Simon
Carter
a,
Andy S.
Fisher
*b,
Phill S.
Goodall
c,
Michael W.
Hinds
d,
Steve
Lancaster
e and
Sian
Shore
f
aHull Research & Technology Centre, BP, Saltend, East Yorkshire HU12 8DS
bSchool of Geography, Earth and Environmental Sciences, University of Plymouth, Drake Circus, Plymouth, UK PL4 8AA
cBNFL Sellafield, Seascale, Cumbria, UK CA20 1PG
dRoyal Canadian Mint, 320 Sussex Drive, Ottawa, Canada K1A 0G8
eRoyal society of Chemistry, Thomas Graham House (290), Science Park, Milton Road, Cambridge, CB4 0WF, UK
fShell Research and Technology Centre, Thornton, P.O. Box 1, Chester, UK CH1 3SH
First published on 19th October 2011
There is considerable interest in the non-destructive analysis of archaeological or historical materials (glasses, ceramics, paintings, materials etc.). The use of solid sampling techniques that cause minimal damage is therefore still gaining in popularity. This is especially true for laser-based techniques such as LIBS and laser ablation, although the many variants of the X-ray-based techniques are also still proving very popular. Non-destructive analysis is also the ideal scenario for forensic scientists and therefore many of these applications are using similar techniques. The increasing trend to use multiple analytical techniques, ideally simultaneously, to cause minimal damage and to obtain the maximal number of results in the shortest time, is also noted. The technique of LIBS, which offers minimal sample damage and a “stand-off” capability is still gaining in popularity, although there are still question marks regarding its quantitative capabilities for some sample types. There is also considerable interest in the growing area of thin films and depth-profiling. Substantial research is on-going to develop methods to improve depth-resolution and several different approaches have been described in the literature. These approaches often use SIMS with either a lower energy primary beam or a primary beam consisting of polyatomic molecules.
This is the latest review covering atomic spectrometric measurements of industrial materials, metals, chemicals and advanced materials. It follows on from last year's review1 and should be read in conjunction with other reviews in the series.2–6 The reader may also be interested in a 25 year retrospective of Atomic Spectrometry Update reviews.7
A non-destructive sampling technique for elemental analysis of installed metal components was reported by Dannecker and co-authors.12 Metal components and weld joints were abraded with high purity abrasive paper which acted as a sample collector. The abraded materials were dissolved in acid and analysed using ICP-OES. The technique was given the name Quantitative Abrasive Test and was validated using metal certified reference materials (CRM).
Gu et al. reported that the determination of FexOy speciation in solid materials was possible by using glow discharge-mass spectrometry (GD-MS).13 Ratios of Fe+ and FeOH+ intensities were used to discriminate between FeO, Fe2O3 and Fe3O4. The composition changes surrounding sulfide inclusions in steel were determined by high resolution secondary ion mass spectrometry (SIMS) as reported by Williams and co-workers.14 Awane et al. reported an effective SIMS method to estimate background levels of H gas in austenitic stainless steel and the surface before and after H charging to determine the net H charged into the material.15 Previously, this had been a difficult determination because distinguishing the H present in moisture, hydrocarbons and other organic materials from the SIMS signal was impossible. This publication described a method by which the background H signal could be reduced and estimated.
Niobium-stabilised steels have been analysed in several papers over this review period. The oxidation kinetics of un-stabilised and Nb-stabilised AISI 430 ferritic stainless steel at high temperatures were studied by Sabioni and colleagues.16 The microstructure of the oxide films were examined using SEM with energy dispersive XRF detection (SEM-EDX), grazing X-ray diffraction and SIMS. For un-stabilised steel the atmospheric composition strongly affected the oxidation whereas the Nb-stabilised steel was less affected. The corrosion resistance of niobium implanted SS316L steel used in a fuel cell was investigated by Feng et al. who used ICP-OES and X-ray photoelectron spectroscopy (XPS) for the analysis.17 After a potentiostatic test the solution was analysed using ICP-OES in an attempt to determine corrosion products. The surface was examined using XPS after the test, and the depth-profiles indicated that niobium oxide was formed.
Another paper to have undertaken depth-profile measurements was presented by Konarski and co-workers.18 Depth-profile measurements of two types of stainless steel after vacuum heating were achieved using SIMS and GD-MS for the analysis. The vacuum annealing was undertaken at 750 °C and at 900 °C. Results from the two analytical methods were similar, but GD-MS had better resolution because of the lower sputtering energy. In addition, it also used multi-directional ion bombardment whereas the SIMS bombardment was unidirectional.
Pande et al. examined eight ferroalloys for impurities using ICP-OES and an analyser capable of determining C, H and O.19 The microstructure of the alloys was examined using SEM-EDX. The inclusion impurities observed in the microstructure were in agreement with impurities determined using ICP-OES. These two techniques were also used by Eliyahu and co-authors who determined the composition and microstructure of two iron anchors found at a shipwreck site.20 Both anchors had the same composition indicating that they likely belonged to the same wreck.
Keane and colleagues used solution nebulisation ICP-OES and XRD to determine the concentration and speciation of Mn generated in gas metal arc welding fumes originating from stainless steel.21 Other elements, e.g. Fe (both Fe2+ and Fe3+) and Ni were also determined. Once collected, the welding fumes underwent a sequential extraction protocol that involved extraction using phosphate buffered saline and reduction using hydroxylamine followed by phosphate buffered saline extraction. The remnants were extracted using aqua regia. The elemental concentrations in fractions soluble in a biological buffer solution and soluble in aqua regia were determined using ICP-OES. The XRD analysis showed that the fumes material contained multiple oxides of Mn.
Low pressure LIBS was used by Kim and co-workers to determine Cu, Nd and Sm in copper binary alloys.24 The four Nd–Cu and five Sm–Cu alloys were prepared from homogeneously mixed powders and melted at 3000 °C in an arc melting furnace. Linear regression coefficients for Nd and Sm were reported to be 0.9913 and 0.9535 respectively.
Calcagnile and colleagues determined the age of ancient bronze statues using accelerator mass spectrometry (AMS) radiocarbon dating on the organic residues contained in the casting cores.25 Organic materials such as charred wood, vegetable remains and animal hairs were found in the casting cores and used for the analysis. Resano et al. used LA-ICP-MS with different single collector devices (sector field and time of flight (TOF)) to determine Pb isotope ratios in ancient bronze coins.26 Isotope ratio precision values for 207Pb/206Pb and 208Pb/206Pb were in the range of 0.15–0.30% with Pb occurring at an average concentration of 5% in these bronze alloys. An internal standard of Tl was introduced simultaneously through the nebuliser/spray chamber assembly and helped to partially overcome noisy signal problems. It had the additional advantage of being capable of correcting the mass discrimination effects. The TOF-based instrumentation was found to provide better results for the transient peaks produced by the LA system.
Vanhaecke et al. described the determination of platinum group metals (PGM) in lead, produced from fire assay, using the analytical technique of laser ablation (LA)-ICP-MS.27 The fire assay lead button acted as a separation and pre-concentration step for the analytes Ir, Pd, Pt, Rh, and Ru extracted from ore samples. Numerous operating parameters were optimised including the laser spot diameter, fluence and repetition rate. The laser used was an IR laser operating at 795 nm and had a pulse duration of 150 fs. A comparison of argon and helium carrier gases was made, with the argon being chosen for future work because the helium provided no improvement in LOD and also led to sample deposition on the chamber walls. Any spectral interferences observed were overcome by the introduction of ammonia to the reaction cell of the ICP-MS instrument. Calibration was achieved by matrix matched (i.e. greater than 99% Pb) standards, which provided excellent accuracy. Detection limits of < 0.010 μg g−1 for Pd, Pt, Rh and Au were reported.
The interfacial diffusion of Zn into nano-structured copper was studied using SIMS in a paper by Wang and co-workers.28 The SIMS experiment was conducted in the temperature range 358–463 K, at which temperature volume diffusion was negligible.
The analytical technique of XRF was used as part of a two step testing program to identify Pb containing jewellery in a study by Cox and Green.29 Over 1500 pieces of jewellery were purchased in California and analysed using a two step process (XRF analysis followed by laboratory verification). About 4% of the pieces were found to be non-compliant with Californian health standards. Hurst et al. determined Cr, Fe, Mn and Ni in welding fumes using XRF as a means of detection.30 Proficiency test samples were analysed using a conventionally calibrated XRF method and then again with a calibration using the Uniquant fundamental parameter software package. On average, the conventional XRF method was accurate to 92–103% of target values with precision values of 3–7% RSD. The Uniquant XRF method had recoveries of 97–113% with RSD values 3–10%. The authors concluded that the Uniquant method produced results within acceptable analytical expectations. Fujisaki et al. used a high precision milling machine in conjunction with an XRF instrument to provide three dimensional microscopic elemental analysis of inclusions to study fractures inside industrial materials.31 The milling machine removed precisely measured layers of the material prior to each set of XRF analyses to build a three dimensional elemental picture. The procedure was applied to the analysis of braided wire.
The technique of SEM - energy dispersive XRF (SEM-EDX) was used by Saeed and co-workers to examine the cross section of welds made from different brazing alloy fillers.32 It was found that weld properties depended on the alloy filler used and size of the weld nugget formed. A paper by Valerio et al., described the analysis of a collection of 54 bronze artefacts dating from the 9th and 8th centuries BC, using EDXRF, μ-EDXRF, SEM-EDX and optical microscopy.33 The average Sn composition of the good quality bronze collection was 10.1 ± 2.5%. Canovaro and co-workers studied the composition of three 5th century AD bronze coins using XRF and SEM-EDX (in Italian).34 The presence of Pb on the surface of the coins was found to prevent XRF from giving a true measurement of the bulk of the coin without surface preparation. The SEM-EDX measurements on metallographic cross sections of the coins were, however, more successful and provided a more accurate measurement of the coin material.
A paper focusing on integrated analytical methodologies for the study of corrosion processes in archaeological bronzes was written by Alberghina et al.35 The authors found that optical microscopy, SEM-EDX, XRF and LIBS were useful in characterising the corrosion patinas. Depth-profile information of the patina layers and the bulk metal was obtained using LIBS. This depth-profiling also enabled migration of ions from the bulk of the material to the surface to be determined.
A comparison of data from seven laboratories for the determination of Cr, Cu, Fe, Mn, and Zn in an aluminium alloy was presented in a paper by Recknagel and co-workers.36 The results from all laboratories were acceptable and the variety of traceable methods included: instrumental neutron activation analysis (INAA), XRF (of fused beads), ICP-OES and ICP-MS. The element concentrations ranged from 0.05–0.2 wt%.
Analysis of metallic surfaces after cleaning processes was the subject of several research papers. Konarski and Opalinska used SIMS and Fourier transform-infra-red (FT-IR) to study cold plasma cleaning of copper and aluminium.37 Measurements at the nanometer level using SIMS indicated the remaining film consisted of metal oxides, oxidised hydrocarbons and hydrocarbons at the metal surface. The efficiency of the cleaning process was monitored using FT-IR. Laser cleaning of metal artefacts was the focus of two abstracts. Buccolieri and co-authors used a UV laser to clean ancient silver artefacts and monitored the composition of the surface using EDXRF and XRD.38 They found that the concentration of S decreased by up to 20% with a 280 J cm−2 laser treatment and with a laser fluence of 50 J cm−2. After treatment, the coin looked visibly cleaner. Drakaki et al. used XRF to examine the surface of Roman coins after laser cleaning.39 Two lasers were compared for the cleaning process; a Q-switched Nd:YAG laser operating at 1064 nm and with a pulse duration of 6 ns and a GaAlAs diode laser operating at 780 nm and with a 90 ps laser pulse.
Tin-zinc coating alloy on steel produced under different electroplating conditions was characterised using SEM-EDX, XRF and GD-OES in a paper described by Dubent et al.40 Kasamatsu and co-workers used synchrotron radiation (SR)-XRF and ICP-OES for the forensic discrimination of aluminium foils from different manufacturers (in Japanese).41 A comparison of the correlation coefficients for the ratio of element concentrations determined using ICP-OES versus normalised X-ray fluorescent intensity showed good agreement: 0.997 for Cu/Fe, 0.999 Zn/Fe, and 0.999 Ga/Fe. The authors concluded that the SR-XRF method was nearly as effective for the discrimination of aluminium as ICP-OES with fully quantitative analysis. The SR-XRF method was, however, able to analyse 1 mm2 samples of aluminium foil non-destructively.
Kaciulis and colleagues employed XPS, Auger electron spectrometry and photoemission spectro-microscopy to characterise the carbon-metal interface of a Ti6Al4V alloy with SiC fibres (coated with graphite).42 In another paper, Kaciulis et al.43 examined tungsten coatings on CuCrZr alloys using EDXRF, XPS and Auger electron spectrometry. The outer W layer was analysed using ICP-OES. It was found that there was no diffusion of material between layers and that the presence of Ni (0.4%) in the outer W layer was an impurity in the W powder used in the coating. Nickel was not observed in higher purity W. Chang et al. determined the high temperature microstructure of a biomedical titanium alloy using optical microscopy, SEM and EDXRF.44
There was a large number of abstracts that involved different spectrometric methods to analyse metal archaeological artefacts. Galli and co-workers used EDXRF to determine the elemental composition of metal artefacts from the Royal Tomb 14 of Sipan, Peru.45 From μ-Raman spectroscopy measurements, the patina on the artefacts was identified as being malachite, atacamite and magnetite. Torrisi et al. used LA-ICP-MS, XRF, LIBS, SEM, optical microscopy, surface profile analysis and density measurements to analyse ancient Egyptian bronze coins.46 Differences were observed in the patina composition, bulk alloy composition, isotopic element ratios and the surface morphology of the coins. Results also indicated that the coins were produced at two different sites. A group of fake coins was also identified. Figueiredo et al. studied late Bronze Age metal artefacts, metallurgical debris and slag fragments from a habitat site in Portugal using the techniques of optical microscopy, μ-EDXRF, SEM-EDX and XRD.47 Compositional analysis indicated that bronze was typical of the area (13 ± 3 wt% Sn and < 0.1wt% Pb) but unique from both Atlantic and Mediterranean bronze, which tended to have higher Pb values. Mudronja and co-workers analysed a bronze statue raised from the north Adriatic Sea in 1999.48 Dating using 14C AMS on the organic material found inside the statue indicated it was made in the range 100 BC–250 AD. Portable XRF and particle induced X-ray emission spectrometry (PIXE) surface analyses were not accurate compared with cross sectional μ-PIXE measurements because the surface of the statue had deteriorated through contact with the seawater. The authors claimed that accurate alloy composition was only determined when measured a minimum of 0.6 mm below the surface. Lead isotope analysis using multi-collector (MC)-ICP-MS indicated that Pb isotope composition was consistent with lead ores from the Eastern Alps or from Sardinia. It was concluded that the statue was not of Greek origin but probably a Roman copy of the original Greek statue. A paper by Young et al. re-examined ancient Chinese bronze fragments using SIMS, ICP-OES, SR-XRD, Raman microscopy and SEM-EDX and compared these results with those obtained using optical metallography and powder XRD that had been published in 1951.49 The modern methods provided a more complete assessment of the composition of the alloy and the corrosion products. The authors discussed the possibilities and limitations of various instrumental techniques available for studying bronze artefacts. Some of the more routine applications have been discussed in tabular form (Table 1).
Element | Matrix | Technique; Atomisation; Presentation | Comments | Reference |
---|---|---|---|---|
Ag | Silver coins | XRF; -; S | The average composition of Greek silver drachmae were determined to be around 98.3%. | 50 |
Co and Cr | CoCr Alloys | OES; GD, S | GD-OES was used to depth-profile corrosion layers and compositional changes to the alloy from oxygen treatment to evaluate the alloy for use in human implant prostheses. | 51 |
Cu | Steel | MS; ICP; L | Solvent extraction on a microchip using 8-hydroxyquinoline in o-xylene then back extracted into 0.1M nitric acid. The Cu was determined using isotope dilution (ID)-ICP-MS. The total measurement time was 40 s. | 52 |
D | Ti | OES; LIBS; S | Experiments using LIBS in a low pressure helium atmosphere enabled determination of deuterium. A LOD of 40 μg g−1 was obtained. | 53 |
Nb | Nb | OES; ICP; L | Niobium metal was dissolved using sulfuric acid in a PTFE vessel (microwave digestion). Recovery was reported to be 99.90 ± 0.08% and this method was also effective for dissolution of niobium pentoxide and niobium pentafluoride. | 54 |
Nb | Zr–Nb alloy | XRF; -;L | Zr–Nb alloy was dissolved and an aliquot absorbed on a filter paper prior to WDXRF determination of Nb. Precision was better than 0.8% RSD (n = 6) when the concentration determined ranged from 1–8%. | 55 |
Sn | Bronze | MS; ICP; L | A multi-collector instrument was used which provided Sn isotope ratios. This technique was applied to the analysis of bronze and ore samples to establish the provenance of bronze materials. | 56 |
Ti | Steel Wire | MS; ICP; S | Content of Ti increased and then decreased during an electrical arc furnace tapping - refining - continuous casting process. | 57 |
Zn | Steel | AAS; ETA; S | Calibration against aqueous standards. Using a 15 mg sample the LOD was 0.86 μg g−1. | 58 |
Various (3) | Al–Cu–Si alloys | XRF; -; S | The composition of macro-segregation of Al–Cu–Si alloys was determined using XRF. | 59 |
Various | Molybdenum | MS; GD; S | Effects of plasma arc melting on the purification of molybdenum were assessed using GD-MS for the analytical measurements. Purity of the materials was found to be between 99.9943 and 99.9996%. | 60 |
Various | Nickel alloys | OES; LIBS; S | Determination of elements in nickel alloys using LIBS. | 61 |
Various (23) | Un-Alloyed, alloyed and highly alloyed Steel | MS; ICP; S | A femtosecond LA system gave stable and representative results when in scanning mode when large ablation craters of 250 μm diameter were used (temporal RSD of ∼ 10%). The exceptions were for Pb and Bi which yielded a temporal RSD of 46%. A large number of CRMs were used for method validation purposes. | 62 |
Various | High purity zinc | MS; GD; S | Triple vacuum distillation used to purify zinc materials and GD-MS used to determine impurity levels. Metal impurities reduced from 43 mg kg−1 to 0.5 mg kg−1. | 63 |
Various (9 REE) | zirconium | MS; ICP; L | Nine REEs were quantitatively separated from Zr by precipitation using mandelic acid followed by solvent extraction. LOD were in the range 0.9–1.1 μg kg−1 for Ce, Dy, Eu, Gd, La, Lu, Nd, Sm, and Yb. | 64 |
The references quoted in the paragraph above will be discussed at greater length in the relevant sub-sections below.
Flow injection (FI), or derivatives of FI have been used by several workers. An interesting paper was presented by Garcia and colleagues who modified an anionic resin with di-2-pyridyl ketone salicyloylhydrazone to use it as a solid preconcentration phase for Cu determination in ethanol fuel samples, prior to spectrophotometric determination.73 Good agreement with results generated using AAS was observed, and low μg L−1 detection was achieved. The method may be adaptable to use in the field. An ICP-OES instrument fitted with a high temperature single pass spray chamber was used by Sanchez and co-workers for the injection of a sample plug into an air carrier gas stream.74 The aim of the study was to alleviate non-spectral interferences produced by petroleum products and organic solvents (11 alkanes), and to reduce plasma loading. Drop size distributions for primary and tertiary aerosols were measured, and the effect of both solvent type and temperature was studied. It was concluded that heating the injected sample volume to 200 °C mitigated differences in analyte mass transported as a function of the solution matrix.
Vorapalawut et al. developed a method for direct analysis of petroleum samples using laser ablation (LA)-ICP-MS.75 A silica gel plate was impregnated for 30 min with the sample, and then analysed. A doubly focusing sector field mass analyser confirmed that carbon related polyatomic interferences were not present. A significant advantage of the method was that matrix effects observed in nebulisation-based sample introduction were not present, which enabled quantification by external calibration. The method was also applied to the analysis of crude oil and asphaltene samples. This work highlighted the possibility of LA-ICP-MS as a useful technique for routine analysis of petroleum products.
