Ben Fairman*a, Michael W. Hindsb, Simon M. Nelmsc, Denise M. Pennyd and Phill Goodalle
aLaboratory of the Government Chemist, Queens Road, Teddington, Middlesex, UK TW11 0LY
bRoyal Canadian Mint, 320 Sussex Drive, Ottawa, Ontario, Canada K1A 0G8
cIRMM, Reteiseweg, Geel, Belgium B-2440
dShell Research and Technology Centre, Thornton, PO Box 1 Chester, UK CH1 3SH
eBNFL, Sellafield, Seascale, Cumbria, UK CA20 1PG
First published on 28th November 2000
This Atomic Spectrometry Update is the latest in an annual series appearing under the title `Industrial Analysis'. This year's review has followed the changed format introduced last year. Further changes may be made in the near future to reflect the growing interest in certain areas such as semiconductor materials and a continuing decrease in technical advances being reported under other traditional headings.
There has been considerable interest in XRF as a tool for the non-destructive analysis of metallic art and historical objects. Laser ablation continues to be explored for metal analysis. Laser ablation ICP-AES was used to differentiate between coins from different countries based on the elemental composition profiles (or fingerprints).
Improvements to XRF instrumentation and methodology have meant that analysis of used oil reported via this technique is on the increase. The analysis of coal and its by-products once again dominates the Fuels section. Various sample preparation procedures and a host of different analytical techniques have been used for its analysis. Pre-concentration using on-line column techniques coupled with atomic spectrometry is very important for trace metal determination. 8-Hydroxyquinoline (8HQ) has been thoroughly investigated and reported by many as an excellent chelating agent for organic based solutions.
There have been some interesting developments this year which impact on inorganic chemicals analysis in industrial applications, particularly in ICP-MS. Elimination and reduction of spectral interferences using collision cell technology in ICP-MS is becoming a commercial reality, as evidenced by an increasing number of papers dealing with the technique.
Materials Control and Accountancy (MCA) is of utmost importance in the nuclear industry. Analysis, undertaken for the purposes of MCA, provides a `Gold Standard' for any laboratory in terms of accuracy, precision and reliability. This crucial area has seen some development in the period covered by this review for nuclear materials analysis.
This year, coupling to a variety of detectors has proved to be a popular use of ETV for the analysis of refractory samples. Finally, one major disappointment and surprise this year has been the lack of high quality papers and articles which could be selected to grace our Catalysts section.
Element | Matrix | Technique;atomization;presentationa | Sample treatment/comments | Ref. |
---|---|---|---|---|
a Hy indicates hydride and S, L, G and Sl signify solid, liquid, gaseous or slurry atomization, respectively. Other abbreviations are listed elsewhere. | ||||
Al | Aluminum alloys | AA;F;L | Samples dissolved on-line with a high current density anodic electro-dissolution unit | 34 |
Al | Steel | AA;F;L | A 0.5 g sample was dissolved in 20 ml HNO3 (1∶1), cooled, and 20 ml of 40 g l−1 NaOH were added. The resulting precipitate was filtered and re-dissolved in 20 ml HCl (1∶1), 6 g ammonium acetate, 10 ml triethanolamine and water to dilute to 100 ml. An 8-fold sensitivity enhancement was reported with an air–acetylene flame | 35 |
As | Steel | AA;ETA;L | Dissolved samples were treated with ascorbic acid and potassium iodide to reduce analytes to the trivalent state. Analytes were complexed by ammonium dithiophosphoric acid O,O-diethyl ester and adsorbed onto activated carbon. They were then eluted by a small volume of nitric acid | 36 |
As | Steel | MS;ICP;L | Sample dissolved in HCl–HF and passed through a desolvating microconcentric nebulizer to the ICP-MS. Rh added as an internal standard | 37 |
B | Steel | MS;ICP;L | Boron nitride in the dissolved sample was decomposed by treatment with H2SO4–H3PO4 fuming at 290![]() | 38 |
Bi | Steel and aluminium | AA;FI-ETA;L | Bismuth is separated from metal matrices by complexation with ammonium dithiophosphoric acid O,O-diethyl ester and adsorbed onto a mini-column of activated carbon. The Bi complex was eluted with ethanol and was collected in an autosampler cup for determination by ETAAS. Limit of detection was 0.048 µg l−1 | 39 |
C | Steel | XRF;—;S | Sample was collected and polished under low pressure (6 Pa). Calibration linear up to 1% C and limit of detection 0.018% | 40 |
Cr | Metal alloys | AA;ETA;L | Dissolved and diluted sample introduced onto a boron coated pyrolytic graphite tube | 41 |
Cu | Brass and bronze | AA;F;L | Samples dissolved on-line with a high current density anodic electro-dissolution unit | 34 |
Deuter- ium | Titanium | MS;SIMS;S | None | 42 |
In | Aluminum alloy | AA;ETA;L | Sample (0.5 g) was dissolved in 10 ml HCl (left overnight), heated and HNO3 added. Sample diluted to volume with water and determined by ETAAS with a tungsten coated L'vov platform. Limit of detection 8 pg | 43 |
Ni | Aluminum alloys and steel | AA;F;L | After sample dissolution nickel is retained quantitatively on 2-(5-bromo-2-pyridylazo)-5-diethylaminophenol–ammonium tetraphenylborate with microcrystalline naphthalene. After filtration, the nickel complex is dissolved in dimethylformamide and determined by FAAS. Precision is 1.5% RSD for 5 µg g−1 Ni. A column version is also described | 44 |
P | Cast iron | XRF;—;S | Sample was ground to 300 mesh and 2.4 g was mixed with a binder of methylcellulose in a 5∶1 ratio, then compacted at 30 kN to form a disc (25 mm id) | 45 |
Pb | Copper alloys | AA;F;G | Dissolved samples (1–100 mg) in HNO3–HCl and evaporated. Residues re-dissolved in 1 M HCl and 1 µl aliquot introduced into a metal vapour elution column (Mo), then heated to generate metal vapours (gas) | 46 |
Pb | Tin plate | AE;ICP;L | Sample (5 × 5 cm) was heated with 5 ml 0.5 M NaOH and 2 ml 30% H2O2 for 5 min to dissolve the plating layer. The digest was further boiled with 5 ml HCl (1∶1) to remove the remaining H2O2 and diluted to 25 ml | 47 |
Pt | Palladium–platinum alloys | XRF;—;S | None | 19 |
S | Steel | AE;ICP;L | Sample dissolved in a flow through electrolytic dissolution system against a graphite counter electrode in 6 M HCl. Precision: 3.5% RSD for 25 µg g−1 S | 17 |
Sb | Brass | AF;HG-F;L | Dissolved sample neutralized then NaOH added to 8 g l−1. Interfering elements Cu, Fe, Co, and Ni are removed by precipitation | 48 |
Sb | Steel | AA;ETA;L | Dissolved samples were treated with ascorbic acid and potassium iodide to reduce analytes to the trivalent state. Analytes complexed by ammonium dithiophosphoric acid O,O-diethyl ester and sorbed onto activated carbon. Eluted by a small volume of nitric acid | 36 |
Th | Thallium–iron slags | AA;F;L | Sample (0.1–0.3 g) was dissolved in 15 ml 30% HNO3 and evaporated to near dryness. The residue was dissolved in 5 ml 50% HCl and diluted to 50 ml with water | 49 |
V | Metal alloys | AA;ETA;L | Dissolved and diluted sample introduced onto a boron coated pyrolytic graphite tube | 41 |
Zn | Brass and bronze | AA;F;L | Samples dissolved on-line with a high current density anodic electro-dissolution unit | 34 |
Zn | Copper alloys | AA;F;G | Dissolved samples (1–100 mg) in HNO3–HCl and evaporated. Residues re-dissolved in 1 M HCl and a 1 µl aliquot introduced into a metal vapour elution column (Mo), then heated to generate metal vapours (gas) | 46 |
Zn | Zinc based metal coatings | AE;GD;S | None; depth profiling | 50 |
Various | Antimony | MS;GD;S | None; relative sensitivity factors derived from steel reference materials | 51 |
Various (8) | Bullet leads | MS;ICP;L | Samples (25 mg) were dissolved in 2 ml 25% HNO3 and 0.5% HF. Lead was precipitated with 8 ml H2SO4. Supernatant (1 ml) was mixed with 1 ml 1250 µg l−1 In (internal standard), 1 ml 0.5% HF and diluted to 25 ml with 1% HNO3 | 52 |
Various | Cadmium | MS;ICP and GD;L and S | Analytes were adsorbed on Dowex-50X8 from 0.2 M HNO3 + 0.25 M HCl sample solutions. Matrix separation was carried out by eluting analytes with 6 M HCl. GD-MS of solid samples was also described and good agreement was reported between the two methods | 53 |
Various (8) | Cadmium | MS;ICP;L | Sample (0.05 g) was dissolved in 5 ml 10% HNO3 at low temperature, treated with 0.5 µg Y and 0.5 µg Bi (internal standards) and diluted to 50 ml with water. Detection limits 0.005–0.052 µg l−1 for As, Co, Ga, Mn, Pb, Sr, Tl and V | 54 |
Various (7) | Copper alloy statuettes | MS;ICP;L | Small samples (1–10 mg) were taken and dissolved in HNO3–HCl and diluted to 100 ml with water | 33 |
Various | Copper | MS;ICP;S and L | Dissolved samples were analysed by continous and flow injection ICP-MS. Solid alloys were also analysed by laser ablation ICP-MS. Methods were compared | 29 |
Various (4) | Copper alloys | MS;ICP;L | Dissolved samples (HCl–HNO3) were mixed with In as an internal standard. Time of flight mass spectrometric detection. Detection limits: 0.7, 2.5, 11, 15 µg g−1, for Pb, Sn, As and Zn, respectively | 32 |
Various | Copper–nickel, neodymium–aluminum alloys, iron powders | XRF;—;S | Powders were ground and sieved to 400 mesh. The powders were mixed with cellulose and pressed into 30 mm diameter pellets at 236.25 MPa | 55 |
Various (5) | Gold alloys | AE;ICP;L | Sample (20 mg) dissolved in 20 ml aqua regia (3∶1 HCl–HNO3) with heat. Internal standards Y (20 µg l−1) or In (200 µg l−1) added upon dilution | 56 |
Various (20) | High purity tungsten | AE;spark;S | Sample ignited to form WO3, then 500 mg mixed with 250 mg buffer agent containing 5% Li2CO3 and 0.03% Ga2O3, and packed into an electrode | 57 |
Various (8) | Iron | MS;ICP;L | Iron was separated from the analytes using Chrome Azurol B (CAB) to form an insoluble chelate with Fe. A membrane filter was used for separation and a surfactant was added for solubilization of the other metal–CAB chelates | 58 |
Various (9) | Iron | MS;ICP;L | Sample (0.1 g) was dissolved with 1.5 ml HCl, 1.5 ml HNO3, and 3 ml water with heat. The solution was cooled, diluted with 4 ml HCl and 10 ppb Y was added as a marker for recovery. The solution was extracted twice with 10 ml portions of MIBK. The aqueous phase was evaporated to near dryness and additions of 0.1 ml H2SO4, 5 ml HNO3, 5 ml water and 10 ppb In (internal standard) | 59 |
Various (3) | Iron and steel | MS or AE;HG-ICP;L | Dissolved samples were put through a continuous hydride generator to either an ICP-MS or ICP-AES. Detection limits: for As, Bi and Sb 0.5, 0.8, 0.5 ng ml−1 for ICP-AES; and 0.03, 0.02, 0.03 ng ml−1 for ICP-MS | 60 |
Various (2) | Coated steel plates | AE;ICP;L | Samples (50 × 40 mm) of the aluminium–zinc coated steel were stripped of the coating with 60 ml 30% HCL. The solution was mixed with 3 ml of 5% HCl and diluted to 100 ml with water | 61 |
Various (9) | Steel | MS;spark ablation ICP;S | Samples were finished with a surface grinder using a 60 grit abrasive zirconium oxide belt | 6 |
Various (4) | Steel | AE;laser;S | None; detection limits: 6, 50, 80, and 80 µg g−1 for Cr, Ni, C and Si, respectively | 9 |
Various | Steel | MS;FI-ICP;L | The matrix elements were removed from the dissolved solution using a micro-electrolytic cell within the flow injection system | 62 |
Various (11) | Tantalum | MS;ICP;L | The dissolved sample was passed through an on-line ion exchange column. Different groups of analytes were selectively eluted by: (1) 2 M HCl–0.1 M HF (Be, Al, Ti); and (2) 1 M HNO3–0.1 M HF (Cr, Ni, Nb, Mo, Sn, W, Th, U) | 63 |
An overview of the application of micro-beam techniques in the steel industry was written4 which emphasised micro-analysis of surfaces and precipitates. A method for obtaining multi-layered thickness profiles by micro-proton induced X-ray emission (PIXE) spectrometry was reported using TiN-coated steels.5
A spark ablation source was coupled to an ICP-MS for the determination of minor and trace elements in various steel samples.6 Major component matrix elements of 57Fe and 55Mn were used as internal standards for a wide range of elements (Al, B, Co, Cu, Mn, Nb, P, Si and V). A restrictive path was designed to minimise skimmer cone blockage from the quantities of spark eroded material. Detection limits were reported to be below 1 µg g−1.
There were a number of reports concerning solid sample atomic emission spectrometry. An overview of the application of lasers to analysis problems with emphasis on the steel industry appeared in the literature.7 A study of the effect of high temperatures (up to 1200°C) on steel analysis by LIBS, indicated that analytical signals were temperature dependent.8 A slag layer of variable thickness formed at temperatures above 600
°C. This was composed mainly of oxides of chromium, iron, and manganese. In another study,9 LIBS was used for the determination of carbon, silicon and chromium. The lens to sample distance was observed to be a relevant parameter in the LIBS experiment. A study of the saturation effects and the amounts of material ablated10 indicated that ablation efficiency decreased at lower laser energies due to plasma shielding and the energy required for air breakdown. A more efficient ablation rate can be obtained by focusing the beam inside the material free expansion zone of the created plasma.
