established as a universal, powerful and very sensitive multi-element method applicable in all fields determination of elements in dierent matrices using a doublefocusing sector field ICP-MS with a maximum mass resolution of modern science and technology: in materials research, e.g., for the characterisation of high-purity materials, in the semi- of 10 000.The mass resolution required for the separation of oxide conductor industry and microelectronics, in environmental and biological research, in medical science and in geology and ions (MO+) from the atomic ions of the analyte as a function of mass is demonstrated in Fig. 1. With a maximum mass mineralogy.1–4 The determination of elements in the trace and ultratrace concentration range is often dicult owing to poss- resolution of 10 000 for commercial double-focusing ICP mass spectrometers, the isobaric interferences of atomic and oxide ible matrix eects, which could be avoided by matrix separation or selective preconcentration of the trace elements of interest.5–9 ions in the mass range with the most abundant elements in nature can be separated.Maximum mass resolution Furthermore, possible contamination during sample preparation and high blank values, especially from elements abun- (m/Dm#4×106) is required for the separation of interferences of 90Zr18O+ and 108Pd+.All other points in the figure relate dant in nature, lead to incorrect analytical results. Blank values and possible contamination should be minimised by careful to interferences of M16O+ (M=metal or non-metal) with the most abundant analyte ions at a mass which is 16 u higher working under ultraclean conditions and using ultrahighpurity chemicals. In addition to these general problems, trace than mass of M. It is clearly seen that possible interferences of MO+ and analyte ions with mass higher than 70 u are dicult and ultratrace analysis is often disturbed by possible interto separate using, e.g., double-focusing sector field ICP-MS ferences in mass spectra.Whereas due to the low mass dierwith a maximum mass resolution (m/Dm) of 7500. In practice, ence the separation of the interferences of isobaric singly the theoretically required mass resolution is insucient if the intensity of molecular ions is significantly higher than that of † Presented at the 1997 European Winter Conference on Plasma Spectrochemistry, Gent, Belgium, January 12–17, 1997.the analyte ions. Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12 (881–889) 881Further molecular ion formation in the ICP and by the expansion of plasma in the low-pressure interface of the ICP-MS system can be reduced using the cool plasma technique in combination with a shielded ICP torch.33,34 This cool plasma technique leads to a significant improvement in the determination of some elements important in the environment, e.g., Fe, Ca, Na and K, with detection limits in the low ng l-1 concentration range, as was demonstrated by Tanner et al.35 and Georgitis et al.36 using quadrupole ICP-MS.An interesting development proposed by Speakman et al.37 in quadrupole ICP-MS is the use of a collision cell filled with helium in combination with a hexapole ion lens behind the skimmer cone for the thermalisation of ions in order to decrease the energy spread and to dissociate disturbing molecular ions.COMPARISON OF DOUBLE-FOCUSING SECTOR FIELD ICP-MS INSTRUMENTS Some experimental parameters and properties of the commercial double-focusing sector field ICP mass spectrometers Element (Finnigan MAT, Bremen, Germany), PlasmaTrace 2 (Micromass, Manchester, UK) and JMS-Plasma X2 (Jeol, Tokyo, Japan)38–41 are compared in Table 1. All these mass spectrometers are double-focusing sector-field instruments with reverse Nier–Johnson geometry.A double-focusing sector field mass spectrometer combines a magnetic sector field for direction focusing and an electric sector field for energy focusing of ion beams. Double focusing is reached at the point where the two image curves for focusing of energy and direction cross. Fig. 1 Mass resolution required for separation of oxide ions from The performances of commercial double-focusing sector field atomic ions of the analyte as a function of mass.ICP-MS systems are similar with respect to the achievable high sensitivity (n×107 ions s-1 per mg l-1 at In), very low noise (<0.2 ions s-1) and very low detection limits (in the The mass resolution required for the separation of argon molecular ions (ArX+) from atomic ions of the analyte has low pg l-1 concentration range determined at low mass reon in ideal aqueous solutions). A dierence is in the slit been discussed.16 It is possible to separate most ArX+ from atomic ions of the analyte at a mass resolution of 10 000, apart system (and in the price).The slit system of the PlasmaTrace 2 with up to five resolution values allows a variable mass from some metal argide ions in the mass range 80–100 u. In contrast to the curves for the mass resolution required resolution by automatically changing the resolution in order to maximise the transmission of ions for each element. for the separation of MO+ and ArX+ from atomic ions of the analyte as a function of mass with the maximum at a mass of Compared with the fixed slit system of Element, this fully variable slit system has the advantage that the mass resolution about 100 and 90 u, respectively, the mass resolution required for separation of hydride ions (MH+) increases with increasing required for the separation of interferences of atomic and molecular ions can be adjusted.mass, where at mass>89 u a mass resolution (m/Dm) of >10 000 is necessary for the separation of interferences with A double-focusing