Cinosi and colleagues presented a method using total reflection X-ray fluorescence (TXRF) to determine 15 trace metals in various petrochemical products.76 The sample preparation was simple, and consisted of depositing the sample onto the sample holder and evaporating the matrix, prior to analysis. The results compared favourably with the reference method (ASTM D5708) and the advantages and disadvantages of both procedures were discussed. In most cases, LOD were ∼0.005 μg g−1, but were up to 0.05 μg g−1 for a few analytes.
Determination of Cu and Fe in fuel ethanol using ICP-OES equipped with an ultrasonic nebuliser with membrane desolvator was presented by Rocha et al.77 While this procedure has been used in industry for some time, the additional work in this paper relating to the nebulisation efficiencies of anhydrous and hydrated ethanol is valuable. A 30% loss of analyte signal was observed when samples contained 7% water, which highlighted the requirement to carefully matrix match samples and standards. Detection limits were 0.1 and 0.5 μg kg−1 for Cu and Fe respectively.
Alkaline hydrothermal de-ashing and desulfurisation of low quality coal, and its application to hydrogen-rich gas generation was described by Mursito and co-workers.80 Experiments were carried out in a laboratory scale reactor, using high S, high ash coal from Java, Indonesia. The alkaline hydrothermal treatment gave upgraded clean coal with S content of approximately 0.3% (m/m), and ash content of approximately 2.1% (m/m). Sodium hydroxide was added to the hydrothermal treatment, and at 330 °C, hydrogen gas was produced. A variety of techniques including XRD, XRF, FTIR, and 13C NMR, were used to analyse the upgraded coal, to determine whether or not there had been a reduction of inorganic elements and to identify any changes of carbon functional groups.
Maslov et al. investigated the concentrations of natural radionuclides and the feasibility of recovery of U from brown-coal ash of a Mongolian cogeneration plant.81 The concentrations of the elements were determined using instrumental gamma-activation (IGAA) and XRF analysis. Leaching of Th and U was achieved using an anion exchanger using 8M nitric acid and 10% hydrofluoric acid. The resulting ash, which contained neither U nor Th, was suitable for use as a building material.
A new method that enabled analysis of minor and trace elements in petroleum cokes using WDXRF was presented by Gazulla and co-workers.82 The elements present in the coke determine its end use, and a reliable rapid method is therefore highly desirable. Standard WDXRF analysis methods determine only a few elements of interest at the required detection limits, and the standard ICP-OES method requires a long sample preparation time and is labour intensive. The proposed new method was described as being rapid and involved simple preparation of a pressed pellet before analysis using WDXRF. Reliable results were obtained for a wide range of elements (24), with low LOD being obtained. Method validation was achieved by the analysis of a range of CRMs, including SRM 1632c, SRM 2718, SRM 2719, SRM 2685b, AR 2771, AR 2772, SARM 18 and SARM 19.
Pohl et al. reported interesting work on multi-element molecular size fractionation in crude oil and oil residue.83 Micro-chromatography using permeation through gels with increasing exclusion limit was combined with high resolution ICP-MS detection (with a resolution of 4000) to measure the distribution of elements (Co, Cr, Fe, Ni, S, Si, V and Zn) as a function of the molecular mass fraction. The accuracy was determined using CRMs for wear-metals in lubricating oils: SRM 1084a and SRM 1085b.
A method for the direct determination of trace elements in crude oil was proposed by Ricard and co-workers who used femtosecond pulses of an IR laser at high repetition rates (up to 10 kHz) to ablate viscous crude oils prior to the determination of trace elements using ICP-MS.67 An internal glass cap was fitted into the ablation cell to minimise oil splashes and to prevent large particles from spreading into the cell. Both static and dynamic LA were studied, together with the effects of repetition rate and fluence. A strong relationship was observed between the sensitivity and stability of the signal and repetition rate and fluence, although the effects were not linear. In some circumstances the laser beam velocity also influenced the signal. Transport efficiency decreased with increasing repetition rate, and it was proposed that this may have been because of stronger particle agglomeration when increasing the density of the primary particles. The plasma atomisation/ionisation efficiency was also affected by the highest repetition rates. It was noted that matrix effects were observed, and that matrix matching of standards or a standard addition calibration approach was required. The method developed achieved good accuracy and low (ng g−1) detection limits. Several papers discussed the determination of V in crude oil. Dechaine and Gray presented a critical discussion of the distribution and form of V compounds within vacuum residue.70 The implications of this chemistry for possible new separation methods are considered in some detail, giving a good basis for future experimental work. The overview contained 166 references and, among the other topics, it discussed the use of some of the more novel X-ray based techniques; including X-ray absorption near edge spectrometry (XANES) and Extended X-ray absorption fine structure (EXAFS).
The use of LIBS has now been extended to the analysis of crude oil and oil residues. A paper by Fortes and colleagues described the use of LIBS to determine a suite of elements (up to 11) in an attempt to discriminate between crude oils and oil spill residues.84 A clear difference in chemical signature between residues and the crude oil were observed.
Jin and co-workers reported the use of a new catalyst for ultra-fast microwave biodiesel production.87 A yield of greater than 95% fatty acid methyl ester was obtained in less than five minutes under mild reaction conditions using less than 1% (m/m) ZnO/La2O2CO3 heterogeneous catalyst. Analysis of the material using XRF indicated that there was no catalyst leaching into the reaction medium. Scaling up of this work would be an interesting topic for future investigation. Salamatinia et al. investigated the use of an ultrasonic processor in the heterogeneous trans-esterification of palm oil for biodiesel production.88 Response surface methodology, a chemometric approach to determining the effects of different variables on a signal response, was used to evaluate and optimise biodiesel production using two catalysts, barium oxide and strontium oxide. Four variables, including reaction time, alcohol to oil molar ratio, catalyst loading and ultrasonic amplitude were optimised. Mathematical models were developed, and were able to accurately predict the biodiesel yield with less than 5% error for both catalysts. The study confirmed that ultrasonic processing reduced the reaction time and the catalyst loading. Alcohol to oil ratio and amplitude were also optimised, and the results of this optimisation were discussed. The technique of AAS was used to characterise the catalysts.
Corrosion of austenitic 304 stainless steel in biodiesel was considered by Gallina et al.89 Although the paper was in Portuguese, it was worth noting because this grade of steel is used in fuel distribution and storage, and there are few literature reports on corrosion by biodiesel. This study considered the corrosive behaviour in the presence of biodiesel both washed and unwashed with aqueous solutions (0.01 M) of citric, oxalic, acetic and ascorbic acids. The elements determined were Cr, Ni and Fe and the method of detection was AAS. The greatest concentration increase was found for Cr, which had a dissolution rate of 1.78 mg kg−1 day−1, with or without washing.
Abreu and colleagues presented an interesting study on ash deposition during the co-firing of bituminous coal with pine sawdust and olive stones.65 The main objectives were to relate the ash deposit rates with the type of biomass burned, and its thermal percentage in the blend. The thermal percentage of biomass varied between 10% and 50% for both sawdust and olive stones. Deposits were collected with the aid of an air-cooled deposition probe, away from the flame region. Analysis of the deposit was carried out using SEM equipped with an EDXRF detector (SEM-EDX). Results indicated that blending sawdust with coal decreased the deposition rate, when compared with the firing of unblended coal. The deposits from the blended sawdust firings contained high levels of Al and Si, which indicated a high fusion temperature, with less capacity to adhere to the surfaces. In contrast the firing with olive stones increased the deposition rates. High levels of K were present in the ash, which could increase the stickiness.
Element | Matrix | Technique; Atomisation; Presentation | Comments | Reference |
---|---|---|---|---|
Eu and Tb | Solid | TXRF; -; S | Direct analysis of solids for these lanthanides in the fluorescent organic-inorganic hybrid compound gamma-zirconium phosphate – terpyrimidine gave much better accuracy than conventional acid digestions followed by ICP-MS analysis which resulted in losses of Eu and Tb. 5 mg sample sizes were analysed. | 135 |
Hg | Vinegar | AFS; -; CV | Matrix assisted photochemical CV generation. Conditions required for the matrix (acetic acid) assisted reduction of Hg to elemental Hg were investigated. 3% acetic acid was optimal with 30 s irradiation from a 20 W mercury lamp. | 136 |
V | Pharmaceutical formulations | AAS; ETA; L | Cloud point extraction/preconcentration using 8-hydroxyquinoline and Triton X-114. A V enrichment factor of 125 was obtained using a volume of 50 mL of sample. The LOD was 42 ng L−1. | 137 |
Various | Levantine rock art | EDXRF; -; S | Field portable instrumentation is becoming increasingly important for in situ analysis and a portable EDXRF instrument was used. Fe/Ca ratios were indicative of the degree of preservation. | 138 |
Various | Pigment characterisation in cultural heritage analysis | EDXRF; -; S | This paper outlined a comparison of three portable XRF instruments. Advantages and drawbacks of each one noted. | 139 |
Various | Historical documents | EDXRF; -; S | Forensic analysis of historical documents. Different paper manufacturers identified from early 20th century papers. | 140 |
Various | Atmospheric aerosols in and around objects of cultural heritage. | XRF; -; S EPMA | Electron probe microanalysis enabled the analysis of single aerosol particles. An attempt to enhance preventive conservation. | 141 |
Various | Wine, vine leaves and grapes | EDXRF; -; S TXRF; -; L | TXRF used for the quantitation of various analytes in wine and must. EDXRF used for grapes and leaves. | 142 |
Various | Wine | OES; ICP, L IR and a range of physico-chemical parameters | Multivariate data analysis (PCA and discriminant analysis) differentiated wines from different regions. | 143 |
Various | Wine | OES; ICP, L AAS; ETA; L | Numerous chemometric (LDA, soft independent modelling of class analogy (SIMCA), k – nearest neighbours, UNEQ and several artificial neural network approaches) techniques used to differentiate wines. | 144 |
Various (45) | Wine | MS; ICP; L OES; ICP; L | Wines discriminated according to country of origin, regardless of the type of grape. Multivariate chemometric techniques used. | 145 |
Various | Wine | TXRF; -; S | Portable TXRF instrument used. The dry residue of wines resulted in high scattering and precisions of 4–28% RSD, however, detection limits of several tens of ppb were reported. 1 ppm Co used as internal standard. In Japanese. | 146 |
Various | Wine | OES; ICP; HG | Numerous analytes determined using ICP-OES and Cd, Hg and Pb determined using HG-ICP-OES. | 147 |
Various | Wine | MS; ICP; L | High resolution ICP-MS employed to resolve spectral interferences caused by the presence of nitric acid and ethanol. Semi-quantitative analysis also reported. Differences in accuracy were < 20% between the two modes of operation. Samples diluted 10-fold and In used as internal standard. | 148 |
Various | South African wines | MS; ICP; L | Multivariate statistics (PCA and LDA) used to classify wines geographically. 96% of wines and 100% of soils tested were classified correctly. | 149 |
Various | South African wines | MS; ICP; L | B isotope ratios determined by quadrupole ICP-MS and linear discriminant analysis for wine fingerprinting. Sr isotope ratios also determined but showed less potential for discrimination purposes. Other elements (> 20) were also determined. | 150 |
Organo-phosphorus compounds | Nerve agent degradation products | MS; ICP; HPLC | Speciation of organo-phosphates after removal of inorganic phosphorus (at levels of up to 10,000 μg mL−1) at pH 9 by coagulation with calcium chloride and ammonium hydroxide. | 151 |
The technique of LIBS also has significant capabilities and advantages in the analysis of artworks and frescos102and automotive paints.103 In the case of the fresco analysis, the measurement of certain line ratios of characteristic elements gave a good indicator of the pigments employed. Depth-profiling enabled the identification of pigments in the different layers. Non-parametric statistics were employed for the discrimination of automotive paint samples.
A nutritional-based approach to illness prevention is gaining widespread popularity and has resulted in a rapidly growing market for vitamin supplements and other pharmaceutical products. The technique of LIBS was employed by De Carvalho et al. for the determination of macro- and micro-nutrients in such supplements.104 This paper described the experimental set up of the LIBS spectrometer and in this case, the sample preparation of the tablets. Sample preparation was required because of the inhomogeneity of the tablets. However, total dissolution was not required, with homogenisation by cryogenic sample grinding and pelletising in a press being sufficient to ensure that the results were fit for purpose. The sample preparation procedure was described in detail in the paper. Signal to background ratios were optimised by varying the acquisition delay from 1 to 4 μs to allow the background to decay. Repeatability was found to be better when taking shots from across the pellet rather than multiple shots from the same site and issues affecting spectral variability were discussed. Results for Ca, Cu, Fe, Mg, Mn, P and Zn were compared with those obtained from the analysis of acid digests of the samples using ICP-OES. The precision for LIBS measurements was in the range 2 to 16%, with the ICP-OES analysis being generally 3% better than LIBS. In most cases the concentrations of the elements were in agreement. However, a small number of exceptions were noted and the reasons for this remain unknown at this time. The authors proposed that the LIBS procedure was sufficiently accurate and precise to be used for quality control of the elemental composition of tablets. They also promoted the ‘greenness’ of the technique as it required no acid digestion step. Work is on-going to improve the signal intensity and to improve the quantitative capabilities of LIBS. In a paper by Son and colleagues, the application of pulsed jets of argon buffer gas onto the sample area being analysed was shown to enhance the LIBS signal by up to 10 times.105 The synchronisation of the LIBS pulse with the buffer gas pulse enabled this system to be employed in the open atmosphere for a range of sample types. The spatial resolution capabilities of LIBS have enabled the quantitative determination of the migration of active pharmaceutical ingredients to the surface of tablets.106 This was investigated using model tablets containing nicardipine hydrochloride as the active pharmaceutical ingredient. When compared with FTIR mapping, LIBS proved to be quicker and more quantitative and was, therefore, an excellent technique for process analytical technology as applied to the active pharmaceutical ingredient migration problems in tablet formulations.
The elemental characterisation of cotton using LIBS has been shown to have important forensic and fraud detection applications because of its capability for rapid elemental fingerprinting of cotton by region, thus enabling the source of the cotton to be established.107 Visualisation of the regional differences based on elemental composition was achieved by using PCA. Linear discriminant analysis of the LIBS data resulted in the correct classification of greater than 97% of the samples by region in the United States and greater than 81% by State.
De Lucia and Gottfried have described the characterisation of organic molecules using LIBS, particularly for remote detection of N-containing molecules in explosives and explosive residues.108 A series of N-rich molecules including 5-aminotetrazolium nitrate, trinitrotoluene and others were studied and the atomic emission lines and intensity ratios of the constituent elements were shown to be correlated with the mole fractions and stoichiometries of the molecules. It was also shown that the amount of O present in the molecule influenced the emission intensities of molecular fragments such as C2. The chemometric tool of PCA was used to differentiate the molecules based on the emission intensities and ratios. Nitrogen-based explosive residues have also been detected in real time and under ambient conditions by one and two-laser photofragmentation fragment detection spectroscopy.109 The one-laser technique used a laser tuned to 226 nm to generate NO fragments and facilitated their detection using resonance enhanced multi-photon ionisation. In contrast, the two-laser technique employed a laser at 454 nm to generate the gas-phase analyte by matrix-assisted desorption and a second laser operating at 226 nm to generate NO photofragments and to ionise them. The effects of laser energy, analyte concentration and matrix on the signal were reported and the rotational temperatures of the NO photofragments calculated. A positive signal identification on a sample mass as low as 70 pg was achieved. This demonstrated the efficacy of remote laser-based techniques for real-time analysis of explosive residues. A novel, high energy explosive material, FOX-7 (1,1-diamino-2,2-dinitroethylene) was studied by Civis and co-workers, who used both LIBS and selected ion flow tube mass spectrometry (SIFT-MS) for the analysis.110 In this work, LIBS was employed as the energy source to convert small quantities of sample into plasma and to yield a LIBS spectrum of its fragments and atoms. The technique of SIFT-MS is based on chemical ionisation using reagent ions. It can be used to determine the concentrations of trace gases evolved from volatile organic compounds. In this study, SIFT-MS was used to quantify the release of nitrogen oxides and a range of organic fragments after LA of FOX-7. Work designed to improve the discrimination ability of LIBS analysis of explosive residues was undertaken by De Lucia and Gottfried.111 Thirteen organic materials, which included novel high N-containing compounds, conventional military explosives and non-explosive materials were analysed and the data examined using chemometric techniques. The best model for discrimination of these similar materials was partial least squares discriminant analysis (PLS-DA). A better than 95% success rate in the correct classification of 13 types of compound was demonstrated. The use of the whole spectra to build the model gave the best results, but variability caused by the surrounding atmosphere and the substrate was an issue, suggesting that further work was required to improve robustness. This problem was investigated by Lucena et al.112 Here, selected atomic emission lines from atomic C, H, N and O as well as molecular lines from CN and C2 were investigated in plasma generated in both air and helium. This enabled a better understanding of the pathways leading to molecular emissions and hence some useful tools for the discrimination of explosives.
The ability to determine N and P concentrations rapidly is critical for quality control in the fertiliser industry and could ultimately lead to improved acceptance and utilisation of waste composting. The use of LIBS with partial least squares (PLS) calibration for the fast and accurate multi-element analysis of fertiliser samples which had been compressed into pellets has been reported by Yao and colleagues.113 The measurement of K and P under atmospheric conditions was discussed, with the results for N to be reported in a future paper. There was a strong dependence of K and P line intensity on the presence of other elements in the matrix, hence the need for multivariate calibration. The accuracy of the analysis was investigated using test samples and, for a phosphorus pentoxide fertiliser reference material of 13.26 wt %, the absolute error was 0.31 wt %. Similarly, for potassium oxide the reference value was 12 wt % and the absolute error was 0.63 wt %. This should be a valuable technique for ensuring the quality of fertilisers rapidly in the field and this, in turn, could enable the growth in the use of fertilisers based on waste compost. This would be particularly important in developing economies.
Laser ablation as a means of sampling challenging materials for ICP-TOF-MS analysis has been employed by Szynkowska and co-workers for the forensic examination of photocopy and printer toners.114 The isotopic analysis of both black and coloured toners enabled the identification and discrimination of materials from various manufacturers. Both cluster analysis (CA) and PCA were employed to examine the analytical data from 201 samples of black toners and 23 samples of coloured toners. Laser ablation can be a useful technique for the analysis of small samples and in an example by Ma et al. it was employed with ICP-MS detection for the analysis and classification of ballpoint inks during the forensic examination of documents (in Chinese).115 This work enabled blue ink from 95 different ballpoint pens to be classified into 34 distinct groups according to their metal content, amongst which 26 groups were distinguishable by their element response ratios. The paper on which the ink was written was found to have no impact on the analysis. The authors found that of the 95 kinds of ink examined, 88 could be distinguished using this procedure, with good precision and little damage to the samples.
Powdered samples can pose significant analytical challenges. In a novel approach, a transversely excited atmosphere carbon dioxide laser induced helium plasma was employed for the rapid semi-quantitative analysis of very small amounts of powder.118 Sample with mass as low as 4 mg was immobilised on a target by mixing it with silicon grease. The transversely excited atmospheric carbon dioxide laser was directly focused onto the sample in a helium atmosphere, resulting in transfer of energy from metastable helium atoms to the sample vapour. The analytes Cr, Cu and Pb were quantified in soil with detection limits in the low ppm range being achieved.
There is currently much interest in ionic liquids in ‘green chemistry’ as they have very low vapour pressures and therefore eliminate emissions of volatile organic compounds to the atmosphere. Additionally, they have novel solvating properties and have many novel uses. Their ultra-low vapour pressure makes them amenable for study using high vacuum surface analysis techniques, including TOF-SIMS. An investigation by Holzweber et al. into the analysis of five different ionic liquids using TOF-SIMS was therefore timely.120 A range of ionic liquids was analysed under Bi-ion and Bi-cluster ion bombardment. Some very nice mass spectra of these compounds were presented, together with proposed general rules for understanding their fragmentation. Strong evidence for H-bonding was found, backing up previous studies and an estimation of their hydrogen-bond strength was shown.