A spark ablation device was interfaced to a microwave plasma atomic emission spectrometer and applied to the analysis of steel and brass.11 Detection limits were in the µg g−1 range but unfortunately these were 20 times higher than those obtained with a spark ablation-ICP-AES high resolution, sequential spectrometer. A glow discharge-atomic emission spectrometer with a specially designed glove box for the analysis of steel samples was tested with steel samples.12 This preliminary study showed that carbon, oxygen and nitrogen can be determined with detection limits of 10, 20, and 40 µg g−1, respectively.
A recent patent was issued related to AAS whereby components in molten metal can be determined from the absorption of specific wavelength light by vapour above the melt.13
Solution based methods are mainly summarized in Table 1 for ferrous samples. However, some work warrants special mention. Low energy electron induced X-ray spectrometry and SEM were used to characterize residues from dissolved titanium stabilized steels.14 Compounds TiN and TiC were identified from the shape and structure of the X-ray peak (L transitions). Quantification was also possible by comparing the spectra of known mixtures.
A review article of the application of high resolution ICP-MS to the steel industry15 presented the view that detection limits were not governed by the spectrometer but by sampling and sample preparation (i.e., reagents, dissolution procedures, environment). Another review focused on sample preparation, pre-treatment and separation methods for steel analysis by ICP-MS.16
Sulfur in steel was determined via on-line electrolytic dissolution against a graphite counter electrode in an acidic medium, followed by introduction into an ICP-AES.17 The analysis time was reported to be only sixty seconds per sample.
There has been considerable interest in XRF as a tool for the non-destructive analysis of art and historical objects. An overview of this XRF application described a range of devices from synchrotron beam lines to a transportable spectrometer.20 The use of a small XRF instrument for in-situ analysis of objects was described.21 The small active excitation window (8 mm2) was focused reproducibly by aligning two sighting lasers attached to beam exit and entrance components. The small beam diameter permits elemental mapping and specific site analysis of these valuable samples.
The analysis of alloy nano-particles (<100 nm diameter) of Pt–Rh and Pt–Re by X-ray emission spectrometry was described.22 Quantitative and phase separation information was obtained from this technique, which was coupled with an electron microscope. Proton induced X-ray emission (PIXE) has also been actively used in the analysis of rare materials such as: ancient coins;23,24 and gold artefacts.25,26 The small beam diameter permits elemental mapping and specific site analysis of these valuable samples.
Laser ablation continues to be explored for metal analysis. Laser ablation (LA)ICP-AES was used to differentiate between coins from different countries based on the elemental composition profiles (or fingerprints) and the depth profiles.27 Similarly, binary Cu–Zn alloys were studied using LA-ICP-MS.28 Different ablation rates were observed for various alloy compositions. Linear calibration curves were made by normalizing the Zn intensity to that of Cu or to the crater volume. Trace impurities in high purity copper were determined by this technique with standards made from high purity copper powder mixed and pressed with solution standards.29 This approach gave results that were within 15% of solution based ICP-MS values. This shows promise as a semi-quantitative screening method and the method of standard production has created possibilities for situations where no reference materials exist.
Methods involving dissolved metal samples continue to be a relevant part of metal analysis. The majority of papers involving AAS, ICP-AES and ICP-MS focus mainly on the sample preparation (see Table 1); however, some papers are worth mentioning.
There were two reports of isotope ratio measurements of metallic historical objects. In one study, 207Pb∶206Pb ratios were determined in bronze objects30 by ICP-MS. In another, thermal ionization mass spectrometry was applied to the determination of magnesium isotope ratios in magnesium metal reference material.31 An improvement in the uncertainty by a factor of 3–4 times over ICP-MS was reported.
The use of different detection systems were reported. Time of flight mass spectrometry was used to the determine a variety of elements in copper alloys.32 Lead isotopic and elemental analysis of copper alloy statuettes were obtained by double focusing sector field ICP-MS.33 Uncertainty estimates for Pb isotope ratio measurements were lower than for those obtained by quadrupole-based ICP-MS.
An automated system was reported for the determination of aluminum, copper, and zinc in non-ferrous alloys by FAAS following on-line sample dissolution with an electro-dissolution unit.34 The system was reported to analyse 50 samples per hour (on average).
Sun Xiao-Juan et al.69 discuss alternative sample preparation to determine Pb in gasoline instead of the traditional methods, e.g., the iodine monochloride method. The use of microwave extraction–nitric acid to convert organic Pb into inorganic Pb is reported. The results compare very well with traditional methods and offer an accurate, simple and quick procedure, which does, however, rely on investment in microwave digestion equipment.
The environmental impact of petroleum products is further investigated via the use of in-situ laser induced fluorescence (LIF), which has been applied to contaminated land.70 Various calibration sets were prepared using laboratory reference oil, fuel oil, gas oil, diesel fuel, Brent and German crude oil, all of which were extensively characterized with regard to their photophysical properties. All standards were prepared in a well characterized soil. Although of 30 soils analysed 23 of them were deemed to be contaminated, quantitative in-situ analysis of petroleum products contaminated soil encounters many difficulties attributed to the fact that petroleum products per se are very complex, ill defined analytes in a complex soil matrix. Hence, the limited variation in contaminants and a single soil matrix for standard preparation must add to the uncertainty. However, for the work carried out thus far comparative data with IR was very encouraging.
A review with 13 references of metal microanalysis in petroleum fractions by ICP-MS has been reported.71 Four different sample routes are compared and contrasted and the advantages and disadvantages of the methods described.
As previously discussed in the 1999 review alternative methods that preclude the need for sample dissolution procedures are favoured by many for coal/fly ash analysis. In this review period several techniques have been used to analyse coal/fly ash in the solid state, although it is fair to say that a certain amount of sample preparation is still required. Zhanget al.72 have developed a method for the determination of Ge in coal ash using WDXRF. The sample was presented to the instrument as a glass bead, and the Compton peak was used for matrix correction. Using synthetic coal ash standards a calibration line with 0.999 coefficient was established, and the method was applicable in the range 10–12000 mg kg−1 Ge in coal ash. The accuracy of the method was evaluated by comparison with results attained by ICP-MS, relative differences between the two methods being <10%. The study found that, after combustion, Ge was enriched in the fly ash rather than the bottom ash, so much so that the fly ash may well be considered as an economical source of Ge.
Two other papers73,74 report the use of XRF for coal and fly ash analysis, the latter reporting on the use of on-line XRF for S, Ca, Fe and Ti determination and the former detailing the use of EDXRF and WDXRF for assessing the environmental impact of coal and ash. This required the samples to be crushed, dried, ground and pressed, further pelletization with boric acid being required for WDXRF. Both papers report excellent accuracy and precision.
Laser ablation ICP-MS has been employed as a rapid method75 for the determination of various elements present at low (ppb) concentrations in bituminous coals. Uncertified coals were used for the calibration standards (representative standards being one of the major drawbacks of LA-ICP-MS); these were subsequently verified using the South African certified coals Sarm 18,19 and 20. A critical analysis of the data showed that, for many of the elements, the results were both accurate and precise.
Pelletization of ashed coal samples was also used in the successful use of rf GD-AES in trace metal analysis.76 However, the evaluation of plasma stability showed that the temporal profile of aluminium decreased with time; this was attributed to water molecules trapped in the ash. To overcome this the samples were prepared as a glass bead, similar to that used in WDXRF analysis.
Silva et al.77 report on the extension of the method reported in 1999, Pb in coal by slurry sampling GF-AAS, to include Cd and Cu. The particle size was reduced from 45 to 37 µm and the diluent remained as 5% (v/v) HNO3, 0.05% Triton X-100 and 10% ethanol. Homogenization was via manual stirring or via the use of ultrasonic agitation. Reference coals used to evaluate the method showed a better correlation when the sample particle size was reduced to 37 µm.
Three papers discuss various aspects of coal analysis after preparing the sample solution via microwave digestion: they are detailed as follows. Richaud et al.78 found that microwave digestion in HNO3 of liquefaction extracts prepared from Argonne Premium Coals and coal pitch tar was a suitable method for determining trace metals in the said extraction solutions by ICP-MS. Sample sizes as small as 3–20 mg were analysed. They reported that the trace element distribution found in the extracts bore little relationship to the corresponding distributions in the starting materials. Mossbauer spectroscopy of the extracts indicated that the high Fe concentrations corresponded to the presence of organometallic Fe compounds and not to pyritic Fe. A sequential microwave digestion method for the determination of sulfate S, pyritic S and organic S in bituminous and sub-bituminous coals with the use of ICP -AES has been reported;79 full details were given. Finally microwave digestion was employed in the determination of Hg in coal. FI-CV-AAS gave a detection limit of 20 ng g−1 and for FI-CV-ICP-MS the detection limit was 4 ng g−1.80
Shiowatana et al.81 report on the use of carbon adsorption and slurry sampling ET-AAS to determine Hg in natural gas liquid and condensate. A 500 ml volume of natural gas liquid was stirred for 10 min with 0.5 g of activated carbon, which is removed via filtration. After drying in air an accurately weighed mass of carbon was introduced into a pre-weighed amount of either 2% boiled tapioca flour or 75% glycerol containing 5 mg ml−1 of Triton X-100, then a homogenized suspension was prepared using a vortex mixer. From this a 15 µl portion of slurry was injected into a pyrolytically coated graphite furnace together with 5 µl of matrix modifier (1 mg l−1 of Pd). This was dried at 120°C for 10 s, pyrolysed at 250
°C for 20 s and step atomized at 1300
°C. The absorbance was measured at 253.7 nm with deuterium background correction. The limit of detection achieved for natural gas liquid was 2 ng ml−1, spike recovery of Hg (added as diphenylmercury) was >90%. Comparative results using acid digestion ETAAS and CV-AAS were in good agreement.
Jose Luis Burguera et al.,82 describe the use of an on-line emulsification FI-ET AAS system to determine Cr in lubricating oils. A simple quick procedure for the on-line preparation of a one phase emulsion in an FI-ETAAS system is detailed. The entire system was computer controlled and independent of the spectrometer. A 1 ml plug of sample solution was injected into a carrier stream of hexane and was subsequently mixed with streams of 3.8% (m/v) NaCl, 5.0% (v/v) Na dodecylsulfate and 5.0% (v/v) sec-butanol. The flowing solutions were subject to sonification, which improved the emulsion's stability. The ETAAS programme parameters were also detailed. The reporter claims that to the best of his knowledge the literature has not reported work on w-o-w emulsions for ETAAS determinations. Good agreement between the certified and found results for two NIST certified materials was attained.
Sn determination in used oil was also reported;83 here the application of fractionary factorial design to the determination of Sn in lubricating oils by continuous flow HG-AAS is described. Microwave digestion was used to prepare the sample, which included a 4 stage programme. The level of Sn studied was in the range 33–108 µg l−1, and in all cases the results were higher than those obtained by a simple dilution in kerosene.
Langer et al.84 interestingly detail a method for the direct analysis of new and used lubricating oils using ETV-ICP-MS. The sample was introduced into the ETV without dilution, and quantification was carried out via aqueous standards. They report this methodology as being reproducible and reliable. The `frozen drop method', where a drop of oil sample falls onto a pellet of dry ice, is solidified and then transferred manually into a modified GF, was investigated as an alternative to conventional acid digestion.
Microwave digestion was once again used for the sample preparation stage for the determination of wear metals in marine lubricating oils.85 A four-stage procedure using nitric acid and hydrogen peroxide is detailed and the reporters claim to have successfully totally digested a larger sample mass than that recommended by the manufacturers. They suggest that this procedure could be an interesting alternative to classical ashing procedures. Comparative data with an alternative technique is given. The reviewer, however, wonders if this is a viable alternative since the cost of advanced microwave systems is not insignificant and systems are still limited to a relatively low sample throughput.
A study to determine a possible link between toxic elements present in jet engine lubricating oils and two hydraulic fluids with the symptoms (dizziness, nausea, lethargy, etc.) of flight crews was undertaken.86 ICP-MS was used to determine the elemental concentrations of Mg, Ti, Cr, Mn, Co, Sr, Zn and W in the oils after they were prepared for analysis by one of the following two routes: (1) dissolution in HNO3 in a closed Parr bomb reaction vessel; (2) dissolution in 60% IPA, 1% Triton-100 and USN-membrane desolvation. Differences between the oils and hydraulic fluids were observed; however, none of the toxic elements, which could be related to the symptoms, were found.
Improvements to XRF instrumentation and methodology has meant that used oil analysis reported via this technique is on the increase.
One paper87 reports the work carried out to improve the ASTM D4927 (WDXRF) method to extend its coverage to 13 elements. The determination of empirical alphas are reported and subsequently compared to the calculated alphas.
The use of EDXRF to monitor the oil condition of the F404 engine on board the F18 weapon system is detailed and compared with the results of a rotrode atomic emission system;88 the benefits of such as system are given.
Pt reforming catalysts are easily poisoned by increased levels of Pb, hence sensitive analytical procedures are required for Pb in crude oils. In preparation for the analysis by ETAAS, the pyrolysis and atomization characteristics of lead alkylsulfon-ate, lead 4-cyclohexane butyrate, tetraethyllead and lead in fuel oil were studied.89 The best procedure is detailed in the paper. The limit of detection from a 20 µl injection was 0.25 ng g−1 with a recovery of 85–106%.
The disposal of combustion products from waste oils can have a detrimental impact on the environment. Concerns are raised over the metal and non-metal content of such oils and also the distribution of metals versus particle size. In this review period three papers have covered various aspects of waste analysis: Study of the distribution of Pb, V, Cr, Ni, Cd, Cu and Fe in particles formed from combustion;90 Low level determination of non-metals (e.g., chlorine, bromine, etc.) by ICP-AES;91 and Determination of heavy metals and S in waste oils by EDXRF.92
This year, the papers submitted in this section have shown a wider coverage in applications and techniques employed in comparison with the last two review periods.
Pre-concentration using on-line column techniques coupled with atomic spectrometry is very important for trace metal determination. 8-Hydroxyquinoline (8HQ) has been thoroughly investigated and reported by many as an excellent chelating agent, showing the best selectivity of the commercial resins. This coupled with the amount of literature available made it a good choice for Howard et al.94 to use as the comparative base for a new chelating agent, poly-(L-cysteine) (PLC). PLC and 8HQ were immobilized on controlled pore glass and used in a flow injection system for the separation of Cd, Pb and Cu from synthetic sea-water, Co and Ni matrices and CRM sea-water. They reported that both resins quantitatively recovered 50 µg l−1 Cd and Pb in synthetic sea-water. However, 8HQ showed low recoveries, 2–4% and 40–50% for the separation of 50 µg l−1 of Cd and Pb, respectively, from a 10000 ppm excess of Ni and Co. PLC maintained quantitative recoveries. Neither 8HQ nor PLC maintained quantitative recoveries for Cu2+. Much more information is detailed in the paper regarding further experiments which illustrate the limitations of the systems plus the extra information available, e.g., stability constants.
A plethora of papers were received during this review period for the analysis of traditional Chinese medicines (TCMs). Three papers95–97 detail the determination of arsenic by ICP-MS, two illustrating the coupling with FI. Two papers98,99 employed suspension-sampling FAAS for the determination of Mn, and Fe, Zn and Ca, respectively. Two further papers100,101 have reported the studies carried out on 10 and 11 different TCMs, respectively, using FAAS.
Atomic spectrometric techniques have been traditionally used to determine various trace metals in all types of drugs; however, in this review period the metal content has been used to determine the mechanisms of certain drugs. The simultaneous determination of Pt and iodine by ICP-MS has been used to study the mechanism of reaction of diiodoplatinum anti-cancer complexes.102 Samples were diluted in KOH with Te as an internal standard. Improved sensitivity (two-fold) was attained by the use of the S-option on the ICP-MS system. The results indicated that iodide release and different kinetics were observed for reactions of diiodo-PtIV and -PtII complexes in low molecular weight fractions of reaction mixtures of diiodo-Pt complexes and human albumin.
Further work investigated metallodrug–protein interactionsvia the use of size exclusion chromatography coupled with ICP-MS.103 Two Pt and three Ru based drugs were included in the study to assess their binding potentials with human serum proteins. Mixtures containing 1.6 µm of Pt based drug or 0.1 mM of Ru based drug and serum were incubated at 37°C. At regular intervals sub-samples were taken and diluted 100 fold with 30 mM Tris hydrochloride buffer of pH 7 and analysed. A Supelco Progel TSK column, with the previously mentioned buffer as the mobile phase, was used to monitor the protein bound drug fraction. The column effluent was fed directly into the ICP-MS instrument and the following isotopes were monitored: 194Pt, 195Pt, 196Pt and 100Ru, 101Ru, 102Ru. The method allowed the kinetics of drug binding to be successfully investigated.
Doped heavy metals as well as those from contamination have a significant effect on the sensitivity of film and photographic papers. The photosensitive layer of films and papers consists of microscopic grains of silver halide suspended in gelatine; to improve the photographic process the silver halide emulsions are usually doped with trace amounts of Pt group elements. Therefore, extremely sensitive techniques are required to determine the low level concentrations of these elements. Krytek and Heumann104 have used several techniques to aid the determination of Ir and Rh. Different sample treatment procedures were coupled with ICP-MS for the determination of Rh and Ir. For Ir, negative thermal ionization isotope dilution mass spectrometry (NTI-IDMS) and ICP-IDMS were explored, using enriched 191Ir spiked solution. For all analysis the removal of the silver matrix proved to be essential and for NTI-MS the gelatine matrix was also removed. Because of the similar chemical and physical nature of Ru and Rh, Ru was used as an internal standard during matrix separation for the determination of trace amounts of Rh. By determining the 103Rh∶99Ru ratio of the separated sample reliable results were attained. It was concluded that ICP-MS procedures could be used routinely and Ir and Rh in photographic emulsions were determined in the 0.5–650 ng g−1 range. NTI-IDMS can be applied as an accurate reference method.
Several papers report the analysis of cosmetics and shampoos for various trace metals.105–107 The most elaborate technique in use was the determination of mercury in cosmetics by flow injection cold vapour generation atomic spectrometry with on-line pre-concentration.108 The method involves the acid treatment of the sample in a focused-microwave digestor, on-line pre-concentration on a C18 column and cold vapour generation AFS determination. The method was applied to the target analyte at the pg level in different eye cosmetics (i.e., eye-liner, eye shadow, eye pencil), which have a maximum allowable level of 0.007% (m/m) Hg.
The latter method gives a clear illustration of the progress made in the coupling of various sample preparation and instrumental techniques together, which enable the trace metals to be determined in difficult matrices.
Solvent extraction for pre-concentration, however, has been dominant. The use of solvent extraction coupled to ETAAS appears to be the favoured combination. A detection limit of 3.3 ng l−1 was achieved for CrVIvia the use of this approach.109 The sample was aspirated (5.5.ml) into a FIA manifold (schematic diagram shown in the paper) and merged with a reagent stream of MIBK (0.22 ml min−1) and 0.065% ammonium pyrrolidine dithiocarbamate (0.5 ml min−1). The mixture was passed through a knotted PTFE extraction coil, after which the organic extract was separated by a gravity phase separator (diagram given) and delivered (0.1 ml min−1) into a collection tube. The collected organic extract (55 µl) was introduced via an airflow (0.25 ml min−1) into a graphite tube. The operation details are given. The calibration graph was linear from 5 to 600 ng l−1, and the methodology enabled a sample throughput of 24 samples h−1. The method was applied to waste water and synthetic sea-water.
Two other papers report the use of filtration in different guises after solvent pre-concentration. Taguchi et al.110 describe the determination of copper reactive pesticides in water after pre-concentration with a solvent soluble membrane. After acidification the samples (200 ml) were treated with 24% diammonium hydrogen citrate (4 ml), 10% ascorbic acid (4 ml) and 2 ml of 40 mM CuSO4 and left to stand for a few minutes. The pH was readjusted to 6 and the solution applied to a mixed cellulose ester membrane. The membrane was washed with 20 ml of water containing 1 drop of 0.1 mM methyl violet. The dyed area was cut out and dissolved in methoxyethanol. Detection limits in the order of 0.2–0.5 µg l−1 in the original sample were achieved: a novel way to pre-concentrate.
The second paper111 reports the use of solvent extraction coupled with cobalt(III) oxide collection for the determination of As in environmental and geological samples. The basics of the system were to acidify the sample solutions, mix with KI and extract with benzene, then stir the organic layer with 3 ml of IBMK and 30 ng of cobalt(II) oxide powder. The mixture was next vacuum filtered on a PTFE membrane filter. The contents of the filter were then slurried with water prior to analysis by ETAAS. A detection limit of 12 µg l−1 was achieved with a calibration linear to 500 µg l−1.
The other thrust in the solvent area is the continued use of membrane desolvation systems that reduce the solvent loading on plasma systems. Generally, the more volatile the solvent the better the removal capabilities of such systems. Gasoline can thus be aspirated directly into the plasma; under normal conditions this would extinguish the plasma immediately. The use of a microporous PTFE membrane desolvator has been evaluated for the on-line removal of organic solvents to facilitate the determination of trace contaminants in solvents by ICP-AES.112 Three organic solvents, isopropyl alcohol (IPA), methanol and DMSO, were studied. They report via monitoring the C I emission a reduction of the solvent by 82–89%. The net intensity of Fe, Al and Cu was increasing.
Akinbo et al.113 experimented with a flat sheet desolvator (FSMD). The system enabled the introduction of methanol without extinguishing the plasma. The bluish green colour of the plasma due to methanol was returned to its original colour by increasing the desolvator counter current gas to 8 ml min−1. The maximum emission intensities of analyte signals depended upon the counter current gas flow rate. For Cl the maximum intensity was observed at 5 ml min−1. The FSMD was also used as an interface between HPLC and the MIP-AES system. This was demonstrated using 2,6-dichlorobenzene and 5,7-dichlorohydroxyquinoline as analytes.
The next move forward was the use of a double membrane desolvator, (DMD).114 A system coupled to a MCN was built and tested, using isopropyl alcohol, on both an ICP-AES and an ICP-MS. The DMD enabled a reduction (50%) in the length of membrane normally used, with an increase in surface area. The influence of sweep gas was studied and reported.
Some interesting applications using electrothermal vaporization AAS have been reported this year. A wavelength modulation technique has been employed for the measurement of Li isotope ratios, using diode laser ETAAS.115 In this work, the relative absorption sensitivity for 6Li was reduced by tuning the laser to apply 2f-wavelength modulation (2f-WM) in the centre of the 6Li D1 fine structure component. Conversely, the absorption sensitivity of 6Li could be increased by tuning the modulation to the maximum in the red wing of the 2f-WM line profile. By calculating the 2f-WM line strengths, overlapping absorption lines could be deconvoluted, thereby facilitating the measurement of 7Li∶6Li isotope ratios as large as 2000. This work identified a potentially viable alternative for measuring Li isotope ratios, compared to ICP-MS or TIMS, both of which are adversely affected by large mass discrimination effects in the mass region of Li.
The use of hydrofluoric acid to decrease the background absorption signal caused by sodium chloride in ETAAS has been reported.116 Addition of a small volume of 40 to 80% HF solution to the samples lead to the formation of NaF during atomization, which has a much lower absorption signal than NaCl. The authors reported that addition of HF caused no damage to the pyrolytic material or the quartz windows of the atomizer.
A number of articles in which flame AAS was used as the detection system have appeared this year. A study of the release of Ni and Cr from stainless steel cooking utensils into food during cooking was performed.117 Ingestion of Ni and, to a lesser extent, CrVI can exacerbate dermatitis in people who are already Ni-sensitive. The authors measured Ni and Cr in 11 different meals cooked in different grades of stainless steel utensils as well as in glass saucepans. A significant difference in Ni and Cr content could be identified between glass and stainless steel utensils, but this difference was found to be low compared with the levels of Ni and Cr already contained in the food. Hence, there was reported to be no benefit in Ni sensitive patients switching to glass in place of stainless steel utensils.
A procedure for indirectly measuring Si in cobaltosic oxide by flame AAS has been described.118 A fairly complex sample preparation procedure was performed involving two reconstitution and evaporation stages, mixing with ammonium molybdate and finally an organic solvent extraction step. During the procedure, the trace Si was converted to a heteropolymolybdate Si species, which was collected in the organic extract. The Si was quantified indirectly by measuring Mo, facilitating a detection limit for Si of 0.2 µg ml−1.
On a more fundamental note, a review on the subject of acid interference effects on analyte signals in atomic spectrometry was published this year.119 Acid interference effects are complex and their magnitude is dependent on several variables, including nebulizer flow rate, spray chamber geometry and plasma parameters. In the review, the methods which have been developed to overcome such interferences are discussed and a strategy for minimizing these interferences is presented.
Applications of emission detectors feature strongly in the literature again this year. The use of a microwave plasma device, operated with either Ar or an Ar–air mixture, for the determination of tetraethyllead, via Pb emission at 283.31 and 405.78 nm, has been reported.120 This instrument can be operated with either air or argon as the carrier gas. Optimum detection limits (using the more sensitive 405.78 nm emission line) in the region of 0.02 ppb Pb were reported and the instrument response was linear over three orders of magnitude. This performance is not as good as can be achieved by ICP-MS, but the instrumentation has the advantage of being simpler and cheaper to purchase and operate. It is likely that the range of applications to which this type of instrumentation can be applied is large, but the small linearity range may be a limitation.
An on-line monitoring system for the determination of boron in dichlorosilane, using ICP-AES detection, has been described.121 On-line monitoring of the gas was achieved using a modified sampling and gas introduction system. Calibration of the system was performed with a mixture of diborane and argon, together with dichlorosilane, using a standard additions approach. This procedure offered a detection limit of around 0.6 µg B per gram of dichlorosilane, which was sufficient for the application. From a process analysis perspective, it will be interesting to see if future developments make this type of instrumentation viable for on-line analyses.
An elaborate system for the measurement of Ge in zinc electrolytic solutions, using ICP-AES, has been reported this year.122 Germanium is present in zinc ores and is a problem as it adversely affects the electrodeposition of zinc during the refinement of this metal. The system which was developed was comprised of an on-line separation stage, followed by hydride generation. Germanium in the samples was complexed with HCl (8.5 M), thereby forming GeCl4, which was extracted on-line into xylene. The two liquid phases were separated on-line and the xylene phase then mixed with NaBH4, dissolved in a dimethylformamide–acetic acid mixture, to generate GeH4. For the liquid–liquid separation stage, a gravitational separator and a membrane separator were evaluated. The generated GeH4 was separated via a gas–liquid separator and transported to the plasma by an argon stream. Detection limits for Ge of 1 and 3 ng ml−1 were achieved using the membrane and gravitational separators, respectively. Despite its apparent complexity, the procedure was sufficiently robust for the successful measurement of three real samples, the results of which compared well with those of a routine method.
The use of a parameter related internal standard (PRISM) method for measuring Ni in high salt and high acidity samples, using ICP-OES detection, has been evaluated as an alternative to conventional internal standardization.123 The PRISM approach is based on the assumption that the emission intensity of an analyte signal is affected only by changes in the excitation temperature of the plasma and changes in the sample introduction efficiency from sample to sample. The authors used La and Y to monitor the effect of changes in the excitation temperature and sample introduction efficiency, respectively, on Ni emission signals in aqueous samples. Changes in these parameters were simulated by varying the forward power and sample uptake rates, using a factorial analysis approach. The results of these experiments were used to derive correction factors for the Ni emission intensity for a range of high acid and high salt matrix samples. This approach was more efficient than conventional internal standardization for correcting emission intensities for higher matrix samples, but for lower salt content samples and lower analyte concentrations there was no discernible difference. The method seems to be beneficial for specific applications but, if the samples can be diluted, using the conventional internal standard approach will probably be faster and just as effective.
As has been the case in recent years, ICP-MS has featured heavily in the literature this year. Of particular interest is the discussion of applications of collision cell technology to reduce or eliminate certain interferences in ICP-MS. Collision cells operate by allowing ions produced in the plasma to collide with a gas such as He, N2 or ammonia in a pressurized cell located behind the sampling interface and before the mass analyser. These collisions result in molecular ion dissociation and exchange reactions as well as thermalization of the sampled ions. A particular benefit of collision cell technology is the near elimination of the 40Ar16O+ interference on 56Fe under normal plasma operating conditions, leading to the possibility of measuring Fe at ppt levels on quadrupole instruments. Interferences on other problematic elements such as K, Ca and Se are also significantly reduced. Völlkopf et al. have applied a collision cell ICP-MS system to the determination of K, Ca, Cr, Fe, As and Se at ppt levels in high purity hydrogen peroxide for the semi-conductor industry.124 The authors found that ammonia was the most effective collision gas for this application. All the elements of interest could be measured under one set of plasma conditions in a single run. Long term stability of the instrument (20 h) was not degraded compared with conventional operation. In a related work, the collision cell ICP-MS instrument was used for the determination of 41 elements in high purity HCl, HNO3 (diluted by a factor of 10 with water) and de-ionized water.125 Ammonia was used as the collision gas and was effective in eliminating 40Ar16O+, 40Ar1H+ and 40Ar2+, but not the 40Ar35Cl+ and 35Cl16O+ interferences, on 75As and 51V, respectively. In addition, the authors indicated that use of a low K quartz torch and Pt cones would probably improve the detection limits for K and Ni, respectively. Nonetheless, currently available collision cell instrumentation is already capable of meeting nearly all the semiconductor industry's specifications. Future developments in the technology will very likely be directed towards the use of other gases or mixtures of gases to eliminate a wider range of interferences. It is less likely at present that collision cells will prove to be useful in environmental analyses because of recombination reactions such as the formation of CaO in place of ArO in the cell, when matrix containing samples are analysed.
Coupling chromatography systems with ICP-MS continues to be of interest. Göessler et al. have used ion chromatography coupled to ICP-MS to separate and quantify inorganic sulfur species, via the 32S16O+ ion signal with the instrument tuned to give high oxide ratios (under cool and hot plasma conditions).126 A mobile phase based on NaOH was found to be necessary to generate the required SO+ signal. A membrane suppressor system was used to remove Na prior to the ICP-MS measurement. Detection limits were in the range from 35 to 270 ppb depending on the sulfur species. Signals derived from sulfide were found to be suppressed for saline samples, under cool plasma conditions, and such samples also reduced the retention time for sulfide. The cause of both effects was identified as chloride ions. This method illustrates an interesting way of indirectly measuring an element which cannot be measured directly with the instrument configuration used. However, when an oxide signal is used in place of the parent ion, care must be taken that the oxide formation rate is kept constant for both standards and matrix containing samples throughout the analysis.
New applications of laser ablation ICP-MS continue to appear. An Nd:YAG laser coupled to a quadrupole ICP-MS was used to measure the major matrix and trace elements in Cd–Zn–Te crystals.127 The spatial distribution of a range of elements across the surface of the material was studied and depth profile experiments were also performed. Using the relative ion signals for the trace and matrix elements, the authors estimated the relative mass fraction of each element in the area of the sample analysed. An excimer laser coupled to a quadrupole ICP-MS was used for the determination of minor and trace elements in eight USGS geological reference materials.128 The samples were fused into pellets with Li2B4O7 as an alternative to a time consuming digestion procedure. Elements of interest were quantified by external calibration with the NIST 612 glass reference material and SiO2 was used as the internal standard. A total of 40 isotopes were quantified down to concentrations as low as 240 ppb in the solid. The results compared favourably with both XRF and solution nebulization ICP-MS and illustrated how LA-ICP-MS could be used as a complementary technique in the analysis of geological materials. Improvements in laser systems and instrument design are now making it possible to quantify elements at ppb levels in the solid. This, coupled with an increasing range of available solid calibration materials, is starting to make LA-ICP-MS a viable alternative to either solution ICP-MS or the less sensitive XRF technique for a range of analyses.
The development of multi-collector ICP-MS (MC-ICP-MS) over the past few years is leading to an increasing number of articles discussing applications of high precision isotope ratio measurements for geochronological, nuclear and high accuracy isotopic composition analysis and atomic weight determination purposes. A particular MC-ICP-MS article of interest this year was a study of the variations in Ca isotopic composition in carbonate materials.129 Under conventional plasma conditions, the very large 40Ar+ signal prevents measurement of 40Ca and, in addition, interference from Ti on masses 46 and 48 prevents satisfactory data being obtained for these low abundance Ca isotopes. So, for conventional ICP-MS operation, Ca isotope ratio measurements are practically limited to masses 42, 43 and 44. The authors measured the 44Ca∶42Ca ratio in an aqueous calcium solution and in digested marine and terrestrial carbonate samples and compared the results with measurements of the NIST 915a Calcium Carbonate reference material. Variations in the 44Ca∶42Ca ratio were expressed as δ44Ca units (deviations in parts per 1000 from the same ratio in NIST 915a). Deviations of up to 0.7‰ were observed, which agreed well with the results of previous studies using TIMS. The source of isotopic deviations of Ca between terrestrial and marine carbonates is believed to be due to fractionation of the latter during carbonate precipitation from the surrounding sea-water environment. Using cool plasma methodology in combination with multi-collector ICP-MS offers the potential for making high precision Ca isotope ratio measurements using 40Ca as well, although ionization suppression from the matrix of digested samples could limit the range of application to simple matrix samples.
Other mass spectrometric methods of detection have been applied in a range of applications this year. Boyle et al. used secondary ion mass spectrometry (SIMS), together with X-ray photoelectron spectroscopy, to study impurities in CdS/CdTe photovoltaic cells.130 SIMS was used to quantitatively determine 12C, 16O, 34S and 35Cl in the thin film voltaic cells. The distribution profiles of these isotopes throughout the cell materials was also studied. The authors found that the cells were tolerant to high concentrations of these impurities, indicating that the low-cost, wet chemical method of manufacture of the devices is potentially viable. Fast atom bombardment mass spectrometry (FAB-MS) was used for the identification and characterization of silicate complexes in calcium chloride solutions.131 Dissolution of silica in aqueous CaCl2 resulted in the formation of several silicate complexes, ranging from the simple monomeric Si(OH)3O− to large cyclic and linear species such as Si4(OH)3O10Ca3−. This latter linear complex type was not found in solutions of NaCl in which silica had been dissolved. This was suggested to be due to the fact that Ca2+ could assist in forming intra-molecular bonds between neighbouring Si-O− groups, whereas Na+ could not. An electrospray ionization mass spectrometry method was developed for the quantitation of perchlorate in drinking water samples.132 Determination of perchlorate has become an issue since this species was discovered in drinking water supplies in the US. Usually, an ion-chromatography method is used for the measurement, but chromatographic retention times alone are not considered to be unique identifiers in a court of law, and so an additional confirmatory method such as mass spectrometry must be applied. However, mass spectrometric methods lack the required sensitivity, and, without prior chromatographic separation, quadrupole mass spectrometers lack the resolution required to conclusively identify the perchlorate ion in the presence of other species. The aim of the work was to use tetraalkylammonium cations and sterically hindered, nucleophilic organic bases to improve selectivity using the electrospray ionization mass spectrometer without losing sensitivity. Selectivity was achieved by the formation of a stable association complex between the base molecule and the perchlorate anion. Using chlorhexidine in methanolic solution as the base molecule, the authors were able to selectively determine perchlorate (via the complex species at mass 605) down to a lower limit of detection of 10 ppb, which compared favourably with the 5 ppb limit achievable with ion chromatographic methods.
A study was performed into the isotopic composition and atomic weight of zirconium, using surface ionization mass spectrometry.133 In this work, enriched 90Zr and 94Zr mixtures were used to calibrate the instrument. The calculated atomic weight of Zr was found to agree closely with the currently accepted IUPAC value and may in future be useful in refining the current value. Thermal ionization mass spectrometry (TIMS) was used for measurement of the absolute isotopic composition and atomic weight of germanium.134 The TIMS instrument was calibrated using gravimetrically prepared synthetic mixtures of enriched germanium isotopes (as high purity oxides). The isotopic composition of Ge in several terrestrial materials was subsequently measured and no isotopic fractionation was found. An alternative procedure for measuring the isotopic composition and atomic weight of Ge was reported by Kippardt et al.135 The approach adopted here was to convert Ge to GeF4, using both direct fluorination and a wet chemical procedure, before measuring the gaseous product by isotope ratio mass spectrometry. A full assessment of the sources of uncertainty in the measurements, ranging from weighing uncertainties to uncertainty in the mass discrimination measurement, was performed following the recommendations of the ISO/BIPM Guide to the Expression of Uncertainty in Measurement. The results of the measurements were compared to previously published values in the framework of improving the quality of the existing data.
Materials control and accountancy (MCA) is of utmost importance in the nuclear industry. Analysis, undertaken for the purposes of MCA, provides a `Gold Standard' for any laboratory in terms of accuracy, precision and reliability. This crucial area has seen some development in the period covered by this review and includes: the certification of a new generation of uranium isotopic CRMs,139multi-collector ICP-MS,140X-ray fluorescence spectrometry,141 improved TIMS methodology142 and rapid sample preparation for TIMS.143 A number of instrumental and procedural developments were reported in the quest to achieve lower uncertainties in the next generation of uranium isotopic CRMs.139 These improvements included: the combined use of very high precision gas mass spectrometry (UF6) and TIMS, upgrade of TIMS detectors, a multi-stage TIMS instrument capable of very high abundance sensitivity measurements, production of new primary isotopic calibration mixtures, more precise instrumental calibration procedures and a new Class 100 clean room for sample preparation. The introduction of multi-collector ICP-MS (MC-ICP-MS) instruments is certain to have a dramatic effect upon MCA measurements. High accuracy and precision measurements of uranium isotope composition were obtained using MC-ICP-MS.140 To correct for drift in the mass bias correction, Pb or Th was added to the sample solutions as internal references. This approach allowed the determination of major and minor isotopes with a precision comparable to benchmark standards. Importantly, this precision could be achieved with a sample throughput that exceeded TIMS by a factor of 4–5. Radioisotope excited, transmission corrected, K-line XRF was used for U and Pu assay.141 A mixed 57Co + 153Gd transmission source was used to correct for variations in absorption. This correction allows a single point calibration to cover the entire desired concentration range up to 300 g dm−3. The chemical equilibration of the spike and sample, and an internal evaluation of the mass bias of the instrumentation, were integral parts of an assessment of TIMS for Pu accountancy.142 It was concluded that the accuracy and precision achievable with TIMS was comparable to that of a reference method (coulometry) and could be used as one of the analytical tools for future certification exercises and inter-laboratory studies. The application of extraction chromatography for the determination of U and Pu in spent fuel solutions by TIMS for MCA was reported.143 After chemical equilibration of the tracer and sample, the sample was loaded onto a single UTEVA cartridge, washed to remove fission products and a sequential elution program used to produce pure Pu and U fractions suitable for TIMS. This rapid separation reduced overall sample preparation times by a factor of three.
Isotopic analysis to the highest standard of precision and accuracy, as offered by TIMS or MC-ICP-MS, may not always be required and more moderate performance may be `fit for purpose'. Quadrupole ICP-MS provided a rapid and cost-effective means of determining uranium enrichment in a processing monitoring role.144 Optimal precision and accuracy was achieved by the use of a single standard. This standard was matched closely to the sample in terms of enrichment and concentration. Similarly, the isotopic homogeneity of a uranium metal billet, produced from two feedstocks of differing enrichment, was assessed by Q-ICP-MS.145 The analysis was sufficiently precise to allow an assessment of the homogeneity of the billet where the aggregate variation in the isotopic composition was about 1%.
Laser induced breakdown spectroscopy was applied to the determination of impurities in uranium and plutonium oxides146 using an optical fiber system for transmission of the laser pulse and collection of the resultant emission. The instrumentation was based around a frequency doubled, Q-switched, Nd-YAG laser and a 1 m spectrograph fitted with a CCD. The plasma was generated in air. Eighteen impurities in UO2 were detected at the 500 µg g−1 level and 12 impurities in PuO2 at the 100 µg g−1 level. Isotopic analysis of U in UO2 was performed by the combination of a laser induced plasma and diode laser induced atomic fluorescence spectrometry.147 Resonant atomic fluorescence spectra were obtained by rapid scanning of the diode immediately after each laser sampling event. Alternatively, time integrated measurements, with the excitation laser fixed at a specific isotope wavelength, were also obtained. Precision and accuracy, for natural abundance U, of 5 and 7%, respectively, were achieved. In both cases of the application of laser induced plasmas, further refinement of the techniques are probably required in terms of limits of detection146 and accuracy/precision.147
Micro-analytical techniques, such as secondary ion mass spectrometry and electron probe microanalysis, were applied to the examination of radioactive particles148 and melted fuel assemblies.149 Particles in soils, swipes and forensic samples were examined by SIMS and identified qualitatively in terms of U and Pu content and the isotopic composition. The latter could be determined with a typical accuracy and precision of 0.5%. Statistically meaningful results could be obtained from a specimen containing 104 atoms m−3 of U contained in particles weighing a few picograms. Fuel assemblies were irradiated to a burn-up of 23 GWd/tU, heated by fission power to about 2500°C and the resultant melted bundles sectioned for optical microscopy and EPMA.149 At mid-height, the bundle had collapsed forming a void and the molten material had pooled. This solidified melt was a solid solution resulting from the fusion of the UO2 fuel and Zircalloy cladding. Oxide and metallic inclusions were observed. The former were rich in Fe and Cr, the latter in Ni and contained small amounts of Mo, Tc, Ru and Pd.
Resonance ionization and accelerator mass spectrometry continue to attract interest. The determination of 129I in air by means of accelerator mass spectrometry offers detection limits of 104 atoms m−3.150 Air (350 m3) was sampled by means of charcoal filters. A tracer was added to the charcoal, before extraction of the analyte of interest, by slurrying the charcoal with water, adding a stable tracer (NaI), treatment with nitric acid and sodium nitrite to convert the tracer to iodine and equilibration of the tracer with the solid. The charcoal was separated, washed and dried, and the uptake of the tracer estimated by determination of iodine, as iodide, in the supernatant (ion chromatography). The iodine was extracted from the charcoal by extraction with aqueous NaOH–NaHSO3 and purified by multiple extractions and back extractions of elemental iodine into chloroform. Finally, the source was prepared by precipitation as AgI and introduced to the AMS. The status of AMS was reviewed briefly with an emphasis on recent technology developments.151 Similarly, the status of the TANDAR AMS facility (Argentina) was reviewed, recent improvements and the focus of the current work on chlorine-36 and Ni beams described.152 The measurement of Pu and U using resonance ionization time-of-flight mass spectrometry, and some of the limitations of the technique, was discussed.151 Samples were vaporized from a Rh filament and ionized by a single color, 3-photon process. Two major limitations were identified. Firstly, the chemistry of the ion source yielded both atomic and molecular species that led to a variety of interferences. Secondly, the low detection efficiency of the instrumentation due to a combination of a continuous atom beam and a pulsed ionization laser was noted. The relative merits of different filament preparation schemes were discussed and a technique proposed that addressed the low duty cycle of the instrumentation. Ionization schemes for the measurement of strontium-90 by diode based RIMS were described.153 Double and triple resonance schemes, used in combination with single mode diode lasers, provided a high degree of optical selectivity. It was suggested that a triple resonance scheme, in conjunction with mass spectrometry, should produce a system with sufficient abundance sensitivity to allow measurement of 90Sr at background environmental levels. The application of multi-step, cw-RIMS to odd isotopes is usually hampered by hyperfine structure.154 The splitting of the transition strength leads to lower ionization efficiencies and large uncertainties in isotopic ratio measurements. The use of appropriate optical pumping schemes overcomes these problems and allows the determination of 41Ca abundances using a double-resonance, three photon ionization in a collimated atomic beam combined with quadrupole mass spectrometry.
Element | Matrix | Technique;atomization;presentationa | Sample treatment/comments | Ref. |
---|---|---|---|---|
a Hy indicates hydride and S, L, G and Sl signify solid, liquid, gaseous or slurry sample introduction, respectively. Other abbreviations are listed elsewhere. | ||||
PETROLEUM AND PETROLEUM PRODUCTS— | ||||
C and H | Petroleum liquids | AE;GC;L | The samples are injected directly onto the column; instrument optimization is discussed | 155 |
Fe, Mo and Sn | Petroleum crude | AA;F;L | After treatment with acids Fe, Mo and Sn are separated from the matrix elements by simultaneous solvent extraction of 5,5′-methylenedisalicylhydroxamic acid complexes from HCl–NaClO4 solution into a Bu Me ketone–tributyl phosphate solution | 156 |
Mn | Gasoline | AE;GC;L | The DB-5-MS column was operated with a 7 psi column head pressure with He carrier gas. O2 and H2 reagent gases were at 25 psi and 70 psi, respectively. The samples were injected directly | 67 |
Pb | Gasoline | MS;ICP-ETV;L | Isotope ratio was used to determine the source of Pb in blood. It was found to be ceramic cookware rather than leaded fuel | 64 |
Pb | Gasoline | AA;GF;L | 5 ml of gasoline and 5 ml HNO3 were added together and irradiated with microwave | 69 |
Pb | Atmospheric particles, leaded petrol, pine needles | MS;ICP;L | The 206Pb∶207Pb ratio was determined by direct analysis in rainwater collected in Scotland to study the effects of leaded fuel | 66 |
S | Coal/petroleum | MS;—;L | Isotope information gathered by MS is investigated, sample preparation is discussed | 157 |
Various (6) | Naphtha | MS;ICP;L | Samples were emulsified with Triton X-100 prior to analysis by ICP-MS. Detection limits of 0.09 and 0.12 µg l−1 for Pb and Hg were achieved | 158 |
Various | Petroleum fractions | MS;ICP;L | Samples were analysed after four different sample routes including: the form of a burnt gas; micro-emulsions; use of surface active agent; and with the aid of oxidizing agents. Advantages and disadvantages of the methods are described | 71 |
OILS, FUELS AND CRUDE OIL FRACTIONS— | ||||
As | Natural gasoline, gas condensate | AA;GF;L | An in situ absorption technique was developed to directly determine arsenic. Pd alloy was used as the solid absorbent | 159 |
Cd, Cu, Pb | Coal | AA;GF;L | The sample is introduced into the ICP as a slurry, which consisted of5–30 mg of coal (ground) in 1.5 ml of a diluent mixture of 5% (v/v) HNO3, 0.05% Triton-100 and 10% ethanol | 77 |
Cr, Cu, Fe, Pb | Lubricating oils | AA;F;L | Sample preparation was via a closed vessel microwave digestion system, using HNO3 and H2O2 | 85 |
Cr | Lubricating oils | AA;ETV;L | A 1 ml plug of sample was injected into a carrier stream of hexane, mixed with streams of 3.8% (m/v) NaCl, 5% Na dodecylsulfate and 5% of sec-butanol. An emulsified sub-sample was collected and introduced to the spectrometer | 82 |
Ge | Coal | XRF;—;S | The Compton peak was used for matrix matching when developing the XRF method. The samples were presented as fused glass beads | 72 |
Hg | Natural gas liquid and condensate | AA;ET;L | Activated carbon is used to extract and pre-concentrate the Hg present in the samples. The sample was presented to the instrument as a slurry | 81 |
Mn | Fuel additive | AA;DL-HPLC;L | Speciation of methylcyclopentadienylmanganese tricarbonyl (MMT) was achieved by a combination of HPLC with diode laser atomic absorption spectrometry (DLAA) | 68 |
Pb | Crude oil | AA;ET;L | Organic palladium and palladium–magnesium chemical modifiers are used to unify the behavior of lead present in different forms in distillation fractions. | 89 |
Pb | Fuel | MS;ICP;L | A low pressure ICP-MS system was used to determine tetraethyllead in fuel. A review of systems coupled together is discussed, e.g., GC-MIP-AE | 65 |
S | Diesel oil | XRF;—;L | The evaluation of EDXRF for the determination of S in diesel fuels and the influence of C/H is reported | 160 |
S | Coal | AE;ICP;L | The samples were digested with microwaves prior to analysis; information re the different forms of sulfur present was investigated | 79 |
Sn | Lubricating oils | AA;HG;L | 1 g of sample was microwave digested via 4 stages | 83 |
V | Oil | AA;F;L | Three sample pre-treatment routes were evaluated: ashing; microwave digestion; and burning in a bomb calorimeter | 161 |
Various (5) | Waste oils | AE;ICP;L | The analysis of non-metals, e.g., Cl, I, P, using the 130–190 nm region, is reported | 91 |
Various (9) | Aviation engine oils | AE;ICP;L | The sample is treated with activated carbon, to reduce the risk of trace element loss, and then digested with a dry digestion method. The digested ash is dissolved in HCl–HNO3 (2∶1) prior to analysis | 162 |
Various | Aviation engine and hydraulic oils | MS;ICP;L | Samples were heated overnight in a Parr bomb reaction vessel with 1 ml HNO3; the resulting solutions were analysed directly. Other samples were dissolved in a 60% propanol–0.1% Triton mixture and analysed using a USN membrane desolvator attached to ICP-MS | 86 |
Various | Waste oils | AA;ET;L | Particulate materials were collected via cartridges which separated them in to size. Each fraction was digested using HCl–HNO3 | 163 |
Various | Lubricating oils | MS;ICP-ETV;L | The sample was injected into the graphite furnace without sample dilution or it was introduced after freezing and placing into a modified graphite furnace | 164 |
Various (6) | Coals, lignites and fly ash | AE,ICP;L | The samples were burnt, muffled at 775![]() | 165 |
Various | Coal | AE;GD;L | The sample preparation was limited to ashing and pressing the resultant sample into discs. Using glass beads was also evaluated | 76 |
Various (9) | Coal | MS;ICP;L | Microwave digestion using concentrated nitric acid was a suitable preparation route for the analysis | 78 |
Various | Oil | XRF;—;L | The use of EDXRF for the analysis of used oils is demonstrated and compared and contrasted to other techniques. No sample pre-treatment was necessary | 88 |
Various | Oil | XRF;—;L | The use of a compact X-ray fluorescence spectrometer to monitor real-time wear metal analysis is described. Sample is introduced form the flowing stream with no pre-treatment | 166 |
Various | Waste engine oil | XRF;—;S | The sample is presented to the XRF as a homogeneous pellet on a mylar backed aluminium ring | 167 |
Various (17) | Fuel oils | MS;ICP;L | Microwave digester was used to extract trace elements from a relatively small samples mass, 20 mg prior to analysis | 168 |
Various (16) | Wood, coal | MS;ICP;L | 1-Methyl-2-pyrrolidinone (NMP) was used to extract samples of wood and coal. These underwent further sample preparation before analysis. | 169 |
Various (24) | Fuel oil | MS;ICP;L | The sample was digested using pressurized closed vessel microwave digestion. The influence of sample size, reagent composition and duration of heating was studied | 170 |
SOLVENTS— | ||||
As | Organic solvents, wines | MS;FI-ICP;L | The samples was introduced directly into the FIA system using a MCN nebulizer into the plasma | 102 |
As | Geological samples | AE;ICP;L | The samples (solids) were digested with a mixture of HNO3, HClO4 and HF (5∶5∶3) solution on a hotplate at 230![]() | 111 |
As, Se | Drinking water | XRF;-;L | The samples were extracted with solvent, which were then analysed for trace metal contamination | 171 |
Cr | Waste water, synthetic sea-water | AA;ET;L | The use of solvent extraction as a preconcentration technique is demonstrated to achieve detection limits down to 3.3 mg l−1 | 109 |
Ca, Cr, Mn | Isopropyl alcohol (IPA) | AE;ICP;L | The use of a double membrane desolvator for the analysis of IPA is described | 114 |
Ru | Bittern | AA;GF;L | The sample was extracted with 10 mM bromothymol blue and 50 mM 18-crown-6. The organic phase was analysed via GFAA | 172 |
Various | Organic solutions | AE;MIP;L | A flat sheet membrane desolvator was coupled to a MIP-AE for the analysis. No sample preparation was required | 113 |
Various | DMSO, IPA, methanol | AE;ICP;L | A microporous PTFE membrane desolvator was built and evaluated for the on-line removal of organic solvents, the samples were introduced directly | 112 |
ORGANIC CHEMICALS— | ||||
Al, Ca, Fe, Mg | Steel solid solution | AE;ICP;L | The method includes low temperature electrolysis using an electrolyte solution containing tetramethylammonium chloride, triethanolamine, glycerol, methanol or the latter substituted by ethylglycerol, absolute ethanol. The elements are separated by liquid chromatography | 173 |
AsV, AsIII, organoarsenic species | Aqueous solutions | XRF;—;S | AsV, AsIII, dimethylarsinic acid (DMAA) and phenylarsonic acid (PAS)were separated via the use of various activated carbon species. PAS was separated using V-loaded activated C, AsV was collected on La-loaded C. AsIII was separated via co-precipitation with ammonium pyrrolidine dithiocarbamate, the precipitate being adsorbed on activated C, and finally DMAA was collected on Zr-loaded C. The various solid fractions were then analysed by ED-XRF | 174 |
Arsenic (sp) and various | Chinese medicine | MS;ICP;L | The samples were digested via conventional wet oxidation and microwave digestion using combination of HNO3–HClO4, HNO3–H2O2 | 96 |
Au | Rock | AE;ICP;L | The rock extract was mixed with HCl to a final concentration of 0.1 M, 1 mM malachite green and 1–3 ml of 20% naphthalene solution in acetone and the mixture filtered | 175 |
13C | Organic and inorganic materials | RM-MS;GC;G | Carbonaceous compounds volatized by a laser are quantitatively converted to CO2 gas by a combustion furnace. The gases are swept by He carrier gas into a GC prior to their determination by isotope monitoring mass spectrometer (IM-MS) | 176 |
Cd, Pb | Chinese crude drugs | AA;GF;L | The samples were digested with 3 ml of concentrated HNO3 | 177 |
Co | Cobalt mesoporphyrin drug | AA;F;L | The drug (100 µl) was mixed with 1% Triton-100 (10 µl) prior to analysis | 178 |
Cr | Gelatine | AA;ETV;L | No sample pre-treatment was necessary | 179 |
Cr, K, Na, Ni | Chinese medicine | AA;F;L | 11 Chinese medicines were analysed following digestion with HNO3 and HClO4 Their possible effects in the therapy of coronary heart diseases are discussed | 101 |
Cr, Mo, Yb | Biological, environmental, high purity samples | AE;ETV-ICP;L AA;ET;L | The influence of PTFE slurry on the refractory elements in ETV-ICP-AE and ETAA is reported | 93 |
Cu, Mn | Single cell protein, cod muscle, freeze dried animal blood | AA;ETV;L | A PTFE knotted reactor was pre-coated with 1-phenyl-3-methyl-4-benzoylpyrazol-5-one and a portion of test solution was injected into the carrier stream and passed through the reactor for 15 s. The absorbed chelates were eluted with methanol which was injected directly into a pyrolytically coated graphite tube for analysis. The detection limits for Cu and Mn were 5.7 and 5 ng l−1, respectively | 180 |
Hg | Cosmetics | AFS;CV;L | The method involves the acid treatment of the sample in a focused-microwave digester and on-line pre-concentration on a C18 column | 108 |
I | Glacial acetic acid | MS;ICP;L | In was added to the diluted samples as an internal standard and the reduction in memory effect of I in glacial acetic was studied | 181 |
I, Pt | Diiodoplatinum anti-cancer complexes | MS;ICP;L | Samples were diluted in KOH containing Te as the internal standard | 102 |
Ir, Rh | Photographic emulsions | MS;ICP;L MS;NTI-ID;L | The silver matrix was removed by dissolving the emulsions in concentrated ammonia solution; for the NTI-IDMS the gelatine matrix of the emulsion was also removed | 104 |
Mn | Chinese medicinal herbs | AA;F;L | The samples were dried and pulverized to 160 mesh, and 1.5 g was suspended in aqueous 1.5 g l−1 agar. The suspension was stable for 10 min | 98 |
Ni, Pb, V | Xylem sap | TXRF;—;L | The effect on organic acid transport in xylem sap of Pb, Ni and V was studied. After the introduction of various nutrients into the soil, three organic acids of the Krebs cycle were measured by RP-HPLC.Simultaneously, the heavy metal content was determined by TXRF | 182 |
Organ-otin,organo-arseniccom-pounds | Aqueoussolutions | MS;HPLC-ICP;L | Solid phase extraction (SPME) was used to extract ionic organotin and organoarsenic without derivatization | 183 |
Pt, Ru | Drugs | MS;GPC-ICP;L | The Pt based or Ru based drugs were incubated at 37![]() | 103 |
Se | Shampoo | AFS,GF;L | The samples are digested using microwave digestion and reduced via continual derivatization (hydride formation) | 107 |
Si | Organic matrices | GEXRF;—;S | The sample was deposited on an optically flat carrier on an air cooled rotating table ready for irradiation | 184 |
Tl | Waste water, fresh water | AA;ET;L | The sample was passed through a tribuyl phosphate extraction resin after treatment with FeCl3 (30 mg l−1) and H2O2 and heating. The Tl was eluted off the column with 0.25% ammonium sulfite and 0.5% ascorbic acid solution. Limit of detection obtained was 3 µg l−1 | 185 |
Trimethyl Se | Urine | MS;MIP;L | The urine was passed through various columns prior to analysis by MIP-MS. They included ODS GH-C18, pre-concentration column, anion exchange resin IC-C75 and, finally, an Asahipac GS-220HQ column | 186 |
Various (32) | High purity metal-organic compounds | AE;ICP;L | Trimethylgallium was decomposed slowly at low temperature without the addition of oxide prior to analysis. 32 elements were determined with detection limits of ∼1 µg g−1 being reported | 187 |
Various (73) | Heroin | MS;ICP;L | ICP-MS data (73) elements was used to fingerprint heroin from various sources | 188 |
Various (6) | Drugs | AA;GF;L TRXRF;—;L | An iminodiacetic acid cellulose (IDAEC) column was used for pre-concentration of the elements. CrIII and CrVI were separated using IDEAC and anion exchanger diethylaminoethyl (DE) cellulose | 189 |
Various | Chinese medicines | AA;F;L | The samples were carbonized and ashed at 550![]() | 100 |
Various | Tobacco | AA;F;L | Tobacco was soaked overnight with HNO3 in a sealed phial. The extract was evaporated with HClO4 to near dryness and the residue was diluted to 25 ml | 190 |
Various | Fluorcarbon resin | AA;GF;L | The sample was placed on a Si wafer and decomposed at 550![]() | 191 |
Various | Amberlite XAD-7 functionalized with chromotropic acid | AA;F;L | The paper details the study to assess the performance of the new polymer matrix | 192 |
W | Drugs | AE;ICP;L MS,ICP;L | A portion of sample was mixed with concentrated nitric acid–H2O (4∶1) | 193 |
INORGANIC CHEMICALS AND ACIDS— | ||||
Ag | Road salt | AA;ETA;L | Samples dissolved in water. Ag preconcentrated and separated from the salt matrix by co-precipitation with Co pyrrolidinedithiocarbamate, followed by re-solubilization | 194 |
Al | Molten germanium | MS;—;S | Germanium samples ablated with Nd∶YAG laser to melt the surface. Al diffusion through the molten Ge studied using SIMS | 195 |
Al | Aluminium chlorohydrate (ACH) | AMS;—;L | ACH labelled with 26Al prepared to study Al absorption from the use of anti-perspirants. Sample fractionated using gel filtration and fractions produced were studied for 26Al using accelerator mass spectrometry | 196 |
Al | Ferrosilicon | AE;ICP;L | Sample digested with a HNO3–HF–HClO4 mixture, evaporated to near dryness, then reconstituted with HCl and H2O | 197 |
As | Hydrogen peroxide, ammonia and water solutions | MS;ICP;L | Samples acidified with HNO3 | 198 |
As | Antimony trioxide | HG-AFS;—;G | Sample converted on-line to SbH3. Arsenic simultaneously converted to AsH3. Hydrides passed through KMnO4 solution where SbH3 decomposes much faster than AsH3. Remaining AsH3 passed into detector | 199 |
As and Sb | Aqueous solutions | AA;ETA;L | Samples were mixed with Pd and Ir chemical modifiers and a study made of the effect of these modifiers on the measurement of As and Sb | 200 |
As and V | Airborne matter reference materials | MS;ICP;L | Samples digested using a HNO3–H2O2–HF mixture in a high pressure digestion system | 201 |
Au and Si | Gold silicide | MS;—;S | Gold coated onto Si substrate and annealed under vacuum at 363![]() | 202 |
C, Ge, S and Si | Toxic hydride gases | AE;—;G | No sample pre-treatment required. Gas chromatographic procedure used to separate gaseous impurities of interest prior to detection | 203 |
Ce | K3Li2Nb5O15 crystals | MS;ICP;L | Samples digested on a hotplate with a H2SO4–H2O2 mixture, then diluted with water | 204 |
Cl | Aqueous solutions | MS;—;G | Cl2 and inorganic chloramines separated from aqueous solution using a flow-through membrane introduction system coupled to the mass spectrometer | 205 |
Cl, Br and I | Aqueous solutions | AE;ICP;L | No sample pre-treatment required | 206 |
Co | XF-210 phosphono-acid dirt preventing agent | AA;F;L | Sample digested with HNO3–HClO4 mixture. Digest evaporated to 0.5 ml, then diluted to 10 ml with water | 207 |
Co, Fe and Ni | Lithium carbonate and potassium carbonate melts | AA;F;L | Sample (1 g) dissolved in concentrated HNO3 (3 ml), next diluted to 50 ml and then aspirated directly into the instrument | 208 |
Cr | Tannery waste water | AE;ICP;L | Samples were first digested with a HNO3–H2SO4 mixture, then KMnO4 was added to convert CrIII to CrVI in the form of Cr2O72− | 209 |
Cu, Fe and Ni | Caustic soda | AA;F;L | Sample dissolved in concentrated HCl (to pH 2). Ethanolic phenolphthalein (1 drop) added, solution diluted and purified NaOH added (slight excess). Samples introduced to instrument via a flow injection manifold | 210 |
Er and Nd | Bismuth tellurite optical crystals | AA;ETA;L | Samples were dissolved with concentrated HCl and mixed with a chemical modifier (triammonium citrate) | 211 |
Fe and Ni | Carbon monoxide | FT-IR;—;G | None performed. Fe and Ni carbonyls were measured and quantified against in-house prepared standards | 212 |
I | Glacial acetic acid | MS;ICP;L | Samples were diluted 1∶3 with water, spiked with In as internal standard and introduced to the instrument using flow injection, with an ammonium hydroxide carrier stream. Standard additions used for quantification | 181 |
I | Calcium iodide tablets | AE;ICP;L | Sample dissolved in water. HClO4 and H2O2 were added (precipitate formed) and the supernatant solution aspirated directly into the instrument | 213 |
N | Propionibacteria | MS;—;G | Propionibacteria was cultivated in a yeast extract lactate medium for 3 d. A portion of the culture was mixed with 14N and 15N-enriched KNO3. The 15NO and 14NO formed were passed directly into the mass spectrometer | 214 |
Ni | Alkali metal salts | AE;ICP;L | Aqueous solutions of the salts were mixed with hexamine buffer and passed through an Amberlite XAD-2 column loaded with 1-(2-pyridazol)-2-naphthol. The Ni chelate so formed was eluted with HCl and analysed off-line | 215 |
Pb | Table salt | AA;F;L | Samples dissolved in dilute HNO3. Solution buffered to pH 4, then APDC added. The chelates formed were extracted into IBMK, then back extracted into dilute HCl | 216 |
S | Gaseous samples | XRF;—;G | Samples passed through a flow cell containing a Be film. The film was irradiated with the X-ray source and S emission from the sample flowing through it detected | 217 |
Se | Table salt | AA;F;G | Samples dissolved in water, then mixed with HCl and NaBH4 to produce H2Se, which was then passed into the detector | 218 |
Si | Silicon dioxide on silicon substrate | XPS;—;S | No sample pre-treatment required | 219 |
Si | Airborne matter | MS;LA-ICP;S | Airborne particulates collected on PTFE membrane filters before direct laser ablation of the filters. Data compared with XRF results | 220 |
Tl | Caesium iodide crystals | AA;ETA;L | Powdered samples were dissolved in water. Aqueous NH3 was added, then the sample was mixed with Pd and Mg(NO3)2 chemical modifiers | 221 |
Zn | Zinc phosphate coated steel | XRF;—;S | No sample pre-treatment performed | 222 |
Zn | Nickel electrolyte | AA;F;L | Sample introduced via a flow injection manifold onto a strongly basic anion exchange column. Column eluted with water and eluate directed into the instrument | 223 |
Various | Chemical grade potassium salts | AA;F;L | Sample dissolved in dilute HNO3, then buffered and mixed with a chelating agent. Metal chelates collected on Amberlite XAD4 resin, then eluted, evaporated to dryness and finally reconstituted with 1 M HNO3 | 224 |
Various | Sodium chloride | AA;F;L | Samples dissolved and diluted in water. Main focus of work was on optimizing the FAA to minimize interference effects | 225 |
Various | Hydrogen peroxide | MS;ICP;L | Samples analysed directly using a standard additions approach | 226 |
Various | Barium titanate powders | MS;ICP;L | Samples digested with HCl and a few drops of H2O2, then diluted with water | 227 |
Various | Sediments | MS;ICP;L | Dried samples digested first with XeF2 at high temperature and pressure, then using aqua regia. Digests then diluted with water | 228 |
Various | Inorganic salts | MS;ICP;L | Samples were injected into a carrier stream of ammonium acetate buffer and passed through Dionex Metpac CC-1 chelating resin columns. Trace elements were retained, then eluted with dilute acid and analysed off-line | 229 |
Various | Wood samples impregnated with inorganic preservatives | AE;Laser induced emission;S | No sample pre-treatment required. Nd:YAG laser focused onto the sample and pulsed for 7 ms. Plasma spark formed from which analytes were measured by their emission spectra | 230 |
Various | Osmium powder | AE;ICP;L | Sample heated with HNO3 in a distillation flask. Os matrix removed as oxide vapour (trapped in 10% NaOH solution). Sample diluted and measured directly | 231 |
Various | Trimethylgallium | AE;ICP;L | No sample pre-treatment required | 232 |
NUCLEAR MATERIALS— | ||||
137Cs | Micro-particles | MS;LI;S | Laser desorption and ionization followed by mass analysis in ion trap | 233 |
Cs isotopes | Waste U and MOX fuel | MS;IC-ICP;L | Cs isotopes determined using on-line ion chromatography coupled to ICP-MS. Sample injection = 200 µl, column = CS5 cation exchange (Dionex), eluent = 1 M nitric acid, LOD = 16 pg g−1. Quantification by isotope dilution using natural abundance Cs spike | 234 |
Er + Er isotopes | Er doped, high enrichment Mo–UO2 fuel | MS;TI, GD-MS;- | Double spike ID using 167Er and 233U with chemical separation prior to TIMS. Direct solids analysis using GD-MS with calibration against in-house standards | 235 |
Pu + Pu isotopes | Algae, sediments, soils and lichens | MS;HR-ICP;L | Spiked with 242Pu tracer, dry ashed at 500![]() | 236 |
Pu + Pu isotopes | Environmental materials | MS;ICP;L | Sample spiked with 242Pu tracer, ash extracted 3 times with boiling 8 M nitric acid. Combined extracts taken to incipient dryness, re-dissolved in nitric acid and diluted with water to give final nitric acid concentration of either 2 or 8 M, NaNO2 added to oxidize Pu to PuIV. Analyte separated on either AG1-X4 or TEVA and eluted with either NH4I or quinol. Eluate taken to dryness, residue taken back up in 4% v/v nitric acid, Bi added as internal standard and whole made up to 5 cm3. LOD = 0.05 mBq cm−3 (239Pu) and 0.17 mB cm−3 (240Pu) or 0.02 pg cm−3. Derived 239Pu + 240Pu value agreed with CRMs | 237 |
Pu isotopes | Environmental | MS;ICP, SF;L | After digestion, on-line, sequential separation of Pu using Sr-SPEC and TEVA-SPEC extraction chromatography resins automated using `PrepLab'. Sample introduction to ICP-MS via a CETAC 6000 desolvating MCN. LOD = 9 fg g−1 | 238 |
Pudau-ghters | Pu | MS;TI;— | Age determination of Pu material | 239 |
Pu, Sr, Y | Precipitates from brine storage solutions | XRF;ED;S µXRF;ED;S | Elemental mapping and bulk analysis | 240 |
226Ra | Mineral waters | MS;ICP,SF;L | Pre-concentrated on cation exchange resin. Eluate taken to dryness, digested with HNO3 (1 cm3, 1.1 M), diluted with water. MS in low-resolution mode, conventional pneumatic nebulization, recoveries = 98–102%, LOD = 0.01 pg dm−3 (500 cm3 aliquot). LOD with CETAC MCN 6000 = 0.004 01 pg dm−3 (500 cm3 aliquot) | 241 |
Tc | Environmental | MS;ICP;L | Bulk of matrix removed by evaporation/precipitation. Ru decontamination via TEVA | 242 |
99Tc | Biological | MS;ICP;L | Chemical separation of Tc followed by mass spectrometry | 243 |
230Th, 234U, 235U | Marine sediments | ICP; MS, SF, ID; L | HF–HClO4 or Na peroxide fusion. Analytes separated on an anion exchange column. MS resolution = 4430, yielded 230/232 abundance sensitivity = 5 × 10−7. 229Th and 233U spikes used for quantification | 245 |
233U | THOREX process solutions | AE;HR-ICP;L | Th separated by precipitation as oxalate from 1 M nitric acid. 233U determined in supernatant by high-resolution spectrometry at the UII 385.96 nm line | 246 |
U isotopes | CRM | MS;HR-ICP;L | Accuracy better than 0.2% | 247 |
234U/238U | Natural waters and carbonates | MS;ICP;L | U co-precipitated from sea-waters on Fe carrier, separated on TEVA. Carbonates decomposed with nitric acid. Internal mass discrimination correction based upon 191Ir40Ar+/193Ir40Ar+ | 248 |
U, Am, Cm | Spent fuel | MS;ICP;L MS;TIMS;L | U separated from fuel solution (8 M HNO3) by anion exchange and with 3 M HNO3. U fraction taken to dryness, re-constituted with 0.2 M HNO3 and subjected to ICP-MS and TIMS. Unretained fraction, containing Am and Cm taken to dryness, reconstituted and separated on a 5 µm Nucleosil SA column with an eluent of 2-hydroxy-2-methylbutyric acid at pH 4.1 with on-line α-spectrometry. Isotopic composition of Am and Cm fractions determined using double spike, isotope dilution methodology | 249 |
U, Pu | Calcites and bioassay | MS;ICP;L | LOD for 244Pu = 0.1 fg cm−3 using an ultrasonic nebulizer | 250 |
U, Th | Waters | MS;ETV-ICP;L | Chelation pre-concentration followed by ETV-ICP-MS. LOD = 24 pg (U), 60 pg (TH). LODs determined by blanks | 251 |
U, Th | Waters and pine needles | MS;ETV-ICP;L | On-line matrix removal. Various sample preparation techniques applied to pine needles, wet ashing, dry ashing and Li metaborate fusion | 252 |
U, Th, Pu isotopes | Environmental and wastes | MS;ICP;L | Cross flow and microconcentric nebulizers compared. External precision for NBL 050 at 1 ng cm−3 U = 0.05% RSD | 253 |
Various | Reactor grade U | AE;ICP;L | Uranium oxide ( ∼2.4 g) dissolved in minimum volume of 50% v/v nitric acid, taken to dryness and reconstituted in 50% v/v HCl. Analytes complexed with 1,2 diaminocyclohexane-NNN′N′-tetraacetic acid and 1,10-phenanthroline and U precipitated as hydroxide by addition of excess NH4OH. Sample filtered, washed with dilute NH4OH, supernatant acidified and made up to volume (50 cm3). Recoveries >90% | 254 |
Various (Na–Pb) | U3O8 | XRF;—;S | Single element LOD 10–20 ppm. Measured at peak (50 s) and 2 background positions (2 × 50 s) | 255 |
93Zr, 107Pd, 135Cs | Wastes | MS;ETV-ICP;L | Solvent extraction and chromatographic separation methods described | 256 |
The use of X-ray spectrometry in the study of fine art and archaeological artefacts is a fascinating area of analytical atomic spectroscopy. Conventional XRF can be used to identify pigments,260–263e.g., identification of blue pigments derived from either mineral or synthetic routes on Bronze-Age wall paintings from Greece and Cyprus.262 Simple identification of pigments may not provide sufficient evidence for a full characterization of a painting but the arrangement and sequence of paint layers is considered to be a distinguishing quality of an artistic studio or even an individual painter. This information can be derived from proton induced X-ray emission260,262 or grazing emission X-ray fluorescence spectrometry.
Forensic analysis can be considered as the direct ancestor of the study of fine art and archaeological artefacts by analytical atomic spectroscopy. The determination of trace element profiles was used to distinguish between plastic garbage sacks of nominally the same colour.264 After dissolution in HNO3–H2O2, the trace element profile of the polymers was determined by ICP-MS. Lead isotope ratios were also determined and found to be indicative of products that contained significant levels of Pb based additives, e.g., pigments. These approaches were considered to provide sufficient discrimination of the various products to be useful for evidentiary purposes. However, an attempt to extend this approach to polyethylene cling films was not successful due to the low intrinsic trace element content of these products.
The oxidation and carbon contamination of GaAs wet treatments has been studied by XPS, Auger electron spectroscopy and SIMS.266 Treatments studied include a variety of cleaning and etching pretreatment procedures prior to immersion in either (NH4)2S, Na2S aqueous solutions or S2Cl2 solution in CH2Cl2. It was found that S passivation removes surface oxide and minimizes C contamination in the surfaces treated in (NH4)2S and S2Cl2. Pretreatment in basic solutions showed significantly lower O and C levels than GaAs treated with anionic solutions. The authors also state that surface pretreatment performed ex-situ showed a higher risk of surface contamination prior to S passivation.
The comparison of shallow depth profiles of cobalt implanted silicon wafers as determined by total reflection XRF analysisand Rutherford backscattering spectrometry (RBS) after repeated stratified etching has been described over two papers.267,268 A novel method which combines a stepwise wet-chemical etching of an implanted wafer with total reflection XRF was compared to a traditional RBS method on prepared samples with 50–150 nm Co layers. A rectangular section of Si wafer was oxidized by treatment with 30% H2O2 for 40 min at 650°C. Then, 50 µl of HF was pipetted onto the wafer and, after the surface layer had been dissolved, the acid solution was transferred to a Plexiglas carrier and Se added as an internal standard. The mixture was then evaporated under an IR lamp and analysed by XRF. Depth resolution was 0.6 nm at best and the limit of detection 0.01% or 4 × 1018 Co atoms cm−3 for a wafer sample 3.4 cm2 area. The characteristic parameters of the profiles, e.g., concentration and depth at the maximum, mean depth, FWHM and total dose, showed a relative deviation of only 4–6% between both methods.
The peak concentration of nitrogen implanted into a Si wafer has been determined by the in-situ internal standard implantation of 14N+ followed by the SIMS depth profile analysis of 30Si14N−.269 As an internal standard, the N ions with a known fluence were directly implanted into the sample using a SIMS instrument. The depth profile of 30Si14N− was then measured. The actual concentration of the N was then evaluated from the measured ΔRp, and the ion intensity for each peak. The estimated concentration was in good agreement with the actual concentration and the depth profiles were also compared to the theoretical ones.
The spark source mass spectrometric (SSMS) assessment of boron and nitrogen270 concentrations in crystalline GaAs has been described. The MS instrumentation incorporated GaAs sample electrodes (15 mm, cross-section up to 4 × 4 mm) and a Mattauch–Herzog ion focusing system for simultaneous multi-element detection by Q plate. A detection limit of 4.4 × 1013 cm−3 was achieved for both elements. The SSMS method was also used as a reference method for the calibration of FTIR analysis of B and N in GaAs after comparison with TIMS for the reference determination of B. The method was further modified for the analysis of C,271 again comparing to FTIR data. For C, a detection limit of 1.4 × 1013 cm−3 was achieved by SSMS. The method is being further evaluated for several other elements to assess the suitability of SSMS as a reliable reference method for the analysis of GaAs. The same authors using this technique for C found a strictly linear relation to the total chemical C concentration as measured bySSMS.272 By using charged particle activation analysis as a reference method for C, a new calibration factor f77 = (7.2 ± 0.2) × 1015 cm−1 for the absorption integral at 77 K was derived. Based on the temperature dependence of the absorption, the authors arrived at a calibration factor f300 = (7.5 ± 0.5) × 1015 cm−1 for room temperature measurement.
The quantitative SIMS analysis of impurities in GeN and AlxGa1 − xN films using molecular ions MCs+ and MCs2+ (M = element to be determined) has been reported.273 It was found that under Cs bombardment the MCs2+ ions had a larger ion yield than the MCs+ ions when M was electronegative. Application of these molecular ions has made it possible for the analysis of both electropositive and electronegative elements in a single run. It was also found that these molecular ions minimized matrix effects in the AlxGa1 − xN film matrix. The authors postulate that these molecular ions are formed by recombination processes in which sputtered neutral species (M and/or MCs) combine with Cs+ ions.
A variety of techniques were reported for elemental mapping and depth profiling. These included: radiofrequency glow discharge atomic emission spectrometry;274secondary ion mass spectrometry;275–279secondary neutral mass spectrometry;275,280Auger electron spectroscopy;275 and electron probe microanalysis.275,281–283
Solar control coatings on architectural glasses were investigated comprehensively using Auger electron spectroscopy, SIMS, SNMS and EPMA.275 All methods found a system of two metallic Ag layers embedded between dielectric SnOx layers. Additionally, thin (1–2 nm) layers of Ni and Cr were detected on top of each of the Ag layers. The capabilities of each of the techniques were assessed critically for routine investigation of these coatings. This assessment concluded that the complementary nature of these surface analysis techniques allowed individual measurement artifacts to be deconvoluted and a rigorous characterization of the material to be obtained. The capabilities of rf-GD-AES were demonstrated for this class of coated glasses274 and the presence of an extraneous overlay of a Si containing material (20 nm) confirmed earlier SEM observations.
Defects in glasses were categorized and investigated by appropriate micro-analytical techniques.283 A combination of EPMA and LA-ICP-MS allowed the determination of elemental concentrations to the low µg g−1 level. This capability allowed identification of a specific source for the defect from otherwise non-distinguishable refractories. In a separate study, the migration of alkali metal in glasses under investigation by EPMA was investigated.281 Experimental conditions were derived to minimize this phenomenon and allow reliable quantitative analysis using EPMA.
Glasses of various types remain popular targets for laser ablation based sample introduction. However, in many cases, there is little technological interest in the glass itself. A detailed study and assessment of LA-ICP-MS for the depth profiling of silica based samples reported on elemental fractionation effects, as functions of spot size and laser fluence, for ablation at 1064, 266 and 248 nm.284 It was observed that the geometry of the crater controlled the extent of elemental fractionation and this could become significant when the depth : diameter ratio of the crater is >6. The use of He, rather than Ar, as a carrier gas was beneficial and, as is generally recognized for glasses, ablation in the UV was favorable. The conclusion was, not surprisingly, that a number of experimental parameters affected the utility of laser ablation in obtaining useful concentration : depth profiles. The following conditions all maximized the performance of the system: ablation in the UV, a large crater size at laser fluences well above the ablation threshold and a He carrier gas. The structure of the crater was described in terms of a three section model, i.e., an ablation front, intermediate section and crater opening.
Forensic and archaeological applications continue to grow and are closely related. The sources of variance in the use of ICP-MS for the forensic identification of glass fragments were considered.285,286 ANOVA was applied to the ICP-MS analysis of both NIST CRMs and to an inter-laboratory study of samples taken across a sheet of float glass. The output of this statistical analysis was used to provide a more accurate interpretation of the analytical data for forensic purposes and an increased understanding of the discrimination offered by an ICP-MS analytical methodology. Non-destructive analysis (NDA) of medieval glasses was used to identify manufacturing processes and technologies. This was accomplished by external beam PIXE and a statistical analysis using principle components methodology.287 Initial measurements were also undertaken using XRF, EPMA and LA-ICP-MS. A number of variations on X-ray spectrometry were tested for suitability in determining the major, minor and trace elemental content of medieval glasses.288 Scanning electron microscopy, with a wavelength dispersive spectrometer, was compared with PIXE for trace elemental analysis. Wavelength and energy dispersive spectrometers were compared for major and minor elemental analysis. The reported PIXE method was clearly superior in terms of detection limits but could not detect light elements of importance in glass research, e.g., Na, Al, Mg and Si.
Element | Matrix | Technique;atomization;presentationa | Sample treatment/comments | Ref. |
---|---|---|---|---|
a Hy indicates hydride and S, L, G and Sl signify solid, liquid, gaseous or slurry sample introduction, respectively. Other abbreviations are listed elsewhere. | ||||
POLYMERIC MATERIALS AND COMPOSITES— | ||||
Al | Polymers with gradient and composite structures | AA;—;— | Sample sectioned and sampled locally along gradient or composite structure | 309 |
Br, Pb, Hg | Polyethylene | µ-XRF;—;S | Foils (93–453 µm) prepared. Elemental mapping and bulk concentrations determined. Local concentrations estimated using the backscatter fundamental parameter method | 310 |
C, O | Polycarbonate and aluminized polycarbonates | XPS;—;S | Surface modification during plasma etching and metallization examined | 311 |
Co, Mn, P, Ti | PET | MS;ETV, ICP;S | Direct solids analysis. Pd–ascorbic acid modifier required for P determination. Results compared with XRF and ICP-AE after sample dissolution. | 312 |
Cu; S | Polyimide on Cu substrate | SIMS;—;S | Cu migration into polymer film studied. Polymer film prepared by either coating and curing polyamic acids onto the substrate or by vapor deposition of Cu onto a PI film | 313 |
Cu | TiOxNy films | SIMS;—;S | Cs adduct approach applied | 314 |
Cu, Fe, Co, Ni | Grafted polypropylene sheets | XRF;—;S | — | 315 |
Cu, Fe, Mg, Mn, Zn | Wood pulp | AE;ICP;S | Pyrolytic carbon direct sample insertion probe with in-situ treatment with HCl and NaF. Samples were dried and ashed prior to plasma ignition. LOD = 20–400 ng g−1 | 316 |
Cu, Fe | Ancient manuscripts | AA;ETV;S MS;LA-ICP;S | Manuscript micro-sampled and analysed by slurry sampling ETV-AA with calibration against aqueous samples. LA-ICP-MS used to define distribution of Fe and Cu on the manuscript surface | 317 |
Cr | Polyethylene | AA;—;— | Attempted Cr speciation by co-precipitation on an alumina carrier | 318 |
F | Ca silicate–fluoride composites | XRF;—;S | Sample fused to yield a glass bead using Li borate flux | 319 |
Fe | Cloth | XRF;—;S | Direct determination | 320 |
Pd | Polymers | MS;ETV, ICP;S | Calibration obtained either externally or by single standard addition. Ir or Ar dimer used as an internal standard. Absolute LOD ca. 1 pg or 1 ng g−1 relative | 321 |
Pd, Ag, Ni | Pd alloy films on stabilized alumina or zirconia supports | µ-PIXE;SEM-EDAX;S | Elemental maps and depth profiles obtained | 322 |
Sb | PET | AA;ETV;L | Treated with 4% v/v acetic acid. Characteristic mass 31 pg, recovery = 92–98% | 323 |
Sb | PVC | AA;ETV;S | Rapid screening methodology | 324 |
Sn | PVC | AA;ETV;S | Direct solids analysis with Pd as modifier. Two-stage pyrolysis. PVC (0.1–0.25 mg) suspended in 20 µl Pd modifier, whole pipetted into furnace, aliquot of ascorbic acid added, sample dried at 120![]() ![]() ![]() ![]() | 325 |
Ti, Si | TiSix thin films | Auger ES;—;S | Corrections for preferential sputtering and matrix composition applied | 326 |
Various | Cloth | AA;SE;L | Cloth extracted with a biological simulant (saliva and perspiration) at 40![]() | 327 |
Various | Polyethylene–polypropylene polymer blend | AE;ICP;L | 0.15 g sample wet-ashed with 2.5 cm3 of concentrated nitric acid at high temperature and pressure in a sealed quartz vessel | 328 |
Various | Finger paints | AE, AA;CV, HG, ICP;L | Microwave digestion followed by CVAA (Hg), HGAA (As, Se, Sb) and ICP-AE (Ba, Cd, Cr, Pb) | 329 |
Various | Pigments and fillers in polymers | AA;—;L | Bulk material micro-sampled using fine capillaries. Sub-samples combusted and these residues taken up in dilute acids | 330 |
Various | Pigments on medieval manuscripts | TXRF | Surface micro-sampled by rubbing gently with a cotton wool swab (`Q-Tip'). Abraded material transferred to a glass carrier | 331 |
Various | Pigments on medieval manuscripts and oil paintings | TXRF;—;— | Surface micro-sampled by rubbing gently with a cotton wool swab (`Q-Tip'). Abraded material transferred to a glass carrier | 332 |
Zr | Polyolefins | MS;ID-ICP;L | 0.25–0.3 g microwave digested in sealed vessel with HNO3–HF. Spike of enriched Zr added prior to digestion. Sample cooled, diluted with water and isotope ratio determined by QICP-MS. LOD = 18 ng g−1 | 333 |
SEMICONDUCTOR MATERIALS— | ||||
Al | Si | XRF;—;— | A combination of vapour-phase decomposition and total reflection XRF gave DLs of 2 × 109 atoms cm−2 | 334 |
B | Si | MS;ID-ICP;L | Boron doped thin films were dissolved in 0.3 M LiOH spiked with 10B. Uncertainties were <4% with DLs of 2.1 × 1017 atoms cm−3 | 335 |
Na | Si | XRF;—;— | As for Al. DLs 3 × 1010 atoms cm−2 | 334 |
O | Si wafer | SIMS;—;S | Quantification is based on 16O/30Si secondary ion signal. Problems with signal instability (RSD <2%) were overcome with vacuumizing for 2 h | 336 |
Various | Microelectronic materials | MS;ICP;L | On-line ion chromatography ICP-MS is compared with NAA, GD-MS and ETV-ICP-MS for the analysis of Mo, MoO3, MoSi2.5, W, WO3, W5Si3, As, P and Re | 337 |
Various (4) | Si wafers | MS;ICP;L | HNO3–HF was used to decompose the Si. Si was removed by evaporation as SiF. Recoveries for Fe, Ni and Cr were 95–106%. Microwave digestion gave better results for Zn but worse for Cu | 338 |
Various (9) | Semiconductor materials | XRF;—;S | A rapid thin layer method with the direct digesting of the materials on the substrate. Determination of Cr, Co, Ni, Cu, Zn, Ga, Se, Sb and Y with DLs of 0.034%–0.113% for a 0.5 mg sample | 339 |
Various (11) | YBa2Cu3O7-x | AE;ICP;L | ICP was optimized using the Mn 259.373 nm line for Ca, Mg, Fe, Mn, Al, Ni, Si and Sr. DLs ranged from 2.0 × 10−5–1.2 × 10−4% m/m | 340 |
GLASSES— | ||||
Cr | Metal coated glasses | AE;rf-GD;S | Depth profiling | 341 |
Fe | Metal coated glasses | AE;rf-GD;S | As for Cr | 341 |
N | Glasses and glass ceramics | EPMA;—;S | Energy and wavelength dispersive EPMA applied to determination of N | 282 |
Ni | Metal coated glasses | AE;rf-GD;S | As for Cr | 341 |
O | Nuclear waste glass | SIMS;—;L | Glass leached in solution containing enriched 29Si and 18O | 277 |
Si | Nuclear waste glass | SIMS;—;L | As for O | 277 |
Sn | Float glass | SIMS;—;S | Depth profiles obtained by low energy SIMS (4 keV O2+) with electron beam charge compensation. Quantification from RSF derived from float glass CRM | 279 |
Various | TV screen glass, glass ceramic, quartz | MS;ICP,LA;S | Excimer laser ablation. Results compared to those obtained from wet chemical analysis | 342 |
Various | Low-level radioactive waste, glass precursors | XRF;—;S MS;ICP;L | Mixtures of low active waste and glass precursors heat treated and resultant material analysed | 343 |
Various | Optical fibres | AE;ICP;L | Three sample preparation methods:100 mg of glass fused with 1 g of metaborate flux and resultant melt taken up in 150 cm3 concentrated nitric acid; slurry nebulization with 10 mg glass dispersed in 100 cm3 surfactant solution and Mn added as internal standard; 25 mg of glass taken up, with gentle heating, in 100 cm3 of 4 M boric acid and 3% v/v HCl | 344 |
Various | Glass fibres | SNMS;—;S | Depth profiling on as-prepared fibres; fibres exposed to humidity, water leach | 280 |
Various (5) | Glass-polyalkenoate cement | SIMS;—;S | Elemental mapping and depth profiles | 276 |
CERAMICS AND REFRACTORIES— | ||||
Ba | BaTiO3 ceramics | EPMA;—;S | BaTiO3 and Y2O3 were used as standards. A Joel JXA 840A electron probe microanalyser under a 20 kV, 50 nA beam current was used. Results suggest Y incorporated preferentially at the Ba-sites | 345 |
Cr | Silicon nitride | AE;ETV;Sl | Sample (50 mg) was treated with 0.5 ml of 60% PTFE emulsion, 0.5% agar, and 0.1% Triton X-100. RSDs between 1.9–4%. DL for Cr 1.58 ng ml−1 | 346 |
Cu | Rare earth oxide | AA;F;L | Spectral interferences are overcome by the use of PLS. Linear range 0–20 µg ml−1, RSD 2.84–5.10%, with recovery of 100.1% | 347 |
Cu | Silicon nitride | AE;ETV;Sl | As or Cr. DL for Cu 1.05 ng ml−1 | 346 |
Eu | Rare earth oxide | AA;F;L | Spectral interferences are overcome by the use of PLS. Linear range 2–18 µg ml−1, RSD 1.07–1.24%, with recovery of 98.6% | 347 |
Eu | Standards | XRF;—;S | Samples were separated and preconcentrated on thorin modified XAD-7. Preconcentration factors of ×500 were obtained. DLs of 13.8, 17 and 15.7 µg l−1 for Sm, Eu and Gd, respectively | 348 |
Gd | Standards | XRF;—;S | As for Eu | 348 |
Sm | Standards | XRF;—;S | As for Eu | 348 |
Ti | BaTiO3 ceramics | EPMA;—;S | As for Ba | 345 |
Various (15) | Eu, Gd | AE;ICP;L | Spectral interferences are overcome by the use of high resolution sequential spectrometer | 349 |
Various (15) | Dy,Sm | AE;ICP;L | Spectral interferences are overcome by the use of high resolution sequential spectrometer | 350 |
Various (15) | Geochemical samples | MS;ICP;L | Non-spectroscopic matrix effects were minimized using matrix matched standards and using Rh and In as internal standards | 351 |
Various | Gadolinium oxide | MS;ICP;L | Groups of REE were determined after various extractions using 2-ethylhexylhydrogen-ethylhexylphosphonate chromatographic separation | 352 |
Various | Praseodymium oxide | MS;ICP;L | Internal standards worked for REE except Tb which required a separation step using chromatography. Recoveries were 89.1–105% with DLs of 0.02–0.09 µg g−1 | 353 |
Various (14) | Eu2O3 | MS;ICP;L | Internal standardization successfully compensated for matrix effects for 13 REE. Tm had to be separated using chromatographic separation. DLs were in the range 0.005–0.021 µg l−1, RSDs 1.4–8.1% and recoveries 84–112% | 354 |
Various (7) | Praseodymium oxide | MS;ICP;L | Sample (10 mg) was decomposed with 2 ml of 4 M HCl, evaporated, re-dissolved in 2 M HCL plus Re internal standard. REE DLs were 5–21 µg l−1 with recoveries of 89.1–102% | 348 |
Various (14) | Lanthanum oxide | AE;ETV-ICP;L | Sample was treated with PTFE emulsion to produce volatile fluorides. DLs range from 2 ng ml−1 for Yb to 130 ng ml−1 for Ce with RSDs <5% | 355 |
Various | Yttrium oxide | MS;ICP;L | Sample (10 mg) was dissolved in a quartz vessel with HNO3–H2O (1∶3) on a hot plate. The effect of the Y matrix could be eliminated with Ga and an internal standard | 356 |
Various (21) | BaTiO3 powders | MS;ICP;L | To correct for polyatomic interferences a `blank' matrix was prepared and subtracted from results. Main impurities found were Sr and Ca | 227 |
Y | BaTiO3 ceramics | EPMA;—;S | As for Ba | 345 |
CATALYSTS— | ||||
C | Spent MnO2–CeO2 | XPS, S-SIMS;—;— | Carbonaceous deposits characterized | 357 |
C | Pt–Al2O3, Pd–SiO2 catalysts | SIMS, XPS;—;— | Carbonaceous deposits characterized | 358 |
O | Spent MnO2–CeO2 | XPS, S-SIMS;—;— | As for C | 357 |
This year, the use of ETV for the analysis of refractory samples, coupled to a variety of detectors, has proved to be popular. Peng et al. have analysed silicon nitride by a variety of methodologies. Using a PTFE emulsion, the vaporization behaviour of silicon and three refractory elements (Al, Ti, and Y) have been studied289 using ICP-AES detection. It was found that when using slurry samples and a 60 s ashing step at 700°C, approximately 90% of 100 µg of Si3N4 could be decomposed and evaporated without trace element loss. Detection limits varied from 0.11 µg g−1 for Al to 0.09 µg g−1 for Ti with RSDs ranging from 1.9–4.2%. This method was modified to include the determination of Cu, Cr, Al, Y and Ti.290 A 1% m/v slurry of Si3N4 was dispersed with an ultrasonic wave vibrator for 20 min before injection of 10 µl into the furnace. The results were in good agreement with those obtained by a dissolution based pneumatic nebulization ICP-AES method. Solid sampling ETAAS has been used for the direct determination of 11 trace elements in high-purity tungsten trioxide and high-purity tungsten blue oxide.291 The extremely high background from the matrix was eliminated by reducing the tungsten to the metallic form with the addition of hydrogen as a purge gas during the pyrolysis stage. Calibration was achieved using aqueous standards. Detection limits ranged from 0.07 ng g−1 for Mg, Na, and Zn to 1.7 ng g−1 for Fe. The multi-element analysis of graphite and silicon carbide by solid sampling ETV-ICP-AES has been reported.292 Vaporization of the trace metals was facilitated by the addition of Freon 1,2 gas (3 ml min−1) to the Ar carrier gas. Sample sizes of 2–16 mg of SiC (<5 µm particle size) produced detection limits of 5–250 ng g−1 for 15 target elements. The analysis of high-purity tantalum powders by ETV-ICP-AES has been published.244 Up to 14 trace elements were determined using an automated AWD 10 workstation to load the samples (8–40 mg) onto the pyrolytically coated graphite platform. Matrix removal was achieved during the ashing stage (1000
°C, 5 s). Calibration was performed using aqueous standards pipetted onto the platform. Detection limits ranged from 5 ng g−1 for Ag to 250 ng g−1 for K.
Several papers have been received this year on the analysis of cement and concrete products and materials. The determination of Th and U in activated concrete by ICP-MS after anion-exchange removal of the matrix293 has been reported. Although a few µg ml−1 of some matrix elements such as Al and Ca were not separated from the target analytes, no interference could be observed with the final determination. Instrument detection limits were 2.3 pg ml−1 for Th and 1.8 pg ml−1 for U with precisions of generally <7% in the solid. The accuracy of the method was tested using the GSJ rock standard with recoveries ranging from −18% to 0.4% for Th and −14% to −5.7% for U.
An unusual variation on the concrete theme has been published by Groenewold et al. with a description of the analysis of the nerve agent VX (O-Et,S-2-diisopropyl-aminoethyl Me phosphonothiolate) on the surface of concrete samples.294 The authors used an ion-trap SIMS instrument to determine VX down to an absolute quantity of 5 ng on a concrete chip. To get down to these levels of detection the m/z 268 and 128 ion fragmentation was measured using MS-MS where 268 corresponds to [VX+H]+ and 128 corresponds to a diisopropylvinylammonium isomer that is formed by the elimination of the phosphonothiolate moiety. The authors were able to show that VX degraded on the concrete surface over time. However, the degradation products of VX were shown to stay detectable on the surface of the concrete for some length of time. The authors did not mention whether these degradation products were toxic as well *** UNDECLARED ENTITY ***
A ceramic and refractory section would not be complete without some mention of what has been going on in the world of laser ablation. The laser ablation ICP-MS analysis of sintered silicon nitride has been reported.295 The surface of the SiN was subjected to ablation for 3 min using a Q-switched Nd:YAG laser (150 mJ) operating at 1064 nm. The laser sampled particles were carried in a Ar stream and passed through 10 ml of 0.1 M nitric acid, which was subsequently analysed by ICP-MS. The method was successfully used for the determination of Mg, Ti, Mn, Co and W in sintered silicon nitride. A method for the quantitative determination of trace elements in single SiC crystals using LA-ICP-MS has been developed.296 As above, a 1064 nm Nd:YAG laser was used in `free running' mode as this provides better signal-to-noise ratios than when using the laser in Q-switched mode. No sample preparation was necessary other than purifying the crystal surface with HF. Calibration was achieved by adding multi-element standard solution to SiC powder. The powder was dried and pressed into pellets with carbon powder as a binding material. Recovery rates from the analysis of SiC reference materials ranged from 95 to 101%. Detection limits were between 3 × 10−9 g g−1 for V to 4 × 10−8 g g−1 for Cu. RSDs at the 1 × 10−6 g g−1 level were <10% when the intensities of 10 craters, each with 500 laser shots, were averaged.
One last word for this section is that a larger than normal selection of papers has been noticed for the analysis of archaeological ceramic-type samples. Methods used range from NAA,297 DCP,298 PIXE,299–301 XRF,302–305 ICP-AES,306 and LA-MS.307
This journal is © The Royal Society of Chemistry 2000 |