sector field ICP-MS system with Mattauch–Herzog geometry and multi-channel ion detection atomic ions. This means that in many cases a decrease in molecular ion formation is required for the determination of (photodiode-array detector) is used for the simultaneous detection of ions, especially for the detection of transient signals, as ultratrace elements, especially when a complicated matrix is to be analysed. described by Cromwell and Arrowsmith.42 This simultaneous ion detection is important for precise isotope measurements.The ways of decreasing molecular ion formation in order to improve the detection limits in double-focusing ICP-MS, by The major limiting factor on the precision of isotopic ratio measurements by ICP-MS is the instability of the argon applying the same as used in quadrupole ICP-MS, i.e., by applying special sample introduction systems, matrix separa- plasma.Therefore, a multi-collector ion detection system is realised by the double-focusing ICP-MS Plasma 54 prototype tion, enrichment of trace impurities, suppression of molecular ion formation, etc. The combination of electrothermal vaporis- (VG Elemental, Winsford, Cheshire, UK) with nine Faraday detectors.43 Recently, a single-focusing magnetic sector field ation or a hydride generator with ICP-MS (e.g., for the sensitive determination of Se or As,27–30) which is applied in order to ICP-MS instrument with multi-collector (Iso-PlasmaTrace, Micromass)44 was introduced for the determination of precise reduce some of the disturbing interferences by selective separation of the analyte, leads to a loss of the multielement capability isotopic ratios.Multi-collection using the fully adjustable Faraday cup permits the measurement of isotopic ratios down of ICP-MS. Furthermore, using an ultrasonic nebulizer, the microconcentric nebulizer (both with desolvator), direct injec- to a precision of better than 0.002%.As an alternative to double-focusing sector field ICP-MS, a tion nebulizer or thermospray nebulizer–membrane separator31,32 for small sample volumes, multielement analysis can quadrupole ICP-MS system was recently developed by Ying and Douglas45 with a maximum mass resolution of 9000. At a be carried out in the ultratrace concentration range with detection limits at the ng l-1 level.The separation of matrix mass resolution of 5000 the sensitivity is 106 ions s-1 per mg l-1 of element in solution. This sensitivity is comparable to that elements or of the organic matrix or enrichment of selected trace elements, such as by extraction, ion chromatography or of commercial double-focusing sector field ICP-MS systems operated at the same mass resolution. In the low-resolution HPLC, can be performed before mass spectrometric measurement or on-line coupled to ICP-MS.5–9 mode the sensitivity is 108 ions s-1 per mg l-1 of element in 882 Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12Table 1 Comparison of commercial double-focusing ICP mass spectrometers Element PlasmaTrace 2 JMS-Plasma X2 (Finnigan MAT) (Micromass) (Jeol) MS configuration Reverse Nier–Johnson Reverse Nier–Johnson Reverse Nier–Johnson Rm=16 cm Rm=20 cm Rm=31 cm Re=10.5 cm Re=26.5 cm Re=22.3 cm Accelerating voltage UB=1–8 kV UB=1–6 kV UB=1–5 kV Mass range 2–260 Da 2–300 Da 2–500 Da Mass resolution setting 3 resolutions 5 resolutions Continuous Maximum mass resolution 7500 10 000 10 000 (m/Dm) Ion detection Analogous; Analogous; Analogous; (single ion collector) ion counting ion counting ion counting Noise <0.2 counts s-1 <0.2 counts s-1 <0.2 counts s-1 Sensitivity 2×107 counts s-1 per mg l-1 at In 5×107 counts s-1 per mg l-1 at In 2×107 counts s-1 per mg l-1 at In (m/Dm=300) (m/Dm=400) (m/Dm=500) solution.The continuum background is significantly higher 20 mg l-1 for Zn.Most elements are measured with high sensitivity at low mass resolution. (about 1000 ions s-1) compared with that of a commercial double-focusing sector field ICP-MS instrument. A decrease ICP-MS is used in some applications for the determination of only one ultratrace element. Takaku et al.48 described the in the background and clarification of the commercial production of such mass spectrometers are major tasks in further determination of silicon in ultrahigh-purity water for microelectronics and investigated the instrumental blank due to the development work.The application of double-focusing sector field ICP-MS in quartz plasma torch, nebulizer and spray chamber. For the determination of silicon the contamination problems are most trace and ultratrace analysis is not restricted to the determination of some ‘dicult’ elements (Cr, Cu, Ni, Fe, V, As, Se, important. The silicon contamination from the original quartz plasma torch was determined to be 5–20 mg l-1. The back- P, Al, Si, S, Co, Ti, Mn, Sc) in environmental samples, highpurity metals or semiconductors, ceramics, etc., by the mass ground of the PlasmaTrace double-focusing ICP-MS system could be decreased by about one order of magnitude using a spectrometric separation of molecular ions from the atomic ions of the analyte.The largest application field in dierent platinium torch and Teflon nebulizer.The determination of Si using the PlasmaTrace double-focusing sector field ICP-MS laboratories is the highly sensitive determination of many other elements (e.g., Li, Be, B, Na, noble metals, lanthanides, actin- system in order to separate the interferences with 12C16O+ and 14N2+ at a mass resolution of about 2000 in semiconductor- ides) carried out at low mass resolution, which will be discussed in the following sections. grade water yielded <1 mg l-1 in comparison with the Si concentration in Milli-Q-purified and de-ionized water of 30 and 35 mg l-1, respectively.For the determination of Si by DETERMINATION OF IMPURITIES OF TRACE double-focusing sector field ICP-MS in the ng l-1 concen- AND ULTRATRACE ELEMENTS IN WATER tration range, a preconcentration step by evaporation in a SAMPLES clean bench is necessary. A similarly important analytical task for the microelectronics ICP-MS using an ecient nebulizer in order to introduce industry is the determination of P in high-purity water or aqueous solutions is ideal for the highly sensitive multi-element acids.The determination of phosphorus (mass spectrometric analysis of water samples for the microelectronics industry separation of the analyte ions 31P+ from interfering molecular (high-purity water) and environmental research (terrestrial ions 15N16O+ and 14N16O1H+) was performed in our labora- water, drinking or rain water, waste water or snow). The tory at a fixed mass resolution of 3000 using an Element detection limits of most elements attained by double-focusing double-focusing sector field ICP-MS system with detection ICP-MS in the low-resolution mode were in the extremely limits in the 10 ng l-1 concentration range.low pg l-1 concentration range for high-purity water, as dem- Narasaki and Cao41 developed a method for determining onstrated by Yamasaki et al.46 using direct sample introduction As and Se by hydride generation ICP-MS using a JMS-Plasma in an ultrasonic nebulizer combined with a PlasmaTrace X2 double-focusing sector field ICP-MS instrument (Jeol), double-focusing sector field ICP-MS system.The detection which allows a high sample introduction eciency owing to limits for elements which were measured at a higher mass matrix separation. The detection limits were 30 ng l-1 for As resolution compared with ICP-QMS varied for dierent and 60 ng l-1 for Se. Although the mass resolution required elements from 0.019 ng l-1 for Sc through 2.09 ng l-1 for Fe for the separation of interferences of 82Se+ and 40Ar2H2+ is to 360 ng l-1 for Si.They determined around 40 ultratrace about 3500, the measurements were carried out at m/Dm= elements in terrestrial water. In further work, Yamasaki et al.47 10 000 owing to the relatively high disturbance by 40Ar2H2+. determined especially the lanthanide and actinide concen- The method developed using hydride generation ICP-MS was trations in terrestrial water in Japan with a precision of 5% applied to the determination of trace levels of As and Se in (RSD) with the same experimental arrangement.Although the river water in the low mg l-1 to ng l-1 concentration range. isobaric interference of 153Eu+ with 137Ba16O+ ions at a mass resolution of 7500 could be separated, they used the mathematical correction of interferences of BaO+ ions on dierent DETERMINATION OF ULTRATRACE isotopes of lanthanide ions in order to avoid a serious decrease ELEMENTS IN ENVIRONMENTAL AND in sensitivity.The detection limits for lanthanide and actinide BIOLOGICAL SAMPLES elements in terrestrial water samples without sample pretreatment are in the low pg l-1 concentration range and The application of double-focusing ICP-MS is demonstrated for the determination of elements dicult to determine in <1 pg l-1, respectively. The element concentrations varied in rainwater samples over a concentration range of about six biological and medical research in the ultralow concentration range, e.g., by the groups of Moens23 and Duneman.26 orders of magnitude, from 10 pg l-1 for some lanthanides to Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12 883Riondato et al.49 measured the ultratrace elements P, S, Si, Al, coupled to an Element double-focusing sector field ICP-MS system.51 Iodide in the sample solution was oxidised to gaseous Cr, Mn, and Ti in human serum by flow injection coupled to an Element double-focusing sector field ICP-MS system at a iodine with concentrated perchloric acid.By desolvation of the initial aqueous solution, an increase in ion intensity of about fixed mass resolution of m/Dm=3000. Owing to numerous disturbing interferences the determination of precious metals two orders of magnitude was achieved in comparison with a conventional Meinhard nebulizer with a detection limit of (Rh, Pd, Ag and Pt) at the ultratrace level in biological materials is dicult, as demonstrated by Begerow et al.26 100 ng l-1 in solutions.An improvement of the detection limit in the determination In our laboratory, some environmental standard reference materials (SRMs, e.g., apple leaves, NIST 1515; pine needles, of 129I is dicult owing to interference with the 129Xe isobar. Furthermore, for determination of 129I, on-line isotope dilution NIST 1575; and mussel tissue, BCR 278) were investigated in which the rare earth elements are incompletely certified.For with stable 127I using the flow injection principle in order to improve the accuracy of results has been applied.51 their selective determination after digestion of the samples the rare earth elements were separated from the matrix using Another task in our institute is the nuclide analysis of an irradiated tantalum target. Tantalum was used as the target liquid–liquid extraction.8 The results of the determination of rare earth elements in dierent SRMs using quadrupole material in a spallation neutron source with 800 MeV protons.The determination of the concentration of spallation nuclides ICP-MS without and with isotope dilution and doublefocusing sector field ICP-MS (using an Element system) agree in a highly radioactive solid matrix in the concentration range from 1 ng g-1 to 50 mg g-1 should serve to verify the theoretical well with one another and also with the certified values, if available.All rare earth elements could be determined only by results of spallation yields of tantalum. For the determination of spallation nuclides in irradiated tantalum, an Element double-focusing ICP-MS after matrix separation. The determination of As and Se in biological SRMs after double-focusing sector field ICP-MS system52 can be used after liquid–liquid extraction of the tantalum matrix in order acid digestion of sample and cation exchange (using a Chelex-100 column) was described by Narasaki and Cao.41 to reduce the high 182Ta activity.The method for the determination of trace impurities after matrix separation was developed using high-purity inactive tantalum. In Table 2, the results ULTRATRACE DETERMINATION OF of the trace analysis for inactive tantalum after matrix separa- LONG-LIVED RADIONUCLIDES tion by double-focusing sector field ICP-MS are compared with those of neutron activation analysis (NAA). It can be The special area in which double-focusing sector field ICP-MS is advantageous concerns the ultratrace determination of long- clearly seen that in comparison with NAA, ICP-MS allows more elements to be determined in low concentration ranges.lived radionuclides in environmental and radioactive waste samples at very low concentration ranges. Conventional radio- The main problem in determining long-lived radionuclides in irradiated tantalum is possible interferences of radioactive metric methods are, however, time consuming at low concentration levels and possess a low specific activity and poor spallation nuclides with stable isotopes at the same mass with a dierent atomic number.Therefore, in order to separate precision. The correct and precise determination of long-lived radionuclides is required for the determination of enrichments isobars for rare earth elements, such as the long-lived 151Sm from stable 151Eu, HPLC was coupled with double-focusing of radioactive nuclides due to nuclear weapons testing or fallout from, e.g., the Chernobyl accident in biological samples, ICP-MS in our laboratory.53 The method of separation of rare earth elements with natural isotopic abundances and enriched waters and geological materials.Further, the determination of long-lived radionuclides in radioactive wastes from nuclear stable isotope HPLC–ICP-MS has been described;53 investigations on irradiated Ta samples after matrix separation are reactors for recycling and final storage of radioactive waste is important. In addition to the required isotopic analysis for U, in progress.The determination of 99Tc in environmental samples is Th, Cm, Am and Pu, especially mass spectrometric methods for the sensitive determination of the long-lived radionuclides disturbed by isobaric interference with 99Ru. Yamamoto et al.54 avoided this interference by the separation of 99Tc using 237Np, 129I, 99Tc, 79Se, 107Pd, 135Cs, 93Mo, 93Zr, 151Sm, etc., in the ultralow concentration range are of interest, where other dierent solvent extraction and purification techniques using anion exchange after leaching of the soil sample.The soil radiometric methods experience diculties. Kim et al.50 determined the detection limits for the long- sample was spiked with a 99mTc nuclide before leaching and no loss of 99Tc was observed during this treatment. Yamamoto lived radionuclides 99Tc, 226Ra, 232Th, 237Np, 238U, 239Pu and 240Pu with half lives of 103–1010 years in standard solutions.et al. determined 99Tc with an absolute detection limit of 0.25 pg, corresponding to 0.16 mBq, and analysed sediment Using a PlasmaTrace double-focusing ICP-MS system with an ultrasonic nebulizer, the detection limits ranged from 2 to samples from the Irish Sea. The same group determined 237Np in a similar way after leaching of the soil sample, solvent 20 pg l-1. The sensitivity of double-focusing ICP-MS with the application of an ultrasonic nebulizer was 10 times better than the same arrangement without the ultrasonic nebulizer.Table 2 Determination of impurities in inactive tantalum after matrix Double-focusing sector field ICP-MS has also been used for separation (concentrations in mg g-1) the selective and sensitive determination a single long-lived Element ICP-MS NAA radionuclide, e.g., the determination of 79Se.28 In this case, hydride generation was used in order to improve the sample Al 1.52±0.05 — V <0.006 — introduction eciency and to avoid possible interferences with Fe 0.88±0.09 — isobaric singly charged ions (79Br+), doubly charged atomic Co 0.023±0.006 0.032±0.003 ions (158Gd2+ and 158Dy2+) and dierent molecular ions Zn 0.45±0.15 <0.5 (39K40Ar+, 38Ar40ArH+ and 63Cu16O+).Most of the possible Zr 0.34±0.1 <8 interferences, but not the 38Ar40ArH+ ions, were eliminated. Mo 17.6±0.6 35±3 The detection limit of 79Se using a special hydride generator Ba 0.06±0.02 <7 La <0.006 <0.001 coupled to an Element double-focusing sector field ICP-MS Hf <0.11 0.12±0.2 system was about 100 ng l-1.Similarly to the hydride generator W 208±20 332±30 described by Hoppstock et al.,28 the sample introduction Ir 0.49±0.04 — equipment for the highly sensitive determination of iodine and Pb 0.48±0.05 — also the stable isotope 127I and the long-lived nuclide 129I is 884 Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12Table 3 Comparison of some results obtained on GaAs after selective extraction and anion exchange in order to evaluate the fallout volatilization of the matrix (concentrations in ng g-1) radionuclides from nuclear weapons tests. Whereas in the determination of selected long-lived radio- Element ICP-QMS Double-focusing ICP-MS nuclides the high sensitivity of double-focusing ICP-MS, but Li <10 3.2±0.8 not the multi-element capability, is used, the multi-element Be <5 <1 capability is important for the characterisation of high-purity Mg 190±45 225±4 materials.Al 1076±108 850±10 V 24±8 <5 Sn <20 3.8±1 DETERMINATION OF TRACE AND Te <10 4±1 ULTRATRACE IMPURITIES IN HIGH-PURITY Pb 102±5 32±2 SOLID MATERIALS AFTER DIGESTION Bi <50 25±3 Trace analysis of high-purity materials is important especially in the microelectronics industry. With increasing purity of (with interchangeable ICP and rf GD ion sources) for trace metals, semiconductors and insulators, an improvement of and ultratrace analysis of GaAs has been described.55 ultratrace analytical techniques is required.Takaku et al.25 The applications of double-focusing sector field ICP-MS in determined trace impurity rare earth elements in high-purity trace and ultratrace analysis are summarized in Table 4. Gd2O3 and Y2O3 using double-focusing ICP-MS. Because the mass resolution required for the separation of rare earth element ions and disturbing molecular ions of Gd is higher ISOTOPIC RATIO MEASUREMENTS than was practically possible with ICP-MS, doubly charged An important advantage of double-focusing sector field rare earth element ions were used as analyte ions for the ICP-MS compared with quadrupole ICP-MS is in the area of detection of trace impurities in lanthanide oxides.The ratios isotopic ratio measurements. The precision of the isotopic of doubly charged to singly charged rare earth element ions abundance ratios has typically been limited to 0.1–0.5% with varied from 0.2% for Lu to 11% for Ce.The detection limits the quadrupole ICP-MS systems commercially available. In for rare earth elements were 80 pg l-1–3 ng l-1 using doubly general, isotopic ratio measurements using double-focusing positively charged ions, compared with 6–30 pg l-1 using sector field ICP-MS yield more accurate and precise isotopic singly charged ions by analysing an ideal aqueous solution. ratios owing to better counting statistics and lower noise in The real concentrations of trace impurities of rare earth comparison with quadrupole ICP-MS. elements in high-purity solid Gd2O3 and Y2O3 were measured The correct and precise isotopic analysis of radioactive in the ng g-1 concentration range.elements, such as U and Pu, is required for the determination A method with selective volatilisation of the matrix was of the enrichment of radioactive nuclides due to nuclear developed in our laboratory for the ultratrace analysis by weapons testing in biological samples, waters and geological ICP-MS of GaAs, which is important for microelectronics.materials. Kim et al.50 compared the results of the isotopic This selective volatilisation of GaAs is achieved by converting analysis of a Pu isotopic standard solution containing 20 ng l-1 the matrix elements into their chlorides in a stream of argon Pu by double-focusing sector field and quadrupole ICP-MS.and gaseous chlorine. Owing to the low volatility of the The accuracy and precision of Pu isotopic analysis could be chlorides of gallium and arsenic (melting points, GaCl3 78 °C improved by double-focusing ICP-MS with an ultrasonic and AsCl3 -20 °C; boiling points, GaCl3 201 °C and AsCl3 nebuliser compared with quadrupole ICP-MS by about one 130 °C), it is possible to separate the matrix elements, leaving order of magnitude. behind those impurities which do not readily form volatile Isotopic standard solutions of uranium have been analysed chlorides.A schematic diagram of the experimental arrangeto determine the mass discrimination eect (the mass discrimi- ment for matrix separation is shown in Fig. 2. The GaAs nation was determined to be about 0.1%) during the measure- sample was transferred into the heating chamber in a quartz ment and the dead time eect on the ion detector (about boat. After purge treatment, 0.5 g of sample was completely 26 ns).56 In Table 5, the isotopic ratios of uranium measured volatilized (30 min) in an argon–chlorine stream (Ar5Cl#551) on a digested soil sample are compared with the values from at 220 °C.The small amount of white residue in the quartz an isotopic table.57 An agreement with the natural isotopic boat was dissolved in hot HNO3. abundance of uranium was measured. The precision of The results of trace impurity determination in high-purity determining the 235U/238U isotopic ratio of uranium in a real GaAs after matrix separation using quadrupole and doublegeological sample was approximately 0.6%.The possible inter- focusing ICP-MS are summarised in Table 3. In comparison ferences of molecular ions in the mass spectra of dissolved soil with the results without matrix separation, the detection limits samples in the mass range of actinides are discussed in the were improved from 500 ng g-1 by over two orders of magninext section. tude in the low ng g-1 concentration range.These detection Vanhaecke and co-workers58,59 demonstrated the capability limits are comparable to those of direct solid analysis on GaAs of double-focusing ICP-MS at low mass resolution in determin- by rf GDMS. A comparison of ICP-MS and rf GDMS using ing the isotopic ratios of magnesium and lead. At higher an Element double-focusing sector field mass spectrometer element concentrations isotopic ratios with an RSD of 0.04% were measured. The determination of the 63Cu/65Cu isotopic ratio at a mass resolution of 3000 yielded a precision of about 0.2%.The use of a multiple collector system in double-focusing ICP-MS (Plasma 54 prototype combined with a desolvator), as demonstrated for the determination of the 176Hf/177Hf ratio by Walder et al.61 allows isotopic ratio measurements at precisions of down to 0.002% (RSD) (concentration of Hf, 50 mg l-1). In their investigations of geological, environmental and nuclear materials, Walder and co-workers,60–62,65 showed that it is possible to achieve a precision for isotopic analysis Fig. 2 Experimental set-up for separation of GaAs. comparable to that of thermal ionisation mass spectrometry. Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12 885Table 4 Application of double-focusing sector field ICP-MS in trace and ultratrace analysis Sample Equipment Elements Concentration range Limit of detection Ref. Terrestrial water PlasmaTrace 40 ultratrace elements 20 mg l-1 (Zn) – At m/Dm=400: Yamasaki et al.46 and USN 15 pg l-1 (Tm) 0.5–5 pg l-1 (lanthanides) At m/Dm=3000: 0.02 ng l-1 (Sc) 2 ng l-1 (Fe) Terrestrial water PlasmaTrace Lanthanides and 1–40 ng l-1 0.5–12 pg l-1 Yamasaki and and USN actinides Tsumura47 Ultrahigh-purity water PlasmaTrace Si (after preconcentration) 0.3–35 mg l-1 0.1 mg l-1 Takaku et al.48 (restricted by blank) Biological samples JMS-Plasma X2 and As, Se Biological SRM: In solution: Narasaki and Cao41 (SRM) hydride generator 0.4–26 mg g-1 As 30 ng l-1 (As) 0.4–7 mg g-1 Se 60 ng l-1 (Se) River water River water: low mg l-1 Standard solutions PlasmaTrace 99Te, 226Ra, 232Th, 1–6 ng l-1 0.002–0.02 ng l-1 Kim et al.50 and USN 237Np, 238U, 239Pu, 240Pu Human serum Element V, Fe, Cu, Zn, Ag 0.8 ng g-1 (V)– In solution: Moens et al.23 24 mg g-1 (Fe) 4.3 ng l-1 (Ag) Human serum Element and P, S, Si, Al, Cr, Mn, 0.7 ng l-1 (U) 0.05 mg l-1 (Ti) Riondato et al.49 flow injection Ti, Ag, Cd, Sn, K 226 ng l-1 (Cd) 0.66 mg l-1 (Al ) (restricted by blanks) Human urine Element Rh, Pd, Ag, Pt 0.5–7.6 ng l-1 (Pt) 0.24 ng l-1 (Pt) Begerow et al.26 0.17 ng l-1 (Au, Pd) Biological samples Element Lanthanides after 0.014–20 ng g-1 0.01 ng g-1 Panday et al.8 (SRM) liquid–liquid extraction Biological samples Element and Se(79Se) Biological samples: In solution: Hoppstock et al.28 (SRM), radioactive hydride generator 0.05–1.7 mg g-1 100 ng l-1 waste Biological samples Element and special 129I, 127I Biological samples: In solution: Kerl et al.53 (SRM), radioactive sample introduction 0.3–1.8 mg g-1 50 ng l-1 waste (0.8 ml sample volume) Tantalum Element 39 elements after 0.008–200 mg g-1 1 ng g-1 (Lu, Tm) Becker et al.52 matrix separation 6 ng g-1(V) Environmental samples PlasmaTrace 99Tc after solvent extraction Sediments: In solution: Yamamoto et al.54 (geological samples) 0.8–3.5 ng g-1 0.25 ng l-1 High-purity solids: PlasmaTrace lanthanides Gd2O3: In solution: Takaku et al.25 Y2O3, Gd2O3 2–3200 ng g-1 6–30 pg l-1 Al2O3 Element prototype V, Cr, Mn, Fe, Ga, Co, 0.04–3.7 mg g-1 At m/Dm=300: Jakubowski et al.24 Ni, Cu, Zn, Ce 0.4–400 ng g-1 At m/Dm=3000: 8–1400 ng g-1 Concrete, radioactive Element and 99Tc, 232Th, 233U, 235U 30ngg-1–6.1 mg g-1 In synthetic concrete Gastel et al.68 waste samples laser ablation 238U, 237Np laboratory standard: 0.5 ng g-1 (237Np) 4 ng g-1 (99Tc) 886 Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12Table 5 Results of isotopic analysis of uranium molecular ions in the high mass range of a digested soil sample is shown in Fig. 3. In the mass spectrum of a soil sample Uranium of natural isotopic digested with an acid mixture containing HCl, dierent molecu- Isotope ratio composition57 Uranium in soil sample lar ions in the mass range of actinides were observed. The 233U/238U — <10-6 relative ion intensities of disturbing PbCl+ and PbAr+ molecu- 234U/238U 5.5×10-5 5.6×10-5 lar ions, some radionuclide ions disturbed by molecular ions 235U/238U 7.26×10-3 7.25×10-3 in the investigated mass range and their half-lives are summar- 236U/238U — <10-6 ized in Table 7.Owing to the positive mass defect for actinides, the required mass resolution for the separation of disturbing molecular ions from long-lived actinide ions is relatively low, The determination of possible isotopic variations in nature m/Dm#2000. A systematic study of possible interferences in due to radioactive decays of unstable nuclides will be used in the determination of ultralow concentrations of actinides is geochronology for age determination (e.g., Rb–Sr method necessary. In order to exclude molecular ion formation in this based on the b-decay of 87Rb; Sm–Nd method; Re–Os age mass range for the determination of ultratraces levels of determination or U–Pb method which uses the fission of 238U) actinides, the method of choice will be a selective actinide where precise isotopic analysis is necessary.For geochronology, separation. thermal ionisation mass spectrometry (TIMS) is the most established method for precise isotope measurements. With the improvement in the precision of isotopic ratio measure- SOLID STATE MASS SPECTROMETRY ments by double-focusing sector field ICP-MS, this method In order to analyse solid samples by mass spectrometric also allows such applications if the elements of parent and methods without dissolution steps, a laser ablation (LA) system daughter nuclides are separated chemically.was coupled to double-focusing sector field ICP-MS. The applications of double-focusing sector field ICP-MS in LA–ICP-MS has been realized by dierent groups, e.g., by isotopic analysis are summarized in Table 6 Walder et al.66 Christensen et al,67 using the Plasma 54 doublefocusing ICP-MS prototype. A non-commercial laser ablation FORMATION OF MOLECULAR IONS system was coupled to an Element double-focusing sector field A systematic study of cluster ion and molecular ion formation in an ICP can be used to measure interferences with analyte ions in the determination of trace and ultratrace elements in inorganic materials.14–16,18,20 The knowledge of the formation, abundance distribution and electronic stability of molecular and cluster ions is of great analytical significance if mass spectrometric systems with low mass resolution (e.g., quadrupole ICP-MS) are applied for trace and ultratrace analysis.A correlation has been found for metal argide (MAr+) ion intensities in ICP-MS and binding dissociation energies.16 From such a linear correlation curve, unknown dissociation energies, e.g., CdAr+ and MnAr+, were estimated. A comparison of theoretically investigated bond dissociation energies of argon molecular ions for elements of the second and third periods of the Periodic Table with measured ion intensities by double-focusing ICP-MS yielded a qualitative correlation.Higher intensities of non-metal argon molecular ions in comparison with the species with lower intensities in ICP-MS can be interpreted owing to the higher stability.16 Besides the general interest of molecular ion formation Fig. 3 Part of a mass spectrum in the mass range of actinides. investigated by ICP-MS, a practical example of disturbing Table 6 Application of double-focusing sector field ICP-MS in isotopic measurements Precision of isotopic analysis Sample Equipment Isotopic ratio (at concentration) Ref.Standard solutions PlasmaTrace and USN 240Pu/239Pu 2.0% (20 ng l-1) Kim et al.50 Standard solutions Element 25Mg/26Mg At m/Dm=300: Vanhaecke et al.50 206Pb/207Pb 0.04% (Mg 5 mg l-1; Pb 100 mg l-1 Sediment digests, Element 63Cu/65Cu At m/Dm=3000: Vanhaecke et al.59 human serum (SRM) 0.096% (1 mg l-1) Standard solutions, Element 235U/238U 0.07% (10 mg l-1) Kerl et al.56 uranium metal, and USN 0.23% (100 ng l-1) waste sample Standard solutions Plasma 54 prototype 233U/238U 0.03% (1 mg l-1) Taylor et al64 235U/238U Standard solutions Plasma 54 prototype 176Hf/177 0.002% (Hf 1 mg l-1) Walder et al.63 (Mistral) 208Pb/204Pb 0.05% (Pb 1 mg l-1) Standard solutions Plasma 54 prototype 87Sr/88Sr 0.008% (Sr 1 mg l-1) Walder and Freedman61 235U/238U 0.014% (U 1 mg l-1) Glass (SRM) Plasma 54 prototype 208Pb/204Pb; 206Pb/204Pb; 208Pb/204Pb:~0.1% Walder et al.66 and laser ablation 207Pb/204Pb (426 mg g-1) Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12 887Table 7 Relative molecular ion intensities and possible interferences with actinide ions Half-life Relative ion Possible of actinide intensity interference by Disturbed nuclide Molecular ion (MX+/M+) molecular ion actinide ion years PbCl+ 8.0×10-6 206Pb35Cl+ 241Am+ 4.3×102 207Pb35Cl+ 242Pu+ 3.8×105 208Pb35Cl+ 243Am+ 7.4×103 207Pb37Cl+ 244Pu+ 8.3×107 208Pb37Cl+ 245Cm+ 8.5×103 PbAr+ 4.3×10-6 206Pb40Ar+ 246Cm+ 4.7×103 207Pb40Ar+ 247Cm+ 1.6×107 208Pb40Ar+ 248Cm+ 3.4×105 13 Shao, Y., and Horlick, G., Appl.Spectrosc., 1991, 45, 143. mass spectrometer in our laboratory,68 where a method for 14 Nonose, N. S., Matsuda, N., Fudagawa, N., and Kubota, M., the determination of some radioactive nuclides (e.g., 99Tc, 129I, Spectrochim. Acta, Part B, 1994, 49, 955. 232Th, 233U, 237Np and 238U) in non-conducting radioactive 15 Nui, H., and Houk, R. S., Spectrochim. Acta Part B, 1996, 51, 779.waste samples such as cement or concrete was evaluated. The 16 Becker, J. S., Seifert, G., Saprykin, A. I., and Dietze, H.-J., J. Anal. detection limits are more than one order of magnitude lower At. Spectrom., 1996, 10, 643. 17 Becker, J. S., and Dietze, H.-J., J. Anal. At. Spectrom., 1995, 10, 637. using double-focusing sector field ICP-MS in comparison with 18 Kobota, M., Fudagawa, N., and Kawase, A., Anal. Sci., 1989, LA–ICP-QMS and reach values <1 ng g-1 (e.g., 233U 5, 701. 0.6 ng g-1; 237Np 0.5 ng g-1). 19 Reed, N. M., Cairns, R. O., Hutton, R. C., and Takaku, Y., J. Anal. At. Spectrom., 1994, 9, 881. 20 Tanner, S. D., J. Anal. At. Spectrom., 1995, 10, 905. CONCLUSIONS 21 Yamasaki, S.-I., and Tsumura, A., Anal. Sci., 1991, 7, 1135. 22 Yamasaki, S.-I., Tsumura, A., Takaku, Y., Microchem. J., 1995, Double-focusing ICP-MS is a useful tool for the ultrasensitive 49, 305. multi-element determination of trace impurities in environmen- 23 Moens, L., Verrept, P., Dams, R., Greb, U., Jung, G., and tal materials and for the characterization of high-purity Laser, B., J.Anal. At. Spectrom., 1994, 9, 1075. samples. The high sensitivity and the very low detection limits 24 Jakubowski, N., Tittes, W., Pollmann, D., Stuewer, D., and achieved, in the pg l-1 concentration range, used in the low- Broekaert, J. A. C., J. Anal. At. Spectrom., 1996, 11, 797. resolution mode, are the most important features. In many 25 Takaku, Y., Masuda, K., Takahashi, T., and Shimamura, T., J.Anal. At. Spectrom., 1994, 9, 1385. cases, if the mass resolution required for the separation of 26 Begerow, J., Turfeld, M., and Dunemann, L., J. Anal. At. analyte ions and disturbing molecular ions is higher than was Spectrom., 1996, 11, 913. practically possible, a further decrease in molecular ion forma- 27 Boonen, S., Vanhaecke, F., Moens, L., and Dams, R., Spectrochim. tion and matrix eects by special sample introduction methods Acta, Part B, 1996, 51, 271.or preparation of sample is required for the determination of 28 Hoppstock, K., Becker, J. S., and Dietze, H.-J., J. Anal. At. ultratrace elements. The precision of isotopic analysis by Spectrom., submitted for publication. 29 Becotte-Haigh, P., Tyson, J. F., Denoyer, E., and Hinds, M. W., double-focusing ICP-MS is about one order of magnitude Spectrochim. Acta, Part B, 1996, 51, 1823. better than that of quadrupole ICP-MS and can be improved 30 Laly, S., Nakagawa, K., Arimura, T., and Kimijima, T., to become comparable to that of TIMS using a multiple Spectrochim.Acta, Part B, 1996, 51, 1393. collector for the simultaneous detection of ion beams of 31 Montaser, A., Tan, H., Ishii, I., Nam, S.-H., and Cai, M., Anal. isotopes. Chem., 1991, 63, 2660. 32 Vanhaeke, F., Van Holderbeke, M., Moens, L., and Dams, R., J. Anal. At. Spectrom., 1996 11, 543. The authors would like to thank V. K. Panday (New Delhi, 33 Sakata, K., and Kawabata, K., Spectrochim.Acta, Part B, 1994, India) and R. S. Soman (Cincinnati, USA) for investigations 10, 1027. on matrix separation carried out at this laboratory. 34 Kawabata, K., Yamanaka, K., and Kishi, Y., paper presented at the 1997 European Winter Conference on Plasma Spectrochemistry January 12–17 1997, Ghent, Belgium. REFERENCES 35 Tanner, S. D., Paul, M., Beres, S. A., and Denoyer, E. R., At. Spectrosc., 1995, 16, 16. 1 Houk, R.S., Acc. Chem. Res., 1994, 27, 33. 36 Georgitis, S., Stroh, A., and Tyler, G., paper presented at the 5th 2 Inductively Coupled Plasma Source Mass Spectrometry, ed. International Conference on Plasma Source Mass Spectrometry, Montaser, A. VCH, New York, 1996. September 1996, Durham, UK. 3 Hieftje, G. M., J. Anal. At. Spectrom., 1996, 11, 613. 37 Speakman, J., Turner, P. J., Haines, R. C., and Merren, T. O., 4 Barnes, R. M., Fresenius’ J. Anal. Chem., 1996, 355, 433. paper presented at the 1997 European Winter Conference on 5 Chen, Z., Fryer, B.J., Longerich, H. P., and Jackson, S. E., Plasma Spectrochemistry, January 12–17, 1997, Gent, Belgium. J. Anal. At. Spectrom., 1996 11, 805. 38 Bradshaw, N., Hall, E. F. H., and Sanderson, N. E., J. Anal. At. 6 Nelms, S. M., Greenway, G. M., and Koller, D., J. Anal. At. Spectrom., 1989, 4, 801. Spectrom., 1996, 11, 907. 39 Morita, M., Ito, H., Uehiro, T., and Otsuka, K. Anal. Sci., 1989, 7 Panday, V. K., Becker, J.S., and Dietze, H.-J., Anal. Chim. Acta, 5, 609. 1996, 329, 153. 40 Giessman, U., and Greb, U., Fresenius’ J. Anal. Chem., 1994, 8 Panday, V. K., Hoppstock, K., Becker, J. S., and Dietze, H.-J, At. 350, 186. Spectrosc., 1996, 17, 98. 41 Narasaki, H., and Cao, J. Y., Anal. Sci., 1996, 12, 623. 9 Taylor, D. B., Kingston, H. M., Nogay, D. J., Koller, D, and 42 Cromwell, E. F., and Arrowsmith, P., J. Am. Soc. Mass Spectrom., Hutton, R., J. Anal. At. Spectrom., 1996, 11, 187. 1996, 7, 458. 10 Barinaga, C. J., Eiden, G. C., Alexander, M. L., and Koppenaal, 43 Halliday, A. N., Lee, D.-C., Christensen, J. N., Walder, A. J., D. W., Fresenius’ J. Anal. Chem., 1996, 355, 487. Freedman, P. A., Jones, C. E., Hall, C. M., Yi, W., and Teagle, D., 11 McLaerty, F. W., Senko, M. W., Little, D. P., Wood, T. D., Int. J.Mass Spectrom. Ion Processes, 1995, 146/147, 21. O’Connor, P. B., Speit, J. P., Chorush, R. A., and Kelleher, N. L., 44 Turner, P., and Wortel, N., 1997 European Winter Conference Adv.Mass Spectrom., 1995, 13, 115. on Plasma Spectrochemistry, January 12–17, 1997, Gent, Belgium, 12 Kluge, H.-J., Bollen, G., Carlberg, C., Moore, R. B., and Quint Q. W., Adv. Mass Spectrom., 1995, 13, 420. Exhibition, Information Note on Iso-PlasmaTrace. 888 Journal of Analytical Atomic Spectrometry, September 1997, Vol. 1245 Ying, J.-F., and Douglas, D. J., Rapid Commun. Mass Spectrom., 57 Isotopic Composition of the Elements, Pure Appl. Chem., 1991, 1996, 10, 649. 63, 991. 46 Yamasaki, S.-I., Tsumura, A., and Takaku, Y, Microchem. J., 58 Vanheacke, F., Moens, L., and Dams. R., Anal. Chem., 1996 68, 567. 1994, 49, 305. 59 Vanheacke, F., Moens, L., Dams, R., Papadakis, I., and Taylor, P., 47 Yamasaki, S.-I., and Tsumura, A., Anal. Sci., 1991, 7, 1135. Anal. Chem., 1997, 69, 268. 48 Takaku, Y., Masuda, K., Takahashi, T., and Shimamura, T., 60 Walder, A. J., Koller, D., Reed, N. M., Hutton, R. C., and J. Anal. At. Spectrom., 1993, 8, 687. Freedman P. A., J. Anal. At. Spectrom., 1993, 8, 1037. 49 Riondato, J., Vanhaeke, L., Moens, L., and Dams, R., J. Anal At. 61 Walder, A. J., Platzner, I., and Freedman, P. A., J. Anal. At. Spectrom., in the press. Spectrom., 1993, 8, 19. 50 Kim, C.-K., Seki, R., Morita, S., Yamasaki, S. I, Tsumura, A., 62 Walder, A. J., Platzner, I., and Freedman, P. A., Int. J. Mass Takaku, Y., Igarashi, Y., and Yamamoto, M., J. Anal. At. Spectrom. Ion Processes, 1993, 8, 19. Spectrom., 1991, 6, 205. 63 Taylor, P. D. P., DeBie`vre, P., Walder, A. J., and Entwistle, A., 51 Kerl, W., Becker, J. S., Dietze, H.-J., and Dannecker, W., J. Anal. J. Anal. Atom. Spectrom., 1995, 10, 395. At. Spectrom., 1996, 10, 723. 64 Lee, D.-C., Halliday, A. N., Int. J. Mass Spectrom. Ion Processes., 52 Becker, J. S., Ku� ppers, G., Carsughi, F., Kerl, W., Schaal, W., 1995, 146/147, 35. Ullmaier, F., and Dietze, H.-J., Ber. Forschungszentrum Ju�lich, 65 Walder, A. J., Furuta, N., Anal. Sci., 1993, 9, 675. 1996, No. 3272, 143. 66 Walder, A. J., Abell, I. D., Platzner, I., and Freeman, P. A., 53 Kerl, W., Becker, J. S., Dietze, H.-J., and Dannecker, W., Ber. Spectrochim. Acta Part B, 1993, 48, 397. Forschungszentrum Ju�lich, 1996, No. 3272, 176; paper presented 67 ChristeJ. N., Halliday, A. N., Lee, D. C., and Hall, C. M., at the 1997 European Winter Conference on Plasma Spectrochemistry, January 12–17, 1997, Gent, Belgium. Earth Planet. Sci. L ett., 1996, 136, 79. 54 Yamamoto, M., Syarbaini, Kofuji, K., Tsumura, A., Komura, K. 68 Gastel, M., Becker, J. S., and Dietze, H.-J., Spectrochim. Acta, Ueno, K., and Assinder, D. J. J. Radioanal. Nucl. Chem., 1995, Part B, submitted for publication. 197, 185. 55 Becker, J. S., Saprykin, A. I., and Dietze, H.-J., Int. J. Mass Paper 7/02178J Spectrom. Ion Processes, in the press. Received April 1, 1997 56 Kerl, W., Becker J. S., Dietze, H.-J., and Dannecker, W., Fresenius’ J. Anal. Chem., 1997, 359, 4. Accepted June 3, 1997 Journal of Analytical Atomic Spectrometry, September 1997, Vol. 12 889