Silicon analysis can present particular challenges and speciation of this element is not well developed. An excellent and very useful review (with 203 refs) of the analysis of total Si and the speciation of Si by hyphenated techniques for environmental, biological and industrial issues, was presented by Chainet and co-workers.123 Of particular note was a very useful section on contamination and analytical artefacts, which present major challenges in the determination of Si. Different Si species including siloxanes, silanes, silanols and silanediols were all discussed.
Speciation of dietary As is becoming increasingly important because of the very different toxicities of the different As species. Slurry sampling with hydride generation (HG) and detection using AAS has been reported by Sun et al. for the speciation of As in dietary supplements.124 The paper gave the slurry preparation procedure and described the HG method which allowed the speciation of As(III) and total As. Spiking studies indicated excellent recoveries. Arsenic speciation in Chinese medicines is also increasingly recognised as being important for the same reasons. In a paper by Jin and colleagues, As(III) and As(V) were determined using HPLC-ICP-MS (in Chinese).125 The species monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) were also determined, but were not found to be present in any of the medicines.
The Hg-based preservative, thiomersal is added to vaccine preparations, including the Hepatitis B vaccine, to prevent bacterial growth. This presents a consideration in the field of Hg toxicity and the vaccine preparations are therefore regulated for Hg concentration. This means that there is a need for a reliable procedure for the determination of Hg in vaccine preparations. Several atomic spectrometric-based analytical methods are available for Hg determination which, under many circumstances, can be less than straightforward. Several procedures including CV-AAS and ETAAS are commonly employed. The most commonly used method is CV-AAS and this requires a sample pre-treatment step to generate elemental Hg. An electrolyte cathode discharge atomic emission spectroscopy (ELCAD-AES) technique for the determination of Hg in solution has been reported by Shekhar and colleagues.128 Although there are currently no commercially available instruments, the authors claimed several significant advantages, including low power consumption, ease of portability for field analysis and sufficiently low LOD for many metals in a range of matrices including tap water, milk and potentially more complex matrices. This paper showed the application of ELCAD-AES as a simple and rapid technique for the direct quantitation of Hg in the Hepatitis vaccine, thus eliminating the need for the formation of elemental Hg. The instrumentation was described in detail and a paper describing the construction of the ELCAD cell was referenced. The optimum experimental conditions were given and the analytical figures of merit described, with linearity over the range 50–50000 ng mL−1 being demonstrated. The 3σ detection limit (n = 10) was 15 ng mL−1 inorganic Hg which was better than that obtained using ICP-OES. Matrix effects were expected to be higher in ELCAD techniques than in ICP-OES because of the lower plasma temperature of around 5000 K. The results were compared with the values for a range of certified references materials and were accurate to within 2.1% relative or better, of the certified value of tuna fish. It was found, however, that the organic Hg present in the thiomersal gave a significantly lower response, possibly because there was insufficient energy in the plasma to break the bonds. It was noted that, because of this, standard addition procedures involving addition of inorganic Hg would be misleading and would give erroneous results. The standard addition of thiomersal was therefore recommended. The detection limit for Hg by direct analysis in the vaccine was 25 ng mL−1 and the accuracy, when compared with ETAAS and CVAAS was within 1.6%–8.7% relative. For the analysis of certain materials, direct analysis may be more expedient than preconcentration or extraction procedures. Photochemical vapour generation of Hg-containing compounds coupled with ICP-OES detection has been reported by dos Santos et al.129 They developed a procedure for the determination of thiomersal in human and veterinary vaccines, also with no sample preparation other than dilution and the addition of formic acid, as a reductant. In this procedure the LOD (0.3 μg Hg L−1) were lower than those quoted in the preceding paper, and recoveries were 93–102%.
Ochratoxins are mycotoxins which are common contaminants of foods and beverages including wine, resulting from fungal contamination. Tests have indicated that they are carcinogenic in rats and mice and consequently EC directives for concentrations of these toxins have been established. Currently, ochratoxins are determined using HPLC-MS and this requires complex extraction and sample clean up prior to analysis. Furthermore, stable isotope-labelled internal standards are required for accurate quantitation. Immunoassays are well established for this kind of analysis, but suffer from limited dynamic range and precision issues for very high or very low concentrations of analyte. An immunoassay for ochratoxins involving labelling with secondary antibodies conjugated with either Au nanoparticles, for detection using ICP-MS, or horseradish peroxidase for detection using photometry has been developed by Giesen and colleagues.130 This work demonstrated that both detection methods enabled quantitation of the toxin over the range 0.01–1 μg L−1, and possessed similar analytical figures of merit. The detection limit was capable of meeting the EC limit of 2 μg L−1. The authors suggested that sample throughput could possibly be increased by employing an assay array with detection using LA-ICP-MS and that detection using ICP-MS could be advantageous where several analytes needed to be determined in a single assay.
Microwave-induced combustion is a useful means of sample preparation that is potentially applicable to a range of matrices. It was described in detail as a means of sample preparation for pharmaceutical samples with ICP-MS analysis in a paper by Nam and co-workers.131 The technique was demonstrated to be a reliable and broadly applicable procedure for trace- and ultra-trace analysis of various analytes in pharmaceutical samples. In the microwave-induced combustion procedure outlined, the crushed sample was added in an ashless filter paper to a quartz basket suspended in the pressure vessel. The paper was soaked with a 1.5 M ammonium nitrate solution which was the igniter. The vessel was then pressurised with oxygen to 20 bar. The absorbing solution was 5 mL of 20% nitric acid. Good quantitative recoveries were achieved with detection limits below 0.01 μg g−1.
Slurry nebulisation as a means of sample introduction and ICP-MS detection has been employed for the direct determination of trace B in high purity graphite powders used in the nuclear industry (in Chinese).132 This procedure required the accurate control of the particle size in the ground sample followed by the addition of polyvinyl pyrrolidone as a dispersant. The analysis was run with a higher mass resolution (0.6 amu) to minimise interference from the carbon matrix and oxygen was added to the plasma to prevent carbon deposition on the cones. Calibration was by aqueous external standards and spiked samples of graphite showed recoveries in the range 97–103%. Slurry sampling ICP-OES has been employed, in a paper by Sousa and colleagues, for the determination of the elemental composition of aqueous suspensions of solid sweeteners.133 Recovery studies showed that the accuracy was between 90–110% with precision better than 5% RSD. Robust plasma conditions should enable slurry nebulisation to be a reliable option for certain analyses, minimising the need for sample digestion or ashing. However, care had to be taken to ensure matrix matching was employed as viscosity effects and ionisation interferences could be significant.
The analysis of trace elements in petroleum products and in refinery streams and crude oils is becoming increasingly important because of the requirement to meet increasingly stringent environmental standards. There is, therefore, the need to quantify elements in these materials at trace levels and ICP-OES or ICP-MS are widely used. However, the introduction of organic samples into the plasma is challenging and often results in difficulties. A possible solution to these problems was to minimise the amount of sample entering the plasma by using air-segmented flow-injection analysis.74 Air segmentation offered several advantages including minimisation of sample dispersion and enhanced solvent evaporation which led to high sensitivity and enhanced energy transfer from the plasma to the aerosol. This paper presented a comprehensive study of the effects of the solvent nature on the sample introduction performance and an evaluation of the extent of the interferences caused by a range of organic solvents. A range of parameters was investigated, including spray chamber temperature, spray chamber design and solvent type. A heated single pass spray chamber in air segmented mode conferred several advantages, including a dramatically decreased sample consumption, increased sensitivity, shorter wash out times and thus improved sample throughput and the elimination of effects caused by changing the solvent composition. The down side was the need for more rapid data acquisition because of the generation of transient signals.
Commercially available instruments for high-resolution continuum source atomic absorption spectrometry (HR-CS-AAS) have been a relatively recent instrumental advance and this reviewer believes that it offers many of the advantages of ICP-OES with the running costs of FAAS. It is also equally applicable for the measurement of molecular absorption bands with high resolution and simultaneous background correction, and this has been demonstrated during the determination of total F in toothpaste using graphite furnace as a means of analysis.134 Transversely heated tubes had zirconium permanent modifier added. Then, before each sample was introduced, palladium plus gallium solutions were introduced and the tube heated. After thermal pre-conditioning, a sample plus more of the gallium solution was introduced and a standard temperature programme was run. During the traditional “atomisation stage”, molecular absorbance from GaF was measured. The LOD (3 σ of 11 blanks) and the LOQ (10 σ) were 0.26 μg L−1 and 0.87 μg L−1, respectively. The detection limit corresponded to around 30 ppm F in the toothpaste. The results found were in good agreement with the manufacturer's stated content and those obtained by GC-MS methodology but, as less sample preparation and no derivatisation was required, the new technique was significantly quicker. Ionic F, however, must be determined by ion selective electrode procedures.
An overview of gemstone analysis during the first decade of the 21st century was made by Breeding and colleagues.153 The advances in gem treatment and synthesis technology, the discovery of new gem sources etc. has led to the requirement of more rapid as well as more accurate analysis. The authors discussed the developments in analytical technology over the decade and noted the rise in popularity of micro-sampling techniques such as LA-ICP-MS and LIBS as well as techniques such as photoluminescence. Also noted was the increased use of assorted X-ray techniques including real-time fluorescence, X-ray mapping and portable instrumentation. The overview contained 82 references.
Santamaria-Fernandez overviewed C isotope ratio measurements using MC-ICP-MS and questioned if it could become a viable alternative to isotope ratio mass spectrometry (IRMS) for rapid analysis of compounds and the characterisation and certification of isotopic reference materials.154 The advantages and disadvantages of the technique were discussed and suggestions of how precision and sensitivity could be improved were made. The overview contained 22 references.
Isotopic reference materials available for research purposes have been reviewed by Vogl and Pritzkow.155 Although not exhaustive (the review did not discuss the analytes C, H, N, O and S or radioactive elements), it did provide a useful list of the isotopic reference materials available. A useful discussion of the limitations of the materials (e.g. homogeneity and uncertainty issues) and of the synthetic isotope mixtures used to characterise them was also presented. The authors concluded that isotopic reference materials should be provided with additional data on isotopic abundancies as well as the delta scale for each element of interest. The review contained 70 references.
A paper by the same research group described the analysis of cements using LIBS.157 After pressing the powdered cement into a pellet, the authors focussed a laser onto the surface and formed a plasma from which the analytes Al, Ca, K, Mg, Mn, Na, Si, Sr and Ti were determined quantitatively. Calibration was achieved by using several standard cement samples. In an attempt to demonstrate that the plasma operated under local thermodynamic equilibrium (LTE) conditions, the plasma electron temperature and the electron density were calculated. Several other research groups have also used LIBS for the successful analysis of cements or concretes. An example, by Sugiyama et al. detected Cl at a concentration of 0.18 kg per cubic metre.158 The instrumental setup included a Nd:YAG laser that operated at 532 nm for the formation of the plasma, a fibre optic bundle to collect the light emitted and a spectrometer with an intensified CCD camera to segregate and then to detect the light emitted by the Cl (at a wavelength of 837.59 nm). When used in double pulse mode, two Nd:YAG lasers were used. Parameters such as the delay between the laser irradiation and spectral measurement, the gate and width of spectral measurement and, when in double pulse mode, the delay between the two pulses, were all optimised to ensure maximal sensitivity. A linear relationship between Cl concentration and signal was observed when powdered samples with Cl content in the range 0.18 to 5.4 kg m−3 were pressed into pellets and used as calibrants. The technique was easily capable of determining Cl at a concentration of 0.6 kg m−3, which is the concentration at which reinforcing bars start to corrode. The authors therefore concluded that the procedure they had developed was fit for purpose. Another application of LIBS, this time used in conjunction with XRF, was reported by Colao and co-workers who analysed historical building materials.159 A total of 48 samples were analysed. One sub-group was used as a reference set because their provenance was known. The majority, however, were treated as “unknowns”. Chemometric analysis (PCA, Linear Discriminant Analysis (LDA) and Soft Independent Modelling of Class Analogy (SIMCA)) of the analytical data obtained clear separation of the eight reference samples. The results of the “unknown” samples were compared with assignments based on petrographical description. It was concluded that LIBS determinations of a relatively few elements enabled the rapid and accurate identification of provenance. Another application reported the analysis of archaeological mortars.160 The authors, Miriello and colleagues, used XRF, LA-ICP-MS and other techniques such as SEM-EDS and X-ray powder diffraction (XRPD) to analyse samples from Pompeii. Again, chemometric analysis of the data (Cluster Analysis) was used and managed to identify three main construction phases.
A paper by Riba et al. related the merits of multivariate calibration techniques during the cement making process.161 Knowledge of the Fe2O3 content is necessary since it is one of the governing factors for the quality of the cement. Using Classical Least Squares analysis of the data obtained using WDXRF and reflection colorimetry, they managed to identify the samples' properties in a rapid, cheap and non-destructive manner; which is an attractive proposition.
Speciation analysis of leachate from cement-based materials was reported by Mulugeta and co-workers.162 Using an anion exchange column, the workers used HPLC-ICP-MS to separate and then detect As(III), As(V), Cr(VI), Mo(VI), Sb(III), Sb(V), Se(IV), Se(VI) and V(V) successfully. A two step multivariate approach to the optimisation process was undertaken in which the effect of sample pH and the mobile phase composition on the peak resolution, peak symmetry and analysis time were the determining factors. Optimal conditions consisted of a mobile phase of 20 mM ammonium nitrate and 50 mM ammonium tartrate at pH 9.5. This was found to elute the analytes in an isocratic manner in less than six minutes. Addition of 1% methanol to the mobile phase enhanced the sensitivity of the detection system. Under optimal conditions, LOD ranged from 0.2 to 2.2 μg L−1. Most sample leachates were found to contain Cr(VI), Mo(VI) and V(V).
A comparison between different permanent modifiers for the determination of Cd and Pb in slurries of sediments and soils using ETAAS was reported by Dobrowolski and colleagues.164 Mixtures of niobium/iridium and tungsten/iridium were studied in detail. The optimal amounts for the Cd determination were 200 μg of niobium and 5 μg of iridium whereas for the Pb determination they were 15 μg of niobium and 200 μg of iridium. The determination of Cd also required the use of the modifier ammonium dihydrogen phosphate to minimise interferences. The slurries were prepared in 5% nitric acid. The optimised procedures were validated by analysing CRMs. Both precision and accuracy were described as “acceptable”. A study of permanent modifiers has also been made by Cervenka and co-workers who compared the electrochemical modification of the graphite platform using two different methods: a cell with 20 mL of solution and a single drop of modifier.165 The permanent modifiers tested were palladium and a mixture of iridium and gold. Using reproducibility as the major criterion, the electrolysis of the drop was the better option. The electrochemically modified platforms were then used for the analysis of solid environmental samples weighed directly onto them (in which 0.1–10 mg of sample was used). The analyte of interest was Hg. Tube lifetime was only 60–70 firings for the deposited gold/iridium modifier but was somewhat improved (100–120 firings) for the palladium modified platforms. Calibration was against aqueous standards stabilised with potassium permanganate. Results obtained were compared with those obtained using thermally deposited modifiers and with those using pyrolysis AAS with gold amalgamation.
Most matrix modifiers work by thermally stabilising an analyte so that higher pyrolysis temperatures may be employed therefore maximising the removal of potential interferences. Some, however, work in the opposite manner to this and actually increase the volatility of the analyte so that it becomes atomised at a lower temperature than potential interferences. There have been two examples of these volatilisation aids in this review period. Chen et al. discovered that 1-phenyl-3-methyl-4-benzoyl-5-pyrazone forms a thermally stable but volatile chelate with Cr(III) but not with Cr(VI).166 They therefore used this to achieve speciation analysis for Cr in water samples using ETV-ICP-MS as the detection method. The Cr(III) was volatilised to detection at only 1000 °C leaving the Cr(VI) in the ETV device. This could then be volatilised at a higher temperature and so the overall result was two distinct Cr signals. The LOD for the Cr(III) was 0.031 ng mL−1 with precision at the 1 ng mL−1 level being 5.3% RSD (n = 9). The second study of this type was by Huang and Jiang, who used 8-hydroxyquinoline-5-sulfonic acid as a volatilisation aid for the ETV-ICP-MS determination of Cd, Co, Cr, Cu, Hg, Ni, Pb, Se, V and Zn in slurried samples of cereals.167 After optimisation to ensure similar sensitivity between slurries and aqueous calibrants, the authors used the ICP-MS instrument in dynamic reaction cell mode with the introduction of 1 mL min−1 methane for the determination of the Cr Cu, Se and Zn. The other analytes were determined separately in normal mode. Several CRMs were used for method validation and a comparison with data obtained using pneumatic nebulisation was also made. Detection limits between 0.6 and 16 ng g−1 were obtained.
Titanium dioxide white powders are used regularly to develop latent fingerprints on dark surfaces. However, it has been noted that different commercially available products have different effectiveness even though nominally, they have the same composition; the main problem being that they adhere to the background as well as the fingerprint. Jones and colleagues have therefore used XRF, XPS, TEM and laser particle sizing in an attempt to elucidate why this should be.170 The results indicated that the titanium dioxide particles contained a coating tens of nm thick that consisted of Al and Si rich material. Traces of Na and S were also observed. It was differences in the morphology, thickness coverage and composition of these coatings that governed the success of the powders.
The use of LIBS as an on-line means of analysis is also still gaining popularity. This is because of its ease of use and standoff abilities. The use of LIBS as an on-line method for the determination of elements (K, Mg and Na) in potassium fertilisers was described by Groisman and Gaft.172 Previously, this work was undertaken by Natural Radioactivity Analysers, which just determined 40K. The use of LIBS to determine all K (rather than one specific isotope) and other trace analytes, was therefore a large step forward. The results from the on-line system were compared with those obtained using another chemical analysis technique and were found to correlate well. Samples from Russia, Belarus and Israel were used in the study. Two studies have reported the use of LIBS for the on-line analysis of gases. In one example, Eseller et al. monitored impurities in hydrogen using a system comprising a frequency doubled Nd:YAG laser to generate a spark in the gas flow and then a miniature spectrometer that covered the wavelength range 620–800 nm and a gated detection system to detect the radiation.173 The effects of pressure on the determination of the contaminants Ar, N and O as well as on the signal of H were observed. Calibration was obtained by studying spectra with different contamination levels of the analytes. Both gated and ungated systems were tested and LOD from each calculated. Although both systems yielded reliable and reproducible results, the ungated LOD were a factor of approximately four higher than those obtained using the gated system. The second paper of this type used LIBS to detect components of air quantitatively using an ungated system and a polarisation filter in the signal collection system to suppress the elastically scattered light.174 This polarisation filter was necessary to ensure that the linear dynamic range of the technique was not compromised. On-line background correction was achieved by running the detector with a doubled repetition rate compared with the laser. Elemental O and H were determined in samples of synthetic, ambient and exhaled air, with LOD being 15 and 10 ppm for O and H respectively.
The on-line analysis of fine and ultra-fine particulate matter using LIBS has been reported by Strauss and colleagues.175 The purpose of the research was to determine the contaminants present in emissions from industrial processes and then also to determine the size-dependent composition. The investigation had several streams, including the optimisation of the laser plasma and the analysis of mono-disperse and poly-disperse aerosols. The system could be calibrated with regard to relative or total elemental concentrations. Fundamental plasma parameters such as electron density and temperature were determined by temporal aspiration of calcium chloride particles. The size-resolved analysis of particles with size range 20–800 nm was performed and the signal response was found to correlate with the mass distribution of the particles.
A paper by Knapek and co-workers discussed the use of LA-ICP-OES as a means of determining Cd, Cr and Cu in high salt samples after electro-deposition of the materials on electrodes.176 Three electrode types were tested and nickel was found to be the best. Other parameters, such as pH, deposition current and time of electrolysis, were all optimised. After deposition of the analytes as a film on the nickel electrode, they were ablated using a Nd:YAG laser and swept to detection. Two different systems were compared, one with a rotating electrode and the other with a static bottom working electrode. Mixed results were obtained, with the latter yielding better LOD for Cr and Cu whereas the former had better LOD for Cd. Precision was typically in the range 3.8 to 10.3% RSD, depending on the analyte and on the concentration. The procedure was applied to the analysis of waste waters.
Ultrasonic slurry sampling is a sample introduction technique that is still being employed by some workers. One of the latest applications has been reported by Hsiao et al. who used it to introduce slurries of 99.999% pure silicon powders and solar grade silicon powders into an ETV-ICP-MS detection system.177 Using citric acid as a matrix modifier, the authors managed to vaporise the analytes Al, B, Cd, Cu, Fe, Ni, P, Pb and Zn to the ICP-MS instrument for detection at the relatively low temperature of 1900 °C. This had the advantage of separating the analytes from the silicon matrix which, being very refractory, remained in the ETV device. This had the overall effect of decreasing polyatomic interferences arising from 28Si29Si+, 28Si16O2+, 28Si17O18O+ and 38Ar28Si+ on 57Fe+, 60Ni+, 63Cu+ and 66Zn+ respectively. Unfortunately, calibration had to be achieved using standard additions which, inevitably, increased the analysis time. The results obtained were in good agreement with those obtained using a sample digestion protocol and pneumatic nebulisation/membrane desolvation. Precision was better than 9% RSD and LOD ranged from 0.4 ng g−1 (Pb) to 50 ng g−1 (Fe).
Three different methods were compared by Figi and colleagues for the determination of assorted analytes in brake pads.178 Two of the methods involved the extraction of the analytes using either a high pressure asher or a microwave extraction technique followed by ICP-OES detection. The other method tested used a hand-held XRF instrument that was capable of determining the analytes without any further sample manipulation. In general, the results from the different techniques were comparable, although the authors did note that only some of the analytes (Cd, Co, Cr, Mn, Mo, Ni, Pb and Sb) could be determined quantitatively using the XRF technique. Other analytes (Bi, Cu, Sn, Sr, V and Zn) could only be determined qualitatively. This was attributed to the matrix characteristics of the sample type. Despite this, it was concluded that the XRF method was an efficient tool for the determination since it could be achieved in situ, hence saving time and cost.
A review (containing 111 references) of gold nanoparticle analysis was prepared by Yu and Andriola.180 The authors summarised the problems associated with the particulates, i.e. what their fate is in the environment and how many particles exist in a dispersion, and then went on to review the methods used to answer the two questions. The methods reviewed included mass spectroscopy, electroanalytical methods, spectroscopic methods and particle counting methods. A recent overview of ICP-MS that discussed the recent trends and developments was presented by Engelhard.181 The overview contained only 54 references but, among the numerous topics, it did discuss some future trends of which analysis in the area of nanosciences and of microparticulates and single cells were included. Other topics touched on included the novel mass spectrometers (e.g. Mattauch-Herzog configuration), focal plane camera array detectors, and methods of detecting fast transient signals, i.e. those obtained using LA or chromatography; etc. The last paper of relevance to this part of the review was prepared by Dubascoux and colleagues, and focused on the coupling of field flow fractionation (FFF) with ICP-MS.182 The authors discussed the assorted merits of the different types of FFF and then gave the theory and many of the practical aspects associated with the coupling. Of particular note, the different components of the coupling were described, with easy reference tables to aid the reader. Operating conditions such as carrier composition, flow rate and nebuliser types that can be used were all discussed; as were the prospects of both qualitative and quantitative analysis. Applications from fields as diverse as the environment, bioanalysis and nanoparticle analysis were all presented. The review contained 64 references.
Other techniques have also been applied for the characterisation of nano- and micro-particulates. Garcia and co-workers characterised single Au and silica particles using ICP-OES by using monodisperse droplets of standard solutions for calibration.185 Commercially available silica (diameters between 470 nm and 2.06 μm) as well as 250 nm Au nanoparticles were obtained, with both material types having narrow mean particle size distribution. By introducing monodisperse droplets of Si and Au standard solutions, the authors demonstrated that ICP-OES signals showed good agreement between the atomic line intensities recorded for particles and droplets from standards that had the same analyte mass. The LOD for ICP-OES measurements equated to particles with diameter of approximately 200 nm and 470 nm for Au and silica nanoparticles respectively; which corresponded to masses of 80 fg for Au and 50 fg for Si. A second paper has also analysed mono-disperse gold nanoparticles, this time using X-ray standing waves (XSW) under grazing incidence geometry as well as SEM-EDX.186 Samples with mean diameter 25 nm, 46 nm, 73 nm, 100 nm, 115 nm and 250 nm were used during the experiment and were prepared on silicon wafer pieces. The smaller diameter particles were able to be analysed using XSW, but the larger ones were too large for the radiation beam produced by the instrument to cover them completely. Experimental results were compared with those produced by simulations.
Nanomaterials may be used in a very wide variety of applications and nanowires, in particular, have very great potential as inter-connectors in nanoelectronic, magnetic, chemical or biological sensors, etc. This point was emphasised by Bustelo and colleagues, who prepared self-ordered arrays of metallic nanowires based on the use of nanoporous anodic alumina and self-assembled nanotubular titanium dioxide membranes as templates.187 Inevitably, the optimisation of synthesis and the quality control of the product will depend on the successful characterisation of the final product. The authors described the use of pulsed radiofrequency (RF)-GD-TOF-MS for depth-profile analysis of two types of materials. One type was self-assembled nanoporous anodic alumina templates filled with arrays of single metallic Ni nanowires as well as arrays of multi-layered Au/FeNi/Au and Au/Ni nanowires; whereas the other was nanotubular titanium dioxide templates filled with Ni nanowires. The RF-GD-TOF-MS analysis enabled a rapid and reliable depth-profile analysis to be obtained which also yielded data on the contamination introduced during the synthesis. In addition, by measuring the elemental to molecular ratio for some species (e.g.27Al+/16O+ and 27Al+/32O2+), it was possible to identify whether or not there had been any leaks in the system and then to rectify it if necessary.
Sample introduction to ICP-based instrumentation using slurries offers many advantages over traditional digestion procedures; namely speed of preparation, decreased sample manipulation and hence decreased contamination levels, no necessity for dangerous reagents (such as hydrofluoric acid) and the possibility of calibration using aqueous standards. However, the technique does have some drawbacks. One of the problems associated with analysing nanoparticles as a slurry is that they can agglomerate forming a much larger assembly of particles. These much larger particles cannot be introduced to a standard ICP-OES or ICP-MS instrument because the nebuliser/spray chamber sample introduction system will discriminate against them; i.e. they will be passed to the drain. Gross underestimates of the trace metal content may, therefore, be obtained. Some workers attempted to overcome the agglomeration problems by the use of dispersants and, over the years, these have shown varying degrees of success. Sousa and co-workers have introduced slurried samples of ferrite nanoparticles as well as nickel and cobalt nanoferrites to an ICP-OES instrument.188 Considerable optimisation of pH was required to ensure that the particles were dispersed and to ensure that the signal produced from the slurry was equivalent to that produced from the same concentration of an aqueous calibration solution. A paper that attempted to address the problem of particle size discrimination during the analysis was published by Duester and colleagues.189 These authors developed a particle size independent method for quantifying metal(loid) oxide nanoparticles and their agglomerates that used the analytical technique of ETV-ICP-MS. The protocol developed was applied to the analysis of nanoparticles in assorted matrices which represented different degrees of complexity.
Metallic catalysts are used during the preparation of both single- and multi-walled carbon nanotubes. These metallic catalysts usually leave a residue in the tube and so must be extracted quantitatively if ultra-high purity tubes are to be produced. Several extraction and analysis techniques for these materials have been examined in a paper by Yang et al.190 The impurities were determined using either ICP-OES or ICP-MS after three different pre-treatment protocols had been used to extract the analytes. These pre-treatments were: water extraction, 1% nitric acid extraction and microwave acid extraction. Results for the total concentrations of analytes in the tubes were obtained using INAA. The results demonstrated that the analytes were very poorly extracted into either water or dilute acid, indicating that they would have very low bioavailability and hence toxicity in the environment. Furthermore, it also raised the possibility that the metallic residues may be used to trace the nanotubes during biological or toxicological studies. Strangely, the results obtained for the ICP-OES analysis of the microwave extracts were in much closer agreement with the INAA results than those obtained using ICP-MS. The authors attributed this to a non-spectral interference induced by carbon residues in the latter technique.
A topic related to nuclear forensics is nuclear safeguards, i.e. systems designed to monitor compliance with the nuclear non-proliferation treaty. The safeguards also require high precision and high accuracy measurements to ensure reliability. It is necessary to be able to measure plutonium at the fg level while measuring the full suite of uranium isotopes (233U–238U) where the total uranium content may be at the ng level. Hotchkis and co-workers presented a paper that described recent improvements in achieving these measurements using AMS as a means of detection.201 A fast isotope cycling system for actinides was commissioned and this led to improved precision, with reproducibility of 4% for actinide isotope ratios. The background level for the key rare isotope 236U was found to be 8.8 fg, for total uranium content in the ng range. The background was limited by 236U contamination rather than ion misidentification. For plutonium the background was at the low fg level. As stated previously, nuclear safeguards must be accurate and precise. There is, therefore, a need to produce reference materials to ensure that results are appropriate. The production and analysis of such materials have been described in a paper by Ranebo et al. who used a vibrating orifice aerosol generator in connection with a furnace system to produce micrometer sized (approximately 1.5 μm), mono-dispersed particles from reference U and Pu materials in solution.202 The exact size and weight of the materials could be controlled by the conditions of the aerosol generator. The particles were then characterised using SIMS. Kraiem and colleagues developed a rapid sample preparation and analysis protocol that used TIMS to investigate U isotopic signatures in real-life particles from “swipes” from a nuclear facility.203 This work also described the use of a carburised filament and multiple ion counting TIMS to increase sensitivity. Environmental dust was collected in different locations within a nuclear facility. The samples were analysed using SEM-EDX to find the U particles of interest. Comparison of the measurement results with reference data evaluated by international safeguards authorities was of key importance for data interpretation. For the majority of investigated particles, detection of U isotopic signatures provided information on current and past nuclear feed operations that compared well with facility declarations. A second paper also analysed swipe samples.204 For the sample preparation a low power microwave-assisted digestion followed by extraction chromatography using TRU (R) resin was applied. The concentration and the isotopic composition of the analytes of interest were determined by sector field ICP-MS. In terms of the accuracy, precision and repeatability of the measurement of 234U, 235U, 236U, 238U, 239Pu, 240Pu and 241Pu, the protocol was found to be adequate. Al-Nuzal reported a method of determining 235U/238U in nuclear spent fuel that was described as being direct, simple and fast.205 Using TIMS as a means of analyte detection, the author improved the quality of the data by placing both the sample and a carefully optimised aliquot of refractory metal oxide solution on the sample filament. Several materials were investigated as the refractory oxide, e.g. magnesium, cerium, thorium and chromium, and it was the magnesium that was found to work the best. The method was described as being very simple, it improved both the accuracy and precision of the collected data, reduced the time required to achieve a steady U signal and hence, the overall time of the analysis. In addition, the level of impurities present was irrelevant. A “guide to the expression of uncertainty in measurement” using the multicollector TIMS determination of U isotopes was presented by Burger and colleagues.206 The advantages of the approach were reputed improvements in the coherency and transparency of the uncertainty calculation, which included contributions from all potentially significant sources of uncertainty to the mass spectrometric measurement result. It identified the dominant causes of the measurement uncertainty and enabled a better understanding, management, and improvement of the measurement process. The paper gave detailed examples of the calculations for most of the common types of TIMS analysis, and included various potential sources of error, including the baseline noise, peak tailing, peak flatness, detector inter-calibration and detector linearity response.
Other applications have reported the analysis of components of a reactor system. A simple application by Smirnova described the determination of the elemental composition of deposits on jackets of fuel elements from zirconium alloys.209 The deposits were removed by using a chemical attack (that had been optimised in terms of temperature and reagent) and then analysed using ICP-OES. Inconel is a nickel-containing superalloy that is a candidate material for the manufacture of high temperature components (the heat exchanger) of a gas-cooled fast reactor. Duval and colleagues reported the oxidation mechanisms at 850 °C using a range of oxygen partial pressures.210 The material was tested over a period of 1–28 days and the oxide layers characterised using a variety of techniques including XPS, GD-OES, XRD and EDX. A LIBS-based system was employed by Mercadier et al. for the analysis of carbon fibre tiles from a fusion reactor.211 An IR laser pulse of 4 ns duration was used to excite H and D within the sample under a reduced pressure argon atmosphere. By measuring the Balmer series of wavelengths for both analytes, the authors found that the H expanded into the argon at a slightly faster rate than the D. This segregation was, however, pressure dependent.
Some papers reported the analysis of nuclear fuels themselves. Satyanarayana and Durani reported the determination of trace impurities in nuclear grade uranium oxide.212 The technique, which involved the dissolution of the sample in nitric acid (6 M), followed by removal of the uranium matrix using di (2-ethyl hexyl) phosphoric acid in toluene and then ICP-OES analysis, was described as being rapid and simple. The method was applied to several synthetic samples and five CRMs. The procedure was optimised in terms of contact time, acidity, aqueous to organic ratio, etc. and then compared with a number of alternative matrix separation protocols. A large number of analytes including the REE and 13 others, were determined with precision values reportedly in the range 1.5–12% for a concentration of 0.2 μg L−1 improving to 1.5–6% for 0.5 μg mL−1. Another example was published by Alamelu and co-workers who determined 16 analytes in nuclear grade thoria by using LA-ICP-MS with an argon fluoride laser using the operating conditions of 20 ns pulse duration, a wavelength of 193 nm at a repetition rate of 20 Hz.213 Method validation was achieved by analysis of three CRMs. The concentrations of the analytes spanned four orders of magnitude, but despite this, the analytical data indicated that all were determined with some semblance of accuracy, with agreement being within 20% of certified values. The protocol developed was compared with LIBS and a dissolution procedure with OES detection. The advantage over the latter was that the LA-ICP-MS system did not produce a stream of radioactive liquid waste. In addition, the mass spectrum was significantly less complex than the OES spectra.
A few papers have discussed methods available for overcoming interferences during the analysis. An example was a paper by Gourgiotis et al. who measured the U and Pu isotope ratios using a collision/reaction cell multicollector ICP-MS instrument.214 Under normal circumstances, there is an isobaric interference from the 238U on the 238Pu, but these authors found that the analytes behaved very differently when carbon dioxide was introduced to the collision cell. Accurate and precise Pu and U isotopic measurements on Pu+ and UO2+ were obtained by using two different carbon dioxide gas flow rates: 0.4 mL min−1 for Pu+ and 1.5 mL min−1 for UO2+. Several parameters were evaluated and, if necessary, corrected for including the instrumental mass bias, oxide isotope abundance and memory blank. Other factors were optimised, e.g. the carbon dioxide flow rates. The results obtained using this method were compared with those obtained following a chemical separation method and were found to be in excellent agreement. A second paper has compared the results obtained using a collision/reaction cell quadrupole-based instrument with those obtained using a sector field ICP-MS instrument.215 Under normal circumstances, the quadrupole analysis would be hampered by both isobaric (e.g.90Zr+ and 90Sr+) and polyatomic interferences, but by the addition of oxygen to the cell, the authors managed to detect ZrO+ and Sr+ hence, effectively separating the analytes. Similarly, the addition of nitrous oxide to the cell managed to overcome the interference of 137Ba+ on 137 Cs+ because the Ba formed BaO+ and BaOH+ whereas the Cs remained unreacted. Once the collision cell parameters had been optimised, the authors compared the results with those obtained using chemical separations of Ba from Cs and for Zr from Sr and analysis using the sector field instrument. A further comparison with gamma spectrometry was also made.
Some papers reported simple applications. In one by Emmert and colleagues, two different laser systems (a Nd:YAG operating at 1064 nm with a ns pulse duration and a Ti:sapphire laser operating at 800 nm and with a fs pulse duration) were compared for the LIBS analysis of depleted uranium.216 The latter, more rapid, laser produced a relatively cool plasma in comparison with the ns one. The determination of low atomic number elements (Al, Mg and Na) in a uranium matrix using a vacuum chamber TXRF instrument has been described by Misra and co-workers.217 Samples were prepared from different mixtures of their solutions and a solution of high purity uranium. The uranium was separated from the analytes using 30% tributyl phosphate in dodecane and then the aqueous remnants containing the analytes were spiked with Sc as an internal standard prior to analysis. After blank correction, the TXRF results obtained indicated an average deviation of 14% from the calculated concentrations of these analytes. However, the data for Mg indicated that significant differences were obtained. Despite this, the authors still concluded that TXRF was a promising technique for such analyses. Mirashi and colleagues described a method by which EDXRF was used to determine Th over the concentration range 7–134 mg mL−1 in phosphoric acid solutions.218 Using matrix matched standards and Y as an internal standard, aliquots of between 2 and 5 μL were placed onto an absorbent sheet, and the sheet presented for analysis to the instrument. A calibration plot was made between intensity ratios (Th L-alpha/Y K-alpha) against respective amounts of Th in the calibration solutions. Results obtained using this method were compared with those obtained using gamma spectrometry and were found to agree within 2.6%. The advantage the method developed had over gamma spectrometry was that it was significantly more rapid and required less sample. A novel method of determining U isotope ratios has been presented by Mahani et al. who used ICP-OES and a multivariate calibration technique (a chemometric method) to overcome spectral overlap.219 The emission spectra from natural and depleted uranium was modelled using a program written in-house. Results obtained using the model were compared with those obtained using TIMS and were found to be in good agreement. Some of the more routine applications have been discussed in tabular form (Table 3).
Element | Matrix | Technique;Atomisation; Presentation | Comments | Reference |
---|---|---|---|---|
241Am | Urine | MS; ICP; L | A comparison of gamma spectrometry, liquid scintillation counting and ICP-MS emergency assays for 241Am was performed. All three methods satisfied the ANSI N13.30 radiobioassay criteria for accuracy and repeatability. The ICP-MS protocol was most sensitive and rapid. Authors warned that other methods are more commonly available and that their speed can be improved through automation. | 220 |
B | Uranium oxide | TIMS; -; L | After dissolution, B extracted using 5% 2-ethylhexane-1,3-diol in chloroform. Sample was then evaporated in the presence of mannitol and sodium carbonate. TIMS analysis using determination of the ion Na2BO2+. | 221 |
B | Boric acid | LIBS; -; S | On-line use of LIBS to detect leaks of boric acid from the reaction vessel head of a nuclear power plant. Q-switched frequency doubled (532 nm) Nd:YAG laser used for excitation and echelle spectrometer equipped with an intensified CCD used for detection. | 222 |
Cd | Uranium | EDXRF; -; L | Continuum source EDXRF used to monitor signal using the Cd K-alpha line. Matrix element (uranium) extracted using 30% tributyl phosphate in dodecane. Calibration standards prepared by using solutions of Cd and uranium in pure nitric acid. Precision was ∼ 2% RSD and results deviated from expected values by < 4%. | 223 |
Ga and In | Uranium | OES; ICP; L AAS; ETV; L | Solvent extraction of analytes using 30% tributyl phosphate in carbon tetrachloride. Measurement using axial ICP-OES and ETAAS. Analytes determined from ppm to % levels by prior dilution with 1 M nitric acid and appropriate (100 mg and 1 g) sample size. The analytical range obtained for Ga and In was 0.1–0.5 μg L−1 and 0.02–0.5 μg mL−1, respectively, using ETAAS and 0.5–500 μg mL−1 using ICP-AES for both elements. T-test indicated that there was no significant difference between the results from the two techniques. | 224 |
129I | Seawater | AMS; -; - | A reference material (IAEA 418, Mediterranean seawater) was developed. The certified value was 2.28 × 108 atoms L−1 and 95% confidence interval was 2.16–2.73 108 atom L−1. Material intended for use in AMS laboratories. | 225 |
129I | Lichens from central Sweden | AMS; -; L | Microwave digestion used to improve speed of analysis and reduce contamination. Samples showed the impact of nuclear reprocessing and of Chernobyl incident. | 226 |
Np | Environmental samples | MS; ICP; L | Behaviours of Np(IV), Np(V) and Np(VI) on various chromatographic resins were tested. Two UTEVA resins used to retain Np from digests of soils and sediments. | 227 |
Np and Pu | Environmental samples | MS; ICP; L | The analytes 237Np, 239Pu and 240Pu converted to the (IV) oxidation state using sulfite and then nitric acid and then separated from concomitant materials using TEVA® resin. Sequential injection analysis used to introduce samples to instrument. Analysis of three CRMs yielded results in agreement with certified or accepted values. Method applied to analysis of soils and seaweeds. | 228,229 |
237Np and Pu isotopes | Soil samples | MS; ICP; L | TEVA resin used to preconcentrate analytes whilst simultaneously removing potential interferences from 238U (from peak tailing on 237Np and 238UH on 239Pu). The Pu was then transferred to DGA resin for further purification. Decontamination factor for U was 106. | 230 |
Pu | Sediment core from coast of Palomares, Spain | AMS; -; - | Both 239Pu and 240Pu determined using AMS at an energy of 670 kV. Results compared with those obtained using alpha spectrometry and were in good agreement. | 231 |
Pu | Air filter samples | AMS; -; - | Both 239Pu and 240Pu determined using low energy AMS from air filters that collected sample over a year-long period. Levels of Pu were greater in the summer months. There was a correlation between Pu and levels of Al and Ti. | 232 |
Pu | Sediment from the Yangtze river | AMS; -; S | 4 pg of 242Pu spike was added to the material before leaching with nitric acid. Extraction of Pu from the leachate was achieved using an ion-exchange column. The Pu was then dispersed in an iron-oxide matrix, mixed with silver powder, pressed into a sample holder and loaded into the ion-source of the accelerator. Data obtained for 239Pu and 249Pu as well as the spike isotope. The depth profile of the core using the Pu data was compared with the more frequently used 137Cs tracer (data from gamma counting). The AMS demonstrated superior sensitivity. | 233 |
239Pu | Urine | AMS; -; L | The method developed avoided many sample preparation steps required for the more normally used alpha spectrometry and was therefore much more rapid (10 h rather than one week). Purification of Pu was achieved by using chromatography columns filled with AG1X2 anion exchange resin. The LOD was sufficiently low (13 fg) to be suitable for use as a routine bioassay. | 234 |
Pu | Marine samples | MS; ICP; L | Multicollector instrument used. Comparison of various separation strategies using AG1-X8, TEVA, TRU, and UTEVA resins undertaken. A combination of AG1-X8 and UTEVA/TRU was optimal and led to a purity of 87% Pu and a decontamination factor of U of 1.2. × 104. The LOD for 239Pu and 240Pu were 0.02 fg L−1 (0.11 fg). Validation using several CRMs used. | 235 |
Pu/U ratios | Microscopic particles from Thule and Palomares incidents | PIXE; -; S XRF; -: S | Two micro-beam technologies compared (μ-PIXE and synchrotron radiation μ-XRF). Results in good agreement. | 236 |
236U and 235U | Human tissue samples | MS; ICP; L | Sector field ICP-MS used to determine analytes. Results for 235U compared with those obtained using alpha spectrometry. Good agreement indicated result for 236U obtained using SF-ICP-MS were accurate. | 237 |
Various | Coal and coal fly ash from Philippine power plant | MS; ICP; L | Samples digested using a mixture of nitric, hydrochloric and hydrofluoric acids with microwave assistance. The analytes 226 Ra, 228Ra, 232Th, 238U and 40K were determined. Results for U and Th were compared with those obtained using gamma spectrometry with correlations of 0.97 and 0.94 being obtained respectively. | 238 |
Three papers have discussed the development of plastic CRMs. Two of them were prepared by the National Metrological Institute of Japan and have been reported (in Japanese) by Ohata and colleagues240 and by Matsuyama et al.241 In the first of these reports, acrylonitrile butadiene styrene resin discs were manufactured and then analysed using EDXRF with no further sample preparation. The results obtained for the analytes Cd, Cr, Hg and Pb were compared with those obtained after a sample digestion procedure and analysis using ICP-MS and were found to be comparable. The study reported the preparation and characterisation of four such resin disc CRMs that were given the numbers NMIJ CRM 8105a, NMIJ CRM 8106a, NMIJ CRM 8115a and NMIJ CRM 8116a. The second study described the preparation of polystyrene resin discs of dimensions 30 mm diameter by 2 mm thickness that contained decabrominated diphenyl ether.241 The resulting materials were analysed using EDXRF, which demonstrated that the Br concentration was very homogeneous both within a disc and between different discs. The decabrominated diphenyl ether itself was determined using both GC-MS combined with isotopic dilution and by HPLC with UV detection; the latter using the technique of standard additions. Results from the different techniques agreed well, enabling the material to be given the certified value of 312 ± 16 mg of decabrominated diphenyl ether per kg of material. The material was allocated the name NMIJ CRM 8108b. The third paper of this type was prepared by Vogl and co-workers, who discussed the reference procedures used to certify the contents of S and the four metals Cd, Cr, Hg and Pb in plastics.242 The procedures discussed included double isotope dilution mass spectrometry as well as TIMS and ICP-MS. The problems associated with memory effects during the determination of Hg and of the dissolution of chromium(III) oxide were addressed in particular. The authors discussed how expanded uncertainty on the results improved from percent levels for older materials such as BCR 680 and 681 to between 0.1 and 0.4% for Cd, Cr, Pb and S and to 1% for Hg for the more recent material CQCM-P106.
Analysis using LIBS followed by a chemometric interpretation of the spectra has also been achieved by Godoi and co-workers who applied their protocol to the analysis of plastic toys.247 The analytes Cd, Cr and Pb were determined. Normally, lixiviation is used to determine the extent and the speed at which toxic analytes migrate. This can be a relatively slow process and so the use of LIBS and chemometric tools to obtain a more rapid analysis would be of considerable use. The authors compared several chemometric tools including partial least squares discriminant analysis (PLS-DA), soft independent modelling of class analogy (SIMCA) and k nearest neighbour. The classification models were first constructed using 40 samples and then 11 other samples were used to test them. Best results were obtained using k nearest neighbour which was capable of correctly predicting 95% for Cd and 100% for both Cr and Pb.
Several other LIBS applications have been discussed in the literature. Of these, many are more than just simple applications, but instead attempt to improve the performance of the technique or apply a mechanistic approach. Khachatrian and Dagdigian performed a study designed to determine whether or not LIBS signals obtained from a polymer were enhanced when the laser wavelength was resonant with a fundamental vibrational transition of the particular polymer.248 Using nylon 12 and poly(vinyl alcohol-co-ethylene) as test polymers, the LIBS signal was found to be enhanced significantly when the irradiation wavelength corresponded to hydride stretch fundamental transitions. The authors extended the study to determine if the same effects were observed in both bulk and spin coated samples. Viskup and colleagues studied the influence of the pulse to pulse delay for a 532 nm double pulse LIBS system applied to the analysis of the technical polymers polyamide, polyvinyl chloride, polyethylene etc.249 The Nd:YAG laser used for plasma induction yielded emission signals from the samples that were detected using an Echelle spectrometer and CCD detector. Pulse to pulse delay times ranging from 20 ns up to 500 μs were tested and were found to enhance the signals obtained for atomic and ionic emission lines when compared with the signals obtained from a single pulse. However, the extent of the enhancement was found to vary between polymer types and also on the pulse delay time. In another study by the same research group, a 13.56 MHz frequency RF generator and a gas nozzle were used to generate a continuous jet of argon and nitrogen plasma which was then used to spatially overlap the laser induced plasma, hence enhancing the signal.250 The authors used short pulses from a 193 nm laser to form the laser induced plasma in the first instance and then the plasma jet to further excite it. Detection of the analytes (Al, Fe, Mn and Zn) was achieved simultaneously using an Echelle spectrometer and a CCD camera. The system was applied to ceramic and polymeric samples. After optimisation of the procedure, in terms of the time delay with respect to the laser pulse, modest improvements in signal intensity (approximately two-fold) for both atomic lines and molecular bands were observed.
Houssiau and Mine discussed the use of TOF-SIMS with very low energy reactive ions.254 The ions of both caesium and xenon (at 250 eV) were used to investigate the possibility of molecular depth-profiling of polymers such as polymethylmethacrylate, polycarbonate and polystyrene. It was found that the caesium ions could be used successfully for all three polymers whereas the xenon primary ions quickly degraded the polymers, although some molecular ions remained detectable for the polymethylmethacrylate. The sputtering yields for the polymethylmethacrylate were significantly higher for both primary ions than they were for the polycarbonate and polystyrene. For the latter two polymers, the caesium primary ions provided a much higher sputtering yield than the xenon. The authors attributed this to an inhibition of the cross-linking reactions. The other overall conclusions were that caesium ions should be used for the task because they have low energy and hence cause less damage and they also enhanced negative ionisation. The other TOF-SIMS application was provided by Piwowar and Vickerman, who determined the effects of molecular weight of polymethylmethacrylate on the spectra obtained from 20 eV primary ions C60+ and Au+.255 Molecular weights of between 2770 and 19940 were tested. Using similar experimental conditions, the yields from the C60+ primary ion were between 15 and 65 times those obtained from the Au+ primary ions. Increasing molecular weight of the polymer had very different effects for the two primary ions, with Au+ providing a decrease in yield with increasing mass and the C60+ giving an increase. The authors concluded that this means that there is a difference in mechanism behind ion formation between the two primary ions.
A speciation application was described by Shao and colleagues, who developed an ultra-sonic extraction technique followed by HPLC-ICP-MS determination of selected polybrominated diphenylethers and polybrominated biphenyl in the polymers high density polyethylene, polystyrene, acrylonitrile butadiene styrene copolymer and polypropylene.257 Toluene was used as the extracting material and succeeded in extracting close to 100% of the analytes with a precision in the range 0.7–5.4%. The extraction method developed avoided problems associated with thermal degradation, as can be seen for some of the analytes. This is clearly one of the most important factors for speciation analysis, since if the speciation is changed during extraction, then unreliable/unrepresentative results are obtained.
Several applications have discussed the analysis of polymers without recourse to extensive sample preparation procedures. This has been accomplished using a variety of techniques. In one example, by Okamoto and co-workers, Br was determined in plastics using ETV-ICP-MS.258 The sample was weighed directly into a tungsten sample cuvette and then covered with an organic compound, e.g. octanol or diisobutyl ketone and potassium hydroxide. The cuvette was then placed in the ETV device and a standard temperature programme run that transformed all of the Br-containing molecules into a simple, thermally stable inorganic salt. This enabled calibration using simple aqueous standards of potassium bromate. A LOD of 0.77 pg Br was obtained, which corresponded to a concentration of approximately 0.31 ng g−1 in the solid polymer when a 2.5 mg sample was used. Another application used solid sampling with ETAAS detection for the determination of Cd and Pb.259 Again, aqueous standards were used for calibration. The authors applied their protocol to the determination of the analytes in polymers from waste electronic materials. Detection limits were 0.1 mg kg−1 for Cd and 0.6 mg kg−1 for Pb. Although these were not overly impressive, they were adequate for the application; i.e. they were fit for purpose. Certified reference materials were used to validate the protocol. Santos et al. have used slurry sampling and ETAAS detection for the determination of Cd, Cr, Pb and Sb in plastics from waste electrical and electronic equipment.260 The pulverised plastic materials were dispersed in dimethylformamide and injected directly into the atomiser. Experimental parameters such as char and atomisation temperatures, the char time and matrix modifier composition were optimised for each of the analytes. Under the optimised operating parameters, the characteristic concentrations were 1.9, 32.3, 54.1 and 9.1 pg for Cd, Pb, Sb and Cr respectively. Unfortunately, calibration had to be performed using the technique of standard additions, with additions in the ranges 5–20, 5–30, 12.5–50 and 25–100 μg L−1 for Cd, Pb, Sb and Cr respectively. Limits of detection were between 0.9 and 6.8 μg L−1 and results from spike/recovery tests yielded values of between 96 and 112%. The authors compared the results obtained using the method they had developed with those obtained using a digestion procedure and ETAAS detection.
Laser ablation followed by ICP-MS detection was used by two research groups. In one application, Stehrer et al. determined the concentrations of 35 elements in waste polymer materials, although the study's main focus was on the analytes As, Cd, Cr, Hg, Pb and Sn.261 Several problems were noted with the analyses. Even when polyethylene standards were used for calibration, the data differed markedly from those obtained when a wet chemical attack was used. The difference was even more marked when glass samples were used for calibration. Other problems were observed and these included the very poor intensity for Hg signals. Analysis of the polymer pellets, as obtained, proved virtually impossible because they were too inhomogeneous. The authors therefore resorted to cryo-milling the pellets to reduce the particle size and then pressed them into powder pellets for the analysis. Even with these procedures in place, the analysis was only a partial success, with only some of the elements providing data in agreement with those from other techniques. May and Watling used LA-ICP-MS followed by the multivariate chemometric technique of stepwise linear discriminant analysis of the data so that discrimination between polycarbonate headlamp lens materials could be made.262 A total of 46 analytes were determined. The advantage of the technique was that minimal damage was inflicted on the sample. Minor variation within a single headlamp lens was observed when different laser shots were made. However, it was still possible to discriminate between most of the lenses tested. The only exceptions were between lenses produced from a single manufacturing plant on the same day.
Two papers have been published that reported the analysis of polyethylene terephthalate bottles using XRF as the means of detection. In a relatively simple application, the analysis enabled the determination of the amount of re-cycled material used in the preparation of the bottle, because the re-cycling process introduced inorganic contaminants.263 Using the chemometric techniques of principal component analysis (PCA) and partial least squares model, the authors identified that the Fe k-alpha line was increased in intensity in re-cycled materials and that the amount of re-cycled material used could be predicted. In the other application, Martin and co-workers reported the speciation analysis of Sb in these materials.264 The results indicated that the Sb introduced as part of the catalyst system was present as clusters of Sb(III) that had dimensions of tens of micrometers. The μ-XRF instrument used demonstrated the ability of synchrotron radiation analyses to both map elemental concentrations and to determine the oxidation state.
The technique of total reflection XRF (TXRF) is known to be extremely sensitive. It has, therefore, also found significant use during this review period. Sparks and co-workers developed a method by which pL quantities of solution could be deposited in programmable arrays on a silicon wafer prior to analysis using TXRF.268 The droplets were spaced no more than 100 μm apart and had a diameter of approximately 5–20 μm. Despite these extremely small dimensions, it was still possible to match the droplet to the analysis spot of the TXRF instrument. The effect of deposition type and matrix on the TXRF signal was also investigated. A second paper that reported the use of TXRF used the well known vapour phase treatment (a method in which hydrofluoric acid vapour is used to aid volatilisation) as a means of sample introduction in an attempt to improve the sensitivity during the analysis of silicon wafers.269 The authors, Takahara et al., and members of an ISO committee (ISO/TC201/WG2) were investigating the method to determine 109 atoms cm−2 levels of metallic contamination on the wafer surface. It was found that vapour phase treatment did improve the sensitivity of the analysis, but by varying amounts that depended on the methods and the process conditions. The study used a fuming chamber in an automated instrument. It was found that a higher signal could be obtained when condensation was formed on the sample surface in a humidifying atmosphere and with a decreasing stage temperature. They then used both SEM and AFM to examine the surface of the sample and found that particles of approximately 4 μm diameter were formed.
Several papers have reported applications of μ-XRF. Two similar papers were prepared by Mino and co-workers.270,271 In the first, the micro-beam produced by a synchrotron radiation XRF instrument at the ID22 beamline of the European Synchrotron Radiation Facility was used to provide a two dimensional micrometric resolved (1.7 × 5.3 μm2) map of an indium gallium arsenide ternary semiconductor film. Information from the map was then inserted into a theoretical influence coefficient algorithm based on the use of only a few reference materials coupled with the fundamental parameter approach. The result was an analytical method that could identify the variation of the thickness and chemical composition of the semiconductor film that had been obtained using the selective area growth method. Results from the algorithm agreed to within 1% of those obtained using high resolution XRD. In the second paper,271 the same group used the selective area growth method to produce a quaternary alloy used for optical fibre communications. The analytical techniques of μ-XRF, micro-extended X-ray absorption fine structure (μ-EXAFS) and μm-resolved photoluminescence were used to measure the effectiveness of the selective area growth method. Again, a μm resolved map of the sample was produced (again by the ID22 beamline). The μ-EXAFS data represented the state of the art achievable from third generation radiation sources and were in qualitative agreement with data from XRD. The authors warned that technical improvements were still required to enable μ-EXAFS to be used quantitatively. The third paper of this type used scanning XRF micro-spectroscopy utilising a synchrotron-based microprobe technique.272 MacDonald and colleagues used the technique to reveal distributions of metallic impurities in relatively impure crystalline silicon feedstock for solar cells. The contaminants were found mainly to be Cu, Fe and Zn, although smaller quantities of Mn and Ni were also observed. Most of the metals were found to be present as discrete particles of up to 60 μm in size, but Cu was more evenly distributed, with only 5% being associated with large particles. Over 50% of the Fe detected was present as large particles at the grain boundaries and this was thought to be because of diffusion and precipitation during cooling.
It has been observed that when caesium is used as a primary ion to sputter secondary ions during SIMS analyses, that some Cs ions become attached to the sample. Berghmans and colleagues have attempted to quantify this phenomenon, to explain why it happens and then discussed methods that can be used to minimise it.274 The presence of even a small amount of oxygen (10−9–10−8 mbar) was found to increase the amount of Cs ions retained on the sample enormously. The amount of Cs retained on the sample was measured using ICP-MS. The authors attributed the increased retention of Cs to the formation of caesium-oxide. As well as an increased retention of Cs, significant variations in the degree of ionisation of the silicon and of the target sputtering yield were also observed. The authors concluded that the data they had collected indicated that the target sputtering yield was directly proportional to the Cs-surface coverage and that this was governed by the amount of oxygen present. It was noted though, that the sensitivity to oxygen increased as the primary ion energy decreased. Since ever-decreasing primary energies are used to obtain high resolution depth-profiling information, this implies that increased oxygen leads to decreased erosion rates. Low energy ion beam irradiation using Cs has also been discussed by Hopstaken et al. who used it to obtain depth-profile measurements from a well characterised indium gallium arsenide/gallium arsenide multilayer structure.275 The authors presented a systematic and quantitative study of the impact of ion species (O2+ as well as Cs+ were used), primary ion impact energy and incident angle on the depth resolution obtained. They found that use of the low energy O2+ ion led to the depth resolution being degraded because of the incorporation of high concentrations of O in the altered layer, leading to “detrimental ion beam induced formation of topography”. The extent of the problem was found to be governed by the angle of incidence of the primary beam. The use of a low energy Cs+ primary beam yielded no such problem, with the sputtering of the gallium arsenide being well behaved with no significant transient yield changes and constant depth resolution. A third paper to use caesium bombardment has been reported by Penley and colleagues who used SIMS to characterise the anomalous behaviour for low dose ion implanted P in silicon wafers.276 It was noticed that the P content at the surface and near surface layers varied as a function of time, especially for very low doses, and so the authors analysed wafers that had been stored in a range of conditions that included heating, high and low humidity and liquid nitrogen. Heating the sample resulted in the most anomalous behaviour, with the concentration of P at the surface increasing by three orders of magnitude. Only samples stored at liquid nitrogen temperatures remained stable with respect to surface P content. The authors then optimised the SIMS operating conditions and analysed a second set of samples that had been prepared in different ways. The different preparation methods included thicker surface oxide layers, preamorphisation and annealing. The P surface content was found to be much more stable when the P was implanted through a 5 nm oxide layer and stored in dry conditions.
As discussed previously, low energy primary ions usually yield better depth resolution and so there has been a drive to decrease the energy to ever lower levels. A paper that used ultra low-energy SIMS to undertake a structural analysis of gallium nitride – indium gallium nitride quantum wells was reported by Morris et al.277 Unfortunately, this type of sample has a high resistivity which restricted the necessary electrical path between the analysed area and the instrument ground potential. This resulted in surface charge accumulation and unstable and unrepresentative depth-profiles being produced. The authors therefore used a technique known as optical conductivity enhancement which reduced the material's resistivity. This resulted in the creation of an electrical pathway between the sample and the holder which helped eliminate charge build-up. The overall result was a more accurate depth-profile.
A few other papers have used SIMS as a means of analysing electronic materials. One example, by Tellez and co-workers, used SIMS depth-profiling and SEM-EDX to investigate metallic inter-diffusion in aluminium gallium nitride/gallium nitride/sapphire heterostructures that had had metallic stacks (titanium/aluminium/nickel/gold) deposited on them.278 The samples had then been annealed at 850 °C under an inert nitrogen atmosphere. This treatment had modified the surface and it was these modifications that were investigated using the analytical techniques discussed. The SIMS conditions were optimised so that good sensitivity was achieved and that the ion-induced mixing effects, produced by the primary beam sputtering, were avoided.
Koh and colleagues developed a dynamic SIMS method by which Cu impurities in silicon wafer could be determined with enhanced sensitivity.279 Electron bombardment of the wafer enabled the Cu to migrate from the bulk to the surface where it was detected. The electron bombardment led to an increase in the negative charge density on the surface and this, in turn, led to an increase in the rate at which the Cu migrated. The increased transport of Cu subsequently led to the increased sensitivity of the SIMS measurement.
Two papers that discussed the use of TOF-SIMS have been published that are relevant to this section of the review, although there are many other reports that have described very simple applications. Herrmann and co-workers described the use of different primary ions (caesium, oxygen and argon), various sputter angles and ion energies during the analysis of gallium arsenide/aluminium gallium arsenide and gallium antimonide/aluminium gallium antimonide heterostructures.280 The authors used SEM to examine the craters produced and to estimate the depth resolution from the surface roughness. The optimal parameters for both sample types were the use of argon primary ions at an angle of 55° to the sample normal and with an energy of 0.5 keV. The other paper, by Veillerot and colleagues, discussed the potential errors obtained using TOF-SIMS when Relative Sensitivity Factors databases are used.281 The authors reasoned that such databases can lead to discrepancies of up to a factor of 10 being achieved. They therefore developed a new method by which silicon wafers were intentionally contaminated with droplets of Fe, Mo, Na and Zn solutions, which were then dried and a large area analysis (a macroraster) of several mm in diameter used to image the entire droplet residue. Such a large sampling site overcame heterogeneity or droplet movement problems. Comparison with data existing in the literature was, in general, in good agreement. However, disparities of a factor of approximately two were still observed occasionally and this was attributed to the inconsistent top surface ionisation.
Although LIBS constitutes the majority of the laser-based techniques used in these applications, LA-ICP-MS has also found application. Choi and colleagues described the use of an array type polydimethylsiloxane microchip that could be used as a sampling platform for pL quantities of photoresist sample.286 The microchips had 5 × 5 and 6 × 7 arrays of islands standing proud from the surface and it was onto these islands that the photoresist sample (94.7 pL) could be placed. After drying the sample using UV irradiation, it was ablated for detection using a 213 nm laser in single shot mode. The technique of standard additions was used for calibration. Detection limits for 27Al, 63Cu and 208Pb were 2.33, 15.4 and 5.72 ng mL−1 respectively, which were approximately 10 times superior to those obtained using spin coating sampling on a silicon wafer. For method validation, the authors used a microwave digestion of the photoresist sample and analysed the digests using ICP-MS. The advantages of the proposed technique were speed, ability to analyse pL volumes of sample and the ease of sample preparation. The other paper of interest to this section compared the use of nanosecond and femtosecond laser ablation methodologies during the LA-ICP-MS analysis of several sample types including glasses, steel, brass and sulfide semiconductors.287 An argon fluoride excimer laser was used for the ns methodology whereas a titanium-sapphire laser was used for the fs one. The authors also used electrostatic sampling of the aerosols produced by the ablation process to coat TEM grids and then studied the morphology and elemental composition of the particles using TEM and TEM-EDX respectively. The results demonstrated that the ns system produced nanosized particle agglomerates for all sample types, but with the conducting samples also producing spherical particles of typically 50 to 500 nm in size. Conversely, the fs laser ablation created a mixture of spherical particles and nanosized agglomerates for all of the sample types. Surprisingly, the differences in elemental composition between the agglomerates and the spherical particles were more pronounced for the fs system than for the ns one, indicating very different ablation and particle formation mechanisms. The authors hypothesised that for conducting samples in the ns system, the mechanism was likely to be a gas to particle conversion followed by agglomeration and additional hydrodynamic sputtering. The mechanism for the fs system was thought more likely to be phase explosion. The overall result was that during LA-ICP-MS determinations, the ns system was more likely to lead to temporal elemental fractionation effects, primarily during the ablation process, whereas these effects were not observed in the fs system. Instead, the fs system could, potentially, lead to ICP-induced fractionation.
The use of an ICP-OES instrument to monitor the process of mercury cadmium telluride in real time has been reported in a preliminary study by Stoltz and colleagues.289 Both Hg and Cd could be detected, but the signal was dependent on factors such as sample area, material composition, etch rate, sample temperature, photoresist area and plasma power. The emission intensity of H decreased dramatically in samples that contained photoresist, which was contrary to the theory that such samples add H to the plasma. It was concluded that H complexed with the photoresist, hence reducing the overall amount available in the process and that this could, at least partially, explain macro-loading issues whereby additional photoresist areas slowed the etch rates of mercury cadmium telluride and cadmium telluride materials.
A paper by Usov and co-workers discussed the contribution of ion beam analysis methods to the development of second generation high temperature superconducting wires.290 One of the crucial steps in this second generation of materials was the development of the buffer-layer architecture; i.e., a layered system that consists of a nucleation layer, diffusion barrier, biaxially textured template and an intermediate layer that provides a suitable lattice match to the superconducting compound. The report described how ion beam-based analytical techniques such as SIMS, RBS, channelling, PIXE, proton induced gamma ray emission (PIGE), nuclear resonant reaction analysis (NRRA) and elastic recoil detection analysis (ERDA) were employed for the analysis of each of the buffer layers and of the yttrium barium copper oxide superconducting film. The results obtained enabled the authors to understand a variety of physical processes occurring during the buffer layer fabrication and therefore helped them to optimise the buffer-layer architecture.
Two simple applications were reported that described the determination of analytes in printed circuit boards as part of the Restriction of Hazardous Substances (RoHS) and Waste from Electrical and Electronic Equipment (WEEE) directives of the European Union. These regulations limit the amount of Cd, Cr (hexavalent), Hg, Pb and some organic compounds that can be used in electronic equipment within the European Union. In the first of these examples, Martin et al. compared different sample preparation procedures prior to EDXRF analysis.291 Different components of the circuit boards were analysed and the data compared with those obtained after the same components had been ground into a fine powder using a novel cryogenic grinding mill. The other study was very similar and also investigated the influence of sample pre-treatment on the precision and reproducibility of the results.292 Five different types of dissolution procedures were tested on samples that had been cut into small fragments or milled. The analysis was achieved using techniques such as ICP-OES, XRF and CV-AFS (for Hg). The results demonstrated that particles with a size of less than 1.5 mm were sufficiently small to obtain accurate results, but that the precision values were often quite poor. To improve the precision to better than 20%, as recommended by British Standard EN62321, a much greater effort to decrease particle size was necessary.
As well as LIBS analyses, other techniques have also been adopted. Austin et al. used LA-ICP-MS to obtain quantitative elemental bio-imaging from biological soft tissues by using metal spiked polymethylmethacrylate films that had been spin-coated onto quartz substrates as calibrants.298 The films contained concentrations of Cu and Zn of between 0 and 400 μg g−1 and the calibration curves produced from the determination of 63Cu and 66Zn yielded correlation coefficients of better than 0.999. Results were found to be in good agreement with those obtained using conventional solution nebulisation ICP-MS. The authors also developed a method by which an internal standard (Ir or Ru) could be used. The internal standard was placed in a film underneath the biological sample and then they were ablated simultaneously to detection.
Laser assisted atom probe tomography was compared with SIMS, TEM and high resolution XRD (HRXRD) for the analysis of a silicon, silicon-germanium multilayered system in a paper by Koelling and co-workers.299 The impact of the laser power on the accuracy was also assessed. Under optimised conditions, the atom probe was found to offer superior depth resolving powers compared with SIMS by a factor of three and provided a decay length of less than 0.4–0.6 nm/dec. The composition and layer thickness values were, however, in good agreement with the SIMS, HRXRD and TEM data.
The same research group has also published several papers describing analytical methodologies. In one, they discussed the use of pulsed RF-GD-OES for depth-profile quantification of metallic coatings.301 The system used a RF power of 50 W, a pressure of 600 Pa, a frequency of 1000 Hz and a duty cycle of 50%. A comparison of emission intensities and emission yield parameters was made with the continuous mode of operation. The pulsed mode was found to have several advantages, including a higher emission yield when the same mass of sample was excited, a more stable signal throughout (continuous mode led to preferential sputtering during the determination of Zn in some sample types and hence, a degradation of precision) and an improvement in depth resolution. Calibration was achieved by using a simple multi-matrix procedure. The process was validated by analysing conductive layers, of samples such as hot-dipped zinc, galvanneal, the back contact of thin film photovoltaic solar cells and tinplates, with thickness spanning the range between tens of nm up to 20 μm. In a similar application, the same research group studied the analytical performance of pulsed RF-GD-OES during the depth characterisation of zinc–titanium dioxide nanocomposite films deposited on nickel or steel.302 Again, the authors explored the relative merits of the pulsed system when compared with continuous mode and then used the technique to determine the thickness and composition of the films using a simple multi-matrix calibration procedure. On this occasion, the operating conditions were somewhat different to the previous paper, with a RF power of 75 W, a pressure of 600 Pa, a 10 kHz pulse frequency and a 50% duty cycle being optimal. Results were comparable to those obtained using the alternative techniques of SEM and ICP-MS. Although GD may be applied to conducting, semi-conducting and non-conducting materials, its applications for the latter are relatively few because it is known to suffer from some problems. These problems are thought to be related to the low power deposited in the plasma because of a voltage drop developing inside the material. This may be quantified as the voltage transfer coefficient and is the ratio between the voltage at the front and at the back of the sample; it is dependent on the sample capacitance (the sample surface, permittivity and thickness). The same research group attempted to minimise these problems by depositing a thin conductive top layer on both sides of the sample, hence increasing their voltage transfer coefficient.303 In addition, the authors determined the effect of applying a magnetic field during the GD analysis and found that it offered the advantages of increased sputtering rates, better ionisation and better excitation efficiencies and, hence, an overall improvement in emission intensity was achieved. Schmitt and co-workers used GD-OES and profilometry to characterise aluminium-doped zinc oxide films that are used for photovoltaics and that had been prepared by atomic layer deposition.304 A procedure was developed that enabled thickness, mean chemical composition and refractive index of the films to be determined simultaneously.
A paper by Lobo et al. discussed the quantitative depth-profiling of As and B ultra-low energy implants on silicon using the technique of pulsed RF-GD-TOF-MS.305 The authors again calibrated the analysis using a simple multi-matrix calibration procedure and then applied it to the analysis of materials that had been prepared under different ion dose and ion energy conditions. Results of the pulsed RF-GD-TOF-MS analysis were compared with those obtained using SIMS and GIXRF and were found to be in good agreement. The authors admitted that further work was required to explore the depth resolution achievable, but concluded that the technique showed a great deal of promise because of its high sensitivity and high sample throughput (because it does not require sampling at ultra-high vacuum). Another research paper to report the use of pulsed RF-GD-TOF-MS was prepared by Molchan and colleagues who used it to determine impurities (B, Cl, Cr, Cu and P) in thin layers of anodic alumina and in anodic tantala films (where B, Cr, H and P were the analytes of interest).306 An orthogonal TOF-MS instrument was combined with a suitable acquisition system to enable the monitoring of the ion species produced during the glow discharge period, i.e. the time during the RF pulse and after the afterglow, where Penning ionisation results in high signals being observed. It was concluded that GD-TOF-MS was an extremely powerful and reliable technique for depth-profile analysis and enabled the analysis of layers of only 2 nm thickness that were enriched with both positive and negatively charged impurities.
A benchtop EDXRF instrument has been used by Queralt et al. to determine the thickness of single layers of gallium nitride thin films (of thickness in the range 400 to 1000 nm) that had been epitaxially grown on sapphire substrate.308 The results correlated well with those obtained using optical reflectance. The indication was that the technique could possibly determine layer thickness down to 5 nm because it had such good precision (better than 2% RSD). Given that the technique of optical reflectance has limitations at very low film thickness, the authors concluded that the EDXRF method offered considerable potential for film thickness measurements of such samples.
The development of several 3D-XRF instruments has been reported in recent years. Tsuji and Nakano have reported the development of a new confocal 3D-XRF instrument that had a fine focus X-ray tube, advanced polycapillary optics and a silicon drift detector with a large sensitive area.309 The depth resolution was found to be 13.7 μm for the Au l-beta line at 11.4 keV, which was a factor of between three and four better than previous 3D-XRF instruments developed by the same research group. The depth resolution available was dependent on the energy range used, the analyte and whether it was an alpha or beta line used. As an example, the Cr k-alpha line gave a resolution of 22.6 μm at an energy of 5.41 keV, whereas the k-beta line had a resolution of 21.6 μm at 5.95 keV. The new and a previous instrument were compared for the analysis of several layered materials as well as a 3D structured material consisting of two cylindrical patterns of Au having μm scale structure.
The non-destructive depth-profiling method of angle resolved X-ray photoelectron spectroscopy (ARXPS) gives information on the elemental content as well as chemical information. It does, however, suffer the drawback of requiring dedicated mathematical modelling and, thus far, has been restricted away from near-surface angles. This latter drawback has partially been tackled in a paper by Oswald and Oswald who presented a method that used a simple algorithm to consider the effects of elastic scattering which occur in standard ARXPS measurements.310 In principle, this would allow the use of the whole angular range, leading to an increase in the information available. The potential of the approach was demonstrated with model calculations for examples of a few thin films.
Many authors used cluster primary ions to achieve enhanced sensitivity. A brief review (79 refs) of the use of cluster ions during SIMS and TOF-SIMS analyses was provided by Seah and Gilmore.312 Numerous aspects were discussed, including the increased secondary ion yields, the raised sputtering threshold energies, damage effects and depth-profiling capabilities. Several examples and applications were discussed including the use of cluster ions for studying molecular structure using the technique of gentle SIMS (G-SIMS). This example demonstrated the necessity of using cluster ions for analysing the weaker, high mass peaks. A comparison between MeV monomer ion and keV cluster TOF-SIMS for the imaging of biomolecules was made by Jones et al.313 The rationale of the research was to develop a method of imaging sub-cellular structures without the need for many of the procedures required for other techniques such as matrix assisted laser desorption/ionisation (MALDI). Other techniques, such as PIXE and Rutherford Backscattering Spectrometry (RBS) used in combination with the TOF-SIMS enabled improved quantitation and sensitivity. The results indicated that spectra obtained using MeV ions showed similarities to those acquired with more conventional keV ion beams of Bi+, Bi3+, Bi5+ and C60+. Another example has been reported by Fujiwara and colleagues, who used iridium carbonyl metal-cluster complexes such as Ir4(CO)12 and Ir4(CO)7 that have molecular weight of greater than or close to 1000 to analyse a contaminated silicon substrate and the room temperature molten salt N,N-diethyl-N-methyl-N-(2-methoxyethyl)ammonium bis(trifluoromethanesulfonyl)imide.314 The sample was bombarded with Ir4(CO)7 at an incident angle of 45° and with an energy of 10 keV. Results were compared with those obtained using a standard Ar+ primary ion beam. It was found that the cluster molecule enhanced the secondary ion intensity and that it caused less fragmentation. Other examples included the use of C60+ sputtering to improve the detection limit of N in zinc oxide using TOF-SIMS315 and for the molecular depth-profiling of model organic films, also with TOF-SIMS detection.316 In the first of the applications, the primary beam had an energy of 10 keV. Compared with the use of a standard Cs+ primary ion, the cluster ion produced an increase in signal intensity by a factor of 200 and an improvement in LOD by a factor of 10. Analysis using XPS indicated that the 10 keV primary beam did not lead to significant carbon deposition at the bottom of the sputter crater in the sample. The second application used a primary ion of C60+ with an energy of 40 keV to analyse the changes in chemical environment of guanine thin films as a function of ion fluence.316 Information on the chemical damage accumulation as well as changes in the molecular ionisation probability was gleaned from the direct comparison of the secondary ion and neutral components of the molecular depth-profiles. Mahoney discussed the use of cluster SIMS with SF5+ and with Bi3+ primary ions for the depth-profiling of stereo-specific polymethylmethacrylate thin films (approx. 140 nm thick).317 The damage characteristics on the sample were measured as a function of temperature (between −75 and +150 °C). Isotactic material tended to suffer worse damage than syndiotactic and atactic materials. In addition, signal precision increased as a function of primary ion (SF5+) dose. Damage was less when elevated temperatures were used during the sputtering process, although other interfacial effects appeared close to the glass transition temperature of the various polymers.
The other main focus of interest for SIMS and TOF-SIMS applications of thin films has been the use of low or ultra-low energy primary beams. Several authors have used low energy O2+ or Cs+ ion beams. Included in this number is a paper by Merkulov and co-workers, who compared the use of both primary ion types during the depth-profile analysis of As, B and P in ultra-shallow implants.318 The ablation rate achievable was 1 and 2 nm min−1 for Cs+ and O2+ respectively when a primary ion beam with an energy of 150 eV was used. The depth-profile results obtained using these sputtering conditions were compared with those obtained using the techniques of high resolution RBS and elastic recoil detection analysis (ERDA). A second paper to have discussed the use of both Cs+ and O2+ as primary ions was published by Drozdov and colleagues.319 These authors determined the elemental composition of a multi-layered nanostructure using SIMS and determined the optimal conditions that enabled a considerably enhanced depth resolution to be obtained. As well as the use of low energy primary ions at an incidence angle of 45° for the sputtering process, the detection of cluster secondary ions was also found to enhance the depth resolution. The slight drawback with the protocol was that analysis of the surface of the crater left after sputtering indicated that it was slightly rougher than the original sample. Despite this, the main contaminants in two different types of multilayer structures were determined successfully. Low energy (250 eV to 1 keV) Cs+, O2+ and Kr+ ion beams have been used by Zhu et al. during the analysis of epitaxial silicon germanide films and a germanium ion implanted silicon standard.320 A study was made of the matrix effects exhibited under various conditions. It was shown that preferential ion yield enhancement of one matrix component over the other could occur as a result of primary ion incorporation. It was, therefore, necessary to define and then apply a matrix yield factor to obtain constant secondary ion yield ratios between the Ge and the Si. Unfortunately, these factors were strongly dependent on operating conditions and so once defined, the operating conditions could not be altered. However, the factors were valid over a very large concentration range. The O2+ primary ion beam was then used for the analysis which concentrated on steady state ion yields and surface transients. The authors reported that unique features of both were observed and that this was only possible through the use of the low energy ion beam conditions. Morris and Dowsett also used ultra-low energy O2+ ions (250 ev–1 keV) for the analysis of silicon germanide films, where x in the formula Si1−xGex could be determined accurately.321 Roughening of the material occurred when x approached 1. The authors therefore established conditions that enabled the whole range (i.e. where 0 ≤ x ≤ 1) to be quantified. These conditions included an ion beam energy of ≤ 500 eV used at an incidence angle that was close to the normal to the sample. The authors undertook a comprehensive study of the useful ion and sputter yield behaviour as a function of the incidence beam energy. A second paper by the same research group used an O2+ ion beam with a range of beam energies to obtain a depth-profile of a silicon germanide sample.322 By plotting the depth-profile as a function of the beam energy (0.25–2.5 keV), it was possible to extrapolate the curve back to zero energy and this was proposed to be a means of removing the effects of atomic mixing and surface transient behaviour. Again, the incidence angle was close to the normal. A monotonic increase in the apparent layer thickness was found with decreasing energy over the range 2.5 to 0.4 keV, but below 0.4 keV the apparent layer thickness was found to decrease. This was attributed to an increase in the relative surface to bulk erosion rate out-weighing any decrease in the true width of the transient region. It was concluded that using “zero beam energy” did not necessarily yield data close to the truth. A second paper to study “zero-energy” SIMS depth-profiling has been prepared by Vanhove and colleagues.323 To achieve monolayer depth resolution, it is imperative that the surface be almost atomically flat during the electron beam-induced etching. However, it was observed that the surface roughness of silicon with electron beam-induced etching and XeF2 were the same as for just silicon with XeF2. This was attributed to subsurface F influencing the unactivated formation of the SiFx reaction layer and, hence, contributing to the roughness. The electron beam-induced etching of silica led to extreme pit formation because of preferential etching of Si at the silica/Si interface. It was concluded that preferential etching can either deteriorate or enhance depth resolution, depending on the sample type. In the case of an ultra-shallow silicon germanide structure, it was found to enhance the resolution. A paper by Seah and co-workers undertook sputter depth-profiling of ultra-thin silicon dioxide layers on silicon to evaluate the variation of the sputter rate over the first few nm.324 Low energy (600 eV) Cs+ ions at an incidence angle of 60° were used as primary ions. When assuming the signal is linear with O content, the sputtering rate fell by a factor of 4.8 with an exponential decay near 1.2 nm. When the signal was assumed to be non-linear with respect to O, then a rapid change in sputtering rate was still observed, but the factor decreased to only 3.5 of that at equilibrium.
Several papers have discussed methods that can overcome interferences. The isotopic comparative method for SIMS was discussed in two papers by the same research group.325,326 In the first of these papers, the method was used to try to quantify high concentrations (up to 40%) of B in silicon. The technique required the use of two specific isotopes in a sample. In the example used, 10B was present at a constant but sufficiently dilute concentration and could be quantified using the conventional relative sensitivity factor method. The other isotope, 11B, was implanted at very high concentration. The intensity of the 10B whose concentration relative to the Si was well known and constant enabled the measurement of the matrix effects induced by the presence of the 11B. The total B content could therefore be calculated. It was found that using Ar+ sputtering, there were no effects on the Si+ ions, but there was a doubling of the B signal. When using O2+ as primary ion, both the Si and B signals per atom increased almost linearly with increasing concentration, with the signal being 2.5 times that given at lower concentration for both elements. Under oxygen saturation, the signals obtained under Ar+ or O2+ bombardment increased for both the Si and B by a similar magnitude, but that increase was relatively small. The results enabled an extended relative sensitivity factor method to be developed. The second paper used 18O and 16O to measure the variations of the relative ion yields of B+, O+, Si+, B− and Si− as a function of O concentration.326 The near-flat profile of 18O measured at weak concentration enabled the 16O to be determined. It was found that the B+ ion yield was very dependent on the O concentration whereas the other ions were not.
Two papers by Py et al. in which silicon germanide materials were analysed, have discussed the use of the full spectrum method.327,328 As discussed previously, matrix effects can limit the usefulness of SIMS measurements and therefore many workers use the measurement of MCs2+ secondary ions, which enables the Ge concentration to be quantified in a Si1−xGex type material over the range 0.05 ≤ x ≤ 0.85. In these two papers, a different approach was described in which negative ion detection using TOF-SIMS was used. When used alone, a fair linear correlation between ion ratios and Ge/Si layer composition ratios was achieved. However, there were still deviations from linearity and this had a detrimental effect on the precision of the measurement.327 The full spectrum method assumes proportionality between composition of the secondary ion beam and the material itself. Silicon germanide layers were depth-profiled using a long cycle time to enable the detection of SinGem clusters, where n and m could be as high as six. Results obtained using this methodology were compared with those obtained using XRD and an excellent linear correlation was observed. The method was found to be extremely precise and reproducible due to the matrix effects being minimised. The second paper extended its use for the depth-profiling of the matrix elements Si and Ge, but also determined the impurities B, C and P in strained SiGe/Si superlattices. Accuracy was assessed by comparison of the results with those obtained using other, more classical, protocols. Overall, the results were in good agreement. Those obtained using the full spectrum method were, however, superior in terms of signal to noise ratio and signal stability. In addition, the method developed also provided slightly enhanced depth resolution.
It is well known that annealing can form defects in materials and that during depth-profiling these may lead to significant errors. A study by Fujiyama and colleagues discussed the influence of these errors on conventional front-sided SIMS depth-profile measurements and then developed robust measurement conditions for the use of back-sided SIMS.329 The paper described how boron difluoride had been implanted in crystalline silicon and then how the material was treated in three different ways: left un-annealed, annealed at 750 °C and annealed at 900 °C. In the un-annealed sample, the results of front- and back-sided SIMS were in good agreement. For both of the annealed samples, the front- and back-sided SIMS results differed markedly, with the depth profiles obtained using back-sided SIMS being significantly more steep than those from front-sided measurements. The authors concluded that this demonstrated that the degradation in front-sided measurements was not purely a function of ion mixing caused by primary ions, but was also related to the crystalline structure in the implanted region. In addition, it was also concluded that back-sided SIMS was capable of obtaining more precise depth-profiles.
Several papers have reported a mechanistic approach during their study, i.e. have attempted to determine why something may work the way it does. The determination of Na depth-profiles in insulating samples using TOF-SIMS with a O2+ primary ion beam is problematic because the Na is known to migrate under the influence of an internal electrical field. The effect of temperature on this effect has been studied by Krivec and co-workers, who discovered that it was a dynamic process “in concordance with the proceeding sputter front”.330 Low temperature was found to decrease the effect somewhat by reducing the Na mobility in the target. The two types of sample used were a Na doped polymethylmethacrylate layer deposited on a silica thin film and Na implanted directly into a silica thin film. The former demonstrated the incorporation behaviour of Na into silica during depth-profiling whereas the latter enabled the examination of Na migration behaviour when defects arising from the implantation process were present in the silica target. Other authors have commented on the difference in results obtained using different depth-profiling techniques. Trigoulet et al. compared GD-OES with TOF-SIMS and found that they produced very similar depth-profiles for samples that had a flat surface, but very different ones when the surface was rough.331 Using anodic oxide films on textured aluminium substrates as samples, the authors concluded that the dissimilarities in results between the two techniques arose because of the different sputtering processes that occur.
Some papers have reported the use of a new instrument or a new sampling chamber. Examples of this include a paper by Xing and colleagues who reported the depth-profiling of nanometre coatings using low temperature plasma probe combined with ICP-MS332 and another by Bendler and co-workers who reported the use of a new neutral caesium evaporation chamber and an ultra-high vacuum “suitcase”.333 In the first of these examples, a low temperature plasma probe, which had a diameter of several tens of μm, was used to sputter the material and convert it into an aerosol whereupon it was transported to an ICP-MS instrument for analysis.332 Examination by SEM of the crater left in the surface after the probe had sampled the material indicated that the crater was less than 10 μm in diameter and that lateral resolution was 200 μm. Depth-profiling a 100 nm layered sample and a multiple layer sample (100 nm aluminium/250 nm silica/100 nm gold/50 nm chromium) was achieved successfully, even under ambient conditions. It was concluded that the technique offered high spatial resolution, rapid analysis speed, was easy to fabricate the probe and was easy to implement. The authors stated that it could act as a complementary technique to more established ones such as GD-OES/-MS, SIMS etc., and could be coupled with an assortment of instrumental detection techniques. The second of these two papers discussed the relative merits of using Cs+ as the primary ion beam (the ability to monitor MCsx+ secondary ions that help reduce matrix effects) and the drawbacks (the Cs+ beam becomes incorporated into the surface layers as well as sputtering the surface).333 This problem means that the sputter yield and the amount of Cs incorporated into the surface are dependent on the primary ion bombardment conditions. The authors overcame this problem by developing a new cation mass spectrometer that was comprised of an evaporator that delivered a collimated and adjustable stream of neutral Cs onto the sample. The Cs–O delivery system was then made compatible with other instrument types by developing an evaporation chamber and an ultra-high vacuum transfer system that was capable of being coupled with SIMS detection. The performance and characteristics of these items were then evaluated.
High energy synchrotron radiation XRF (116 and 75 keV) has been used by Nishiwaki as a forensic tool to discriminate between small fragments of glass, titanium dioxide pigments, automotive white paints, polyester single fibres and ceramic prints on automotive glass (in Japanese).335 Using the k-series of lines, the technique enabled the author to determine both relatively heavy analytes (including the REE) and lighter elements. The trace heavy elements could then be used for fingerprinting purposes and enabled particles with a size of no more than 0.5 × 0.5 mm to be characterised in a non-destructive way.
Another forensic application, by De Young and colleagues, discussed the use of PIXE to discriminate between different types of glass fragments.336 The authors admitted that the technique was not as sensitive as LA- or other ICP-MS based techniques, but it did have the advantage of being entirely sample non-destructive, had a very simple sample preparation procedure and had sufficient sensitivity to act as a method of pre-sorting samples ready for further analysis. It was therefore concluded that although unlikely to be used as a mainstream technique, it could easily find a niche for certain applications. It was also found that the method had a higher sensitivity than some XRF methods for many of the higher mass elements.
The last paper to be reviewed in this sub-section was by Smolders and Krystek; who discussed the chemical characterisation of recycled glass and detailed the entire process from sampling to the analysis using WDXRF.337 Since re-cycled glass is an increasingly used component of new glass products, it must be carefully characterised (along with the other raw ingredients) so that the manufacturers can guarantee a constant high quality of product. A multi-point calibration was achieved using both certified and in-house reference materials. Since the recycled glass component may constitute 400–2000 tons of a batch of sample, the authors had to develop a sampling and then sub-sampling regime that was representative of the sample as a whole. This protocol, along with the final preparation procedure stage (which used a melting of a flux bead) was described.
Several other LIBS applications have been reported. In one, it was used to analyse REE-doped oxyfluoroborate optical glasses.340 The laser beam (Nd:YAG operating at 532 nm) was focussed using a 150 mm quartz lens to give a spot size of 0.02 mm and the light emitted from the induced plasma was collected using a convex lens and transported via a fibre optic cable to the entrance slit of a spectrometer. The intensity at each wavelength over the range 190–900 nm was measured using a CCD detector. The authors, Dwivedi and colleagues, found that several REE, other than the one deliberately doped into the glass, as well as C, Ca and Fe were observable in the spectrum produced. The technique was capable of determining very light elements (e.g. B, F and O) as well as the much heavier elements (Ba, Ca, Er, Eu, Fe, Ho and Nd) that were present in the optical glass. Dehghanpour and Parvin have produced two papers describing the LIBS analysis of amorphous silica glasses.341,342 The decomposition of silicon hexafluoride in the vicinity of the silicon dioxide glass and the penetration into the glass by F was investigated using varying laser wavelengths (e.g. 193, 248, 532 and 1064 nm). The surface of the sample was analysed using techniques such as LIBS, SEM, RBS, WDXRF mapping and energy dispersive X-ray microanalysis. In one example,342 excitation of the sample using an argon fluoride excimer laser created silicon tetrafluoride gas and disulfur decafluoride clusters. No such by-products were observed when a Q-switched Nd:YAG laser was used. A paper by Son and co-workers discussed the use of argon gas jets injected through a pulsed nozzle onto the sample area to be analysed and described their effects on the sensitivity of LIBS analyses.105 By synchronising the buffer gas pulse with the laser beam used to form the plasma and optimising the spatial arrangements between the gas jets and the sample surface, the authors obtained a ten-fold sensitivity enhancement. The mechanism of the enhancement was discussed and the method was applied to the analysis of metals, glasses, paper etc.
The other approach to laser-based analytical techniques was the use of LA as a means of sample introduction to atomic spectrometric detectors. A paper by Le Roux described the use of LA-MC-ICP-MS as a means of Li isotope analysis in natural and synthetic glasses.343 Using a laser operating at 213 nm and in a helium atmosphere, the sample was ablated into the instrument and some of those ions sampled from the plasma then passed through a “low mass, high abundance” skimmer cone. Asymmetric tuning of the quad lens system enabled the simultaneous measurement (using Faraday cups) of both 7Li and 6Li. The “bracketing” system was used for the analysis, i.e. alternating between a standard (BCR-2G) and a sample. This enabled the accurate determination of the Li ratios for concentrations between 3 and 35 ppm Li with precision being better than 1 part per thousand. The second paper of this type was prepared by Mertz-Kraus et al. who determined Th and U isotopic ratios (230Th, 232Th, 234Th, 234U, 235U and 238U) in silicate glasses.344 Instead of using a multi-collector style instrument, which is the more commonly used instrument type for the task of isotope ratio measurements, the authors used a single collector sector field instrument. Ablation of the sample was achieved using a frequency quintupled Nd:YAG laser at 213 nm and with a beam width of 110 μm. Precision and accuracy of the data was assessed by analysing assorted different reference matrices, e.g. glasses (NIST SRM 612, MPI-DING (ATHO-G and T1-G), and USGS (BHVO-2G, NKT-1G and BCR2-G)), 91500 zircon and a travertine. Corrections for background signal as well as for the tailing of 232Th on 230Th and of 235U on 234U had to be made. In addition, further corrections for instrumental mass discrimination and elemental sensitivity also had to be made. The latter two were found not really to be problematic for most sample types, but were for SRM 612. Values for several isotopic ratios were obtained and the results compared with those obtained using TIMS. The data from the two different techniques were found to be in good agreement (within 2%).
The surface of float glass was analysed by Sundberg and colleagues, who discussed the use of a surface ablation cell to achieve highly spatially resolved data (10–20 nm) and with good spectral resolution.346 The results compared very favourably with those obtained using SIMS. The surface ablation cell utilised a wet dissolution of the sample, on a layer by layer basis and used equipment available in most standard chemistry laboratories. The concentration of the acid used as well as the sample pre-treatment and many other experimental parameters were found to affect the results and the step length of the analysis and so had to be optimised. The procedure was judged to be sufficiently successful to have use both in laboratories and in the production environment as an off-line surface characterisation technique. The work was a collaborative programme organised by Technical Committee 2 (Chemical Durability and Analysis), a technical sub-committee of the International Commission on Glass (ICG/TC-02).
The final application of interest was presented by Sjastad et al. and reported the forensic discrimination of glasses according to their Pb isotope ratios.347 Forensic scientists have, for many years, differentiated different glasses according to their refractive index. However, in recent times, the refractive indexes of many glasses are becoming increasingly similar. In addition, glass that has endured a high temperature, e.g. following a fire, may have a somewhat altered refractive index and so comparing it with the parent material can be problematic. These authors therefore developed a simple test to differentiate between glass samples. Glass was first powdered and then dissolved in a mixture of hydrofluoric, hydrochloric and nitric acids (2:1:1) on a hotplate. After evaporation of the acids, the residue was re-suspended in 1 mL of 6 M hydrochloric acid. After heating to dryness, the residue was treated with 10 drops of 0.8 M hydrobromic acid, evaporated to dryness again, re-suspended in 1 mL of hydrobromic acid and centrifuged at 4000 rpm for 30 min. The sample was then passed through AG1-X8 resin to retain the Pb isotopes whilst enabling many of the matrix components to be eluted to waste. The Pb isotopes were then eluted using nitric acid and determined using an MC-ICP-MS instrument. The method was validated using the material SRM 612 and a solution of NIST 981 was used regularly to ensure instrument stability. Mass discrimination was accounted for by determining Tl simultaneously. The statistical method of Tukey's Honest Significant Difference was used to statistically identify which of the samples was the closest match.
Analyte | Matrix | Technique; Presentation; Atomisation | Comments | Reference |
---|---|---|---|---|
Various (61) | Wood ash glass | MS; ICP; LA WD -microprobe | Five sub-types of wood ash glass analysed (early medieval wood ash glass, wood ash glass, early wood ash lime glass, wood ash lime glass, mixed alkali glasses). Differences in age identified from CaO/K2O ratio. Other distinctions could be made from elemental profiles. 29Si used as internal standard and calibration was achieved using NIST 610. | 348 |
Various (61) | Glasses from late antiquity and middle ages | MS; ICP; LA | Soda ash (20), soda lime (16) and wood ash (23) glasses analysed. Concentrations of K, Mg and Na enabled glass type to be determined. Determination of other elements enabled region of origin to be ascertained. | 349 |
Various | Plant ash glass vessels | XRF; -; S | Portable XRF instrument used to analyse glass vessels found in the South Sinai area of Egypt in situ. Chemometric analysis of the data (PCA and cluster analysis) enabled a greater understanding of the chronological changes of the composition of the glasses. NIST 610, 612, 621, 1412, 1830 and 1831 as well as 15 other synthetic glasses were used to assess accuracy and precision. | 350 |
Various | Mineral soda alumina glass | MS; ICP; LA | 486 artefacts analysed using LA-ICP-MS in an attempt to identify different manufacturing centres. Concentrations of Ba, Ca, Cs, Mg, Sr, U and Zr identified 5 different sub-classes of glass. Manufacturing sites appeared not all to originate in South Asia as had previously been assumed and operated over different time periods. | 351 |
Various | Glass beads from Medieval Al-Basra (Morocco) | MS; ICP; LA | Six chemical types of glass determined and light shed on the geographical origin and techniques of manufacture. The CRMs NIST 610 and 612 as well as other materials were used for calibration. Reproducibility was reported to be better than 10% and accuracy was typically within 5–10% of expected values. | 352 |
Various (47) | Glass beads from Southern Africa | MS; ICP; LA | 360 glass beads from 19 sites analysed. Analysis confirmed that 8 types of glass bead previously identified through morphological characteristics did have different glass chemistry. | 353 |
Various (18) | Indo-Pacific glass beads | XRF; -; S | 22 samples of beads analysed using XRF to try and determine whether they were manufactured in India or in Sungai Mas (Malaysia). Results indicated that Sungai Mas manufactured their own beads between the 6th and 13th centuries. | 354 |
Various | Glass tesserae from mosaics | XRF; -; S XPS; -; S | Assorted techniques (SEM, UV-VIS-NIR and XRD) also used for analysis. Chromophore ions identified as being Cu(II) (for green and pale blue tesserae), Cu(0) (for red samples) and Co(II) (for deep blue samples). Opacifiers also analysed and found to be a lead – antimony compound (bindheimite) for yellow and green tesserae, CaSb2O6 in the pale blue samples and Ca2Sb2O7 in the blue samples. | 355 |
Various | Ancient glass from China | XRF; -; S | Portable EDXRF system developed and tested on ancient glass samples from Xinjiang province. Results compared with those obtained using PIXE. The principles, performance and analysis process of the equipment were all discussed. | 356 |
Various | Post medieval Tuscan glass finds | MS; ICP; LA | Results from the analyses were used to distinguish several compositional groups. These groups were consistent with those obtained using data from SEM-EDX for major elements. | 357 |
Various | Stained glass windows from St Anthony's Basilica, Padova, Italy | WDXRF; -; S EPMA | WDXRF analysis of very small fragments of glass taken from corners (where lead strips hold the glass in place). Therefore, no damage observed to the windows. Samples were ground and then fused at 1100 °C with lithium tetraborate ready for analysis. Precision for WDXRF was 0.6% for major and minor elements and approximately 3% for trace elements. The analytes Cl, S, Sb and Sn could not be determined using this methodology because of volatilisation problems. SEM-EDX and XRD also used for analysis. | 358 |
As discussed briefly above, slurry sampling has proven popular in this review period. One example, by Wang and co-workers, used slurry concentrations of nm-sized silicon carbide as high as 30% m/v and ICP-OES as a means of detection.362 After rigorous optimisation, the authors managed to maintain fluidity and stability of the very high suspended solids slurry by using 2% polyethylene imine at pH 4 as a dispersant. As with many other slurry-based techniques, calibration of the instrumental response was achieved using aqueous standards. The very high solids loading of the slurry and very low blank contamination levels enabled exceptionally low LOD to be obtained for Mn and Ti (2 ng g−1) through to Al (100 ng g−1). Another example of slurry preparation of ceramic materials was reported by Amberger and Broekaert.363 These authors determined trace elemental contamination in 1% m/v slurries of boron carbide ceramics with different size distribution using slurry sampling ETV-ICP-OES. The temperature programme used for the ETV process was optimised and the optimal vaporisation temperature was 2600 °C, which was held for 12 s. Doping of the 500 mL min−1 argon flow rate used to transport the vapour to the plasma with 6 mL min−1 Freon R12 (dichlorodifluoromethane) was found to enhance the Ti signal by a factor of 25. The authors set about determining why this should be by injecting 20 μL of slurry onto a L'Vov platform and then going through the temperature programme. Using TXRF of the L'Vov platform surface, the authors discovered that the amount of Ti vaporised increased from 12 to 97% when the Freon R12 was used compared with only argon. A number of analytes were determined (Al, Ca, Co, Cr, Cu, Fe, Mg, Mn, Na, Ni and Ti) and yielded LOD in the range 0.002–2 μg g−1. The authors applied their protocol to three boron carbide ceramics, three boron nitride powders, an aluminium oxide powder reference material (NIST 699) and a silicon carbide powder reference material (BAM S-003). Good agreement with certified values was obtained. A comparison was also made between the data obtained using the slurry sampling ETV-ICP-OES technique with Freon addition and those obtained using slurry sampling TXRF and a wet chemical digestion followed by ICP-OES analysis. The values obtained using the method developed were found to lie in the range 78–128% of the mean results of the other two techniques. The use of a fluorinated matrix modifier has also been reported by Souza and Oliveira who used slurry sampling ETAAS to determine Cr and Mn in alumina.364 These authors coated the graphite platform with niobium carbide (350 μg of niobium) to minimise the amount of interaction of the alumina with the graphite and then used 0.2 M sodium fluoride matrix modifier during the analysis itself to aid decomposition of the sample (and hence improve volatilisation of the analytes). The overall effect was to extend the lifetime of the graphite tube up to 150 firings. Under optimised conditions, a slurry of 50 mg of material suspended in 30 mL of 2% nitric acid could be analysed, although manual agitation of the slurry was required immediately prior to analysis to ensure a homogeneous suspension. An aliquot (20 μL) of slurry (or of calibration standard) was co-injected with 20 μL of the sodium fluoride solution and, after drying, the sample was charred at 1300 °C and then atomised at 2400 °C. The protocol was validated by the analysis of standard reference materials, where the results were found to be in good agreement with certified values. Limits of detection were 66 and 102 ng g−1 for Cr and Mn respectively. Alumina was also the sample matrix of choice for Lu and colleagues, who also used slurry sampling ETAAS for the analysis.365 Although no novel matrix modifiers were used, they succeeded with the analysis after optimisation of the temperature programme and of the lamp current used for the Smith-Hieftje background correction system. Calibration was again achieved using aqueous standards. The results obtained were compared with those obtained using a sample dissolution procedure (sulfuric acid in a PTFE pressure vessel) and were found to be in good agreement. Limits of detection were 0.66, 2.5 and 0.13 ng g−1 for Cu, Fe and Na respectively.
Laser ablation with ICP-MS detection as a means of determining N in different calcium nitrides has been reported by Bohme and co-workers.366 The materials Ca2N, alpha-Ca3N2 and beta-Ca3N2 were prepared using both 14N and 15N. It was found that the 14N signal was completely overwhelmed by N present in the atmosphere and could not be used for detection. However, the 15N produced a distinct signal above background levels and was therefore used during subsequent analyses. The authors described three different quantification strategies in detail. At the time of reporting, the precision and accuracy was approximately ± 5%, but there was hope that further improvements could be made. The LOD was reported as being 5 mg g−1 for 15N.
Two LIBS applications have been reported. One, by Pedarnig and co-workers, has already been discussed in detail in section 3.1.2 because it also reported the analysis of polymeric materials.250 The method reported the enhancement of the LIBS signal by using an RF-plasma jet to further excite the laser induced plasma. The other application, by Liu and colleagues, also reported an enhancement of the LIBS signal, this time using microwave assistance.367 Optimisation of the laser pulse irradiance, microwave duration and surrounding gas enabled an enhancement in signal from an alumina ceramic sample of up to 33 fold to be obtained when compared with an analysis when no microwave irradiation was used. Although the magnitude of signal enhancement was element dependent, it was found to be at a maximum when LA was induced at low irradiance on a large area and in an air atmosphere. Unsurprisingly, the enhancement was found to be greatest for analytes whose transitions had a low excitation energy. Microwave enhancement was less effective when nitrogen was used as the surrounding gas and the authors provided a discussion giving the reasons for this.
In one study, Sakaguchi et al. used SIMS to study O diffusion in magnesium doped zinc oxide ceramics.368 Results indicated that increasing concentrations of magnesium oxide led to increased O diffusion. Differences in the bulk diffusion coefficients when between 1.4 and 11.7% magnesium oxide were added amounted to two orders of magnitude. The grain boundary diffusion coefficients for O in an 11.7% magnesium doped material were approximately 10 times higher than for other samples. Another piece of work by the same research group reported the use of SIMS and a first principles study to detect Cr diffusion in alpha alumina.369 The SIMS experiment demonstrated that Cr diffusion in this material showed a crystallographic orientation dependence; i.e. that Cr diffusion was greater perpendicular to the 0001 direction than parallel to it. Theoretical calculations indicated that both Al and Cr diffusion proceeded via a vacancy rather than an interstitial site and that the crystallographic orientation dependence discussed above would be more profound for Cr than for Al.
A paper by Galindo and co-workers discussed the improvement of the oxidation resistance of aluminium chromium nitride coatings by the addition of a sub-surface titanium nitride layer.370 Oxidation degradation occurred through the diffusion of O into the material and the diffusion of Cr to the surface. The authors varied the depth of the embedded layer and the oxidation time and then used GD-OES to detect any changes in aluminium chromium nitride/titanium nitride depth composition profile and the surface oxidation stoichiometry. It was found that a titanium nitride layer of 300 nm thickness embedded 500 nm from the surface inhibited the inward diffusion of O and promoted the formation of aluminium surface layers. The use of GD-OES in conjunction with cross-sectional SEM proved to be a rapid and accurate technique for depth-profiling and for monitoring the changes occurring during oxidation.
Analyte | Matrix | Technique; Atomisation;Presentation | Comments | Reference |
---|---|---|---|---|
Pb | 19 Tang Sancai pottery glazes | MS; ICP; L | Pb isotope analysis using MC-ICP-MS enabled pottery from two different kilns to be identified. Data from Huanbao kiln indicated the Pb in the glaze came from northern China whereas glaze from Gongyi kiln came from the Yangtze region. | 374 |
Various | Guarani ethnic groups pottery | XRF; -; S | Samples from 13 sites analysed using XRF with an Am-241 source. Chemometric analysis (spidergrams) identified four different sets of samples according to areas of provenance. | 375 |
Various (13) | Historical pavement ceramics from Naples Province, Italy | OES; ICP; L | Atomic spectrometric, thermogravimetric and XRD analysis for the mineralogical composition undertaken. Results of analyses enabled origin of raw materials and provenance of ceramic (i.e. local or imported) to be ascertained. | 376 |
Various | Polychrome lead-glazed Portuguese Faiences | XRF; -; S | Non-commercial μ-EDXRF instrument used for in situ analysis of materials. Spot sizes of 100 pm in diameter could be analysed giving lateral resolution of a few tens of pm. Blue pigments had the elements As, Co, Fe and Mn associated with them whereas the yellow pigment was a combination of Pb and Sb. A statistical treatment was also undertaken to reveal groups of similarities. | 377 |
Various (17) | 44 pottery samples from Tell Jendares in Syria | XRF; -; S | Multivariate statistical analysis (cluster and factor analysis) undertaken on analytical data. Three distinct groups of samples identified through trace elemental composition. | 378 |
Various | Proto-porcelain of Yingguo Graveyard in Henna province | XRF; -; S | Both XRF and XRD used to analyse several sherds. Ca/Al and P/Al ratios enabled distinction between proto-porcelain of the Yingguo site from that produced in the South. | 379 |
Various | Bluish-white porcelain from Fanchang kiln, Anhui province | XRF; -; S | WDXRF used to analyse porcelain bodies from several dynasties. Chemical composition of major, minor and trace elements determined. INAA data supplemented study. | 380 |
Various | Cypriot Byzantine pottery | XRF; -; S | Samples (25) dating between 12th and 15th century analysed. Supplementary data obtained using SEM and XRD. All glazes were found to contain Pb and some also contained Sn. Decoration colours were formed from Co, Cu, Fe and Ni. Principal component analysis used to analyse analytical data. | 381 |
Various | Ancient Chinese Qingbai wares | OES; ICP; L | 28 Qingbai wares from different districts were analysed for major, minor and trace elements. Data for K and 12 other trace elements were found to yield provenance information. | 382 |
Various | Clay cuneiform tablets | XRF; -; S | Portable XRF instrument used to analyse tablets from Hattusa and from el Amarna in a non-destructive way. Results obtained were compared with those from a previous study that used INAA and optical mineralogy. | 383 |
Various | 29 Tang sancai sherds from Liquanfang site, Xi'an city | OES; ICP; LA | Both SEM-EDS and LA-ICP-OES were used. Ceramics with yellowish bodies were calcareous whilst those with red bodies were ferruginous clays. No other calcareous clays used for ceramics in Tang sancai bodies, implying manufacture was influenced by ceramic technology from near east or Central Asia. | 384 |
Various | 102 fragments of Marajoara ceramics | XRF; -; S | EDXRF data obtained then analysed using PCA to identify groups that exhibit similar characteristics. EDXRF system was a portable one that used an acquisition time of 600 s and a beam collimation of 2 mm. | 385 |
Various | Etruscan depurata pottery | XRF; -; S | Portable XRF instrument used to obtain analytical data and then dendrograms and PCA used to classify sample types. The samples had already been classified by archaeological, chemical and mineralogical examination. Analysis of polished areas of sample yielded the “correct” provenance results. Analytical data validated by analysis of SRM 679. | 386 |
Various (20) | Ancient potteries from Himera and Pestavecchia necropolis, Sicily | OES; ICP; L | Microwave digestion of samples prior to analysis. Cluster analysis, PCA and non-statistical analysis all used to treat dataset in an attempt to obtain provenance data. Three main groups identified plus a single “lonely” sample representing a potential fourth group. | 387 |
Various | Renaissance lustred majolica shards from two places in central Italy | OES; ICP; L | Differences in chemical and mineralogical composition between two similar potteries identified because one had much higher Ca content than the other. In addition, hierarchical cluster analysis and PCA used to further discriminate between the two types. Crystallographic analysis using XRD also performed. | 388 |
Various | Red-slipped pottery (2nd to 1st century BC) from Greece | XRF; -; S | Analysis using EDXRF to obtain data and PCA used to establish whether the ceramic was produced locally or imported. Crystallographic analysis using XRD indicated firing temperatures of typically 800–1000 °C. Technique of SEM-EDX also used to analyse the surface. | 389 |
Various | Roman lead glazed pottery | OES; ICP; L TIMS EPMA | Ceramic bodies analysed using ICP-OES after grinding of samples to remove glazes etc. Isotopic signature of Pb in glazes determined using TIMS. Glazing techniques employed were examined using EPMA. Two methods identified. One used lead oxide applied to the clay body and the other used a mixture of lead oxide and quartz. | 390 |
Other applications are more novel in terms of analytical chemistry and they will be discussed in more detail here. A review of LIBS analysis for cultural heritage samples as well as for space exploration was presented by Gaudiuso et al.359 The review, containing 154 references, emphasised the technique's conceptual simplicity and versatility. It also discussed its ability to obtain quantitative and qualitative data by constructing calibration curves or with standard free analysis. The use of portable XRF instruments for the analysis of cultural heritage has been overviewed by Cesareo.371 The paper contained only seven references but discussed how such instruments can be used for the analysis of porcelains, ceramics, stones and a variety of metallic objects in a non-invasive and non-destructive way.
A paper by Zhu et al. reported the development of 17 types of ceramic standard sample that can be used for EPMA and XRF analysis (in Chinese).372 Some of the samples demonstrated high density, low water absorption and good homogeneity of elemental distribution. In addition, they had a similar phase structure to many ancient ceramics and, therefore, the authors concluded that they would be the ideal reference materials for the identification of ancient ceramics.
The combined use of a portable XRF instrument and synchrotron radiation XAS for the non-destructive analysis of Sicilian ceramic fragments dating from the twelfth and thirteenth centuries was reported by Bardelli and co-workers.373 The XRF instrument enabled elemental and spatially resolved data for both major and minor elements to be obtained. This enabled the authors to assign Fe and Mn as the main components of the colouring agent. The XAS was used to obtain more detailed information which the authors then analysed using principal component analysis (PCA) and least squares fitting. The speciation thus determined enabled the authors to ascertain that the pigment was umber, a class of brownish pigments that are characterised by a mixture of hydrated Fe and Mn oxides.
The use of rhodium porphyrin complexes as catalysts for increased carbon monoxide oxidation efficiency has been reported by Takeda and Yamazaki.392 Electrospray ionisation mass spectrometry (ESI-MS) yielded structural information on the macromolecules produced during the cell reactions. The coordination of the carbon monoxide was determined using tandem ESI-MS/MS. This paper described the use of capillary electrophoresis electrospray ionisation mass spectrometry (CE-ESI-MS) as an alternative means of analysis. A special capillary cartridge was used to insert the capillary into the ESI probe. Sample components were separated by differences in electrophoretic mobility before detection using MS. Although found to be an effective method of analysis of the macromolecules, further work was needed to confirm carbon monoxide coordination.
Improvement of the carbon monoxide tolerance of methanol fuelled cell catalysts by metal doping and modification of support materials has been a significant research area. Investigations have included platinum-ruthenium nanoclusters supported on graphite nanofibres,393 platinum-niobium oxide supported on carbon,394 and platinum-tin supported on carbon.395 Particle size and morphology of catalysts were characterised using XRD and TEM. Metal loading was determined using ICP-OES and ICP-MS, whilst surface characteristics were measured using XPS. Similar studies were also reported for hydrogen fuelled cells with platinum supported on carbon spheres,396 multi-walled carbon nano-tubes397 and mesoporous carbon398 being proposed. Again, the characterisation techniques of choice were XRD, TEM/SEM and ICP-OES/ICP-MS.
Bozzini et al. published an interesting paper on the use of synchrotron-based photoelectron microscopy and high lateral-resolution XPS for the mapping of Pt location and chemical state in both used and unused fuel cell membrane-electrode assemblies.399 The synchrotron X-ray beam was focused to a 150 nm spot size using Fresnel plate optics. The membrane-electrode assemblies consisted of a Pt catalyst layer sandwiched between Naffion foil and a gas diffusion layer. It was this layer that was peeled open to gain access to the catalyst for analysis. The analysis found that catalyst aging was linked to the agglomeration of Pt nanoparticles in the catalyst layer. Further to this they also found that Pt had migrated to the gas diffusion layer where it was found to be fixed on the originally Pt- free carbon paper fibres. Agglomeration was detected on the basis of shifts of the Pt 4f7/2 level.
Two separate papers published by Carvalho et al. investigated the effect of the depth-profile of Cu418 and Co419 doped on titanium dioxide thin films. Samples were prepared by DC sputtering of pure metal on to a titanium dioxide thin film prepared in the same way. Samples were annealed at either 100 °C or 400 °C before being chemically etched to remove excess surface metal. The extent of migration of the dopant into the titanium dioxide was measured using Grazing Incidence X-Ray Fluorescence (GIXRF) spectroscopy using synchrotron radiation. It was found that Co migrated easily into the titanium dioxide structure at room temperature (a distance of 19 nm) and annealing had little additional effect (a distance of 29 nm). On the other hand, Cu needed the annealing process for migration, since it migrated only 7 nm at room temperature but had travelled a distance of 31 nm after annealing at 400 °C. Photocatalytic performance was measured by methylene blue discolouration and ESI-MS.
Iron(III) doped titanium dioxide heterogeneous photocatalysts, for the decomposition of phenol, were prepared by Bajnoczi et al. using both a sol–gel method and a flame hydrolysis method.420 It was found that the undoped flame hydrolysis prepared sample performed much better than the undoped sol–gel prepared sample. However, for the sol–gel prepared samples, photocatalytic activity first increased with increasing Fe(III) concentration, reaching a maximum, whilst Fe(III) doped flame hydrolysis prepared samples significantly decreased in activity, even at the smallest Fe(III) concentration. Bulk characterisation using AAS, diffuse reflectance spectroscopy (DRS), XRD and TEM could not explain the differences. Mossbauer and XPS measurements showed that the local structure of the Fe(III) was different in the two series. From the pre-edge and XANES region it was found that Fe(III) was present in a distorted octahedral environment in both series. However, in the flame hydrolysis samples, the extent of this distortion was much more significant. Information obtained in the EXAFS region indicated the structure of Fe2O3 was much less ordered in the sol–gel series than the flame hydrolysis series and vacancies were much more abundant. The authors found that an optimum catalytic performance required a somewhat distorted geometry around the Fe(III) together with as many vacancies in the dopant as possible.
The characterisation of reaction products and intermediates is an important step in the production of new catalyst materials, because it enables the deduction of possible reaction pathways. The use of solid phase micro-extraction (SPME)-GC-MS to measure the reaction pathway for the degradation of 3-chloropyridine using titania nano-photocatalysts was reported by Colmenares and co-workers.421 Different fibres were evaluated as absorbents (polyacrylate, polydimethylsiloxane/divinylbenzene, etc.).
The best one had polyacrylate as the stationary phase. Before use, the fibre was conditioned at 300 °C for 2 h. It was then introduced into a vial containing samples and allowed to extract for 30 min at 40 °C. Masses scanned were in the 30–300 atomic mass unit range. The analysis showed that two parallel degradation pathways existed, involving hydroxylation and ring opening prior to, or after, the release of chloride ions. Bulk catalyst properties were measured using ICP-MS, XRD, UV-VIS and TEM. The excellent sensitivity and surface specificity of TOF-SIMS analysis has been exploited for the surface characterisation of a Pd/TiO2 catalyst.422 Both elemental and molecular species were identified in fresh, de-activated and regenerated catalysts giving valuable information on the use of the catalyst in a hydrodechlorination process.
Moving away from titania, Cu2+ substituted zinc-gallates were produced using a microwave-hydrothermal synthesis method.423 The technique of EXAFS was used to determine that Zn was coordinated mainly in the tetrahedral sites of the spinel lattice, whereas Cu was located in the octahedral sites. It was thought that such a catalyst could be useful for photocatalytic splitting of water. It was suggested by Lin and Chang, that the use of gold nanoparticles deposited on potassium titanoniobate are useful for photocatalytic water splitting.424 Several catalysts were prepared by deposition-precipitation, conventional impregnation or photodeposition methods and were then characterised using XRD, ICP-MS, UV-VIS and TEM. A further method for photocatalytic water splitting using ZnmIn2S3+m (where m = 1–5 integer) was reported by Shen and colleagues.425 Characterisation of the materials was achieved by using XRD, UV-Vis together with XRF and SEM.
Nanocomposites of Zn 1−X Cu X S and Zn 1−X Ni X S were synthesised as photocatalysts for Congo red degradation.426 Characterisation was achieved using UV-VIS, AAS, XRD, TEM and Brunauer Emmett and Teller (BET) measurements. It was found that a blue shift in the band-gap of zinc sulfide occurred with incorporation of Cu and Ni. Nanocomposites of formulae Zn0.94Ni0.06S and Zn0.90Cu0.10S showed the highest activity. Zinc oxide co-doped with Fe and Ni was reported for the photocatalytic degradation of methyl orange in a paper by Fu and colleagues.427 The usual suite of techniques was used to characterise these materials.
A thorough characterisation of visible light active block copolymer-templated bismuth ferrite (BiFeO3) nanoarchitecture films was presented by Reitz et al.428 The team used SEM to determine the morphology of the materials whereas grazing incidence small-angle X-ray scattering (GISAXS) and conventional SAXS were used to determine the nanostructure. Further crystalline information was obtained by employing the technique of wide angle X-ray spectroscopy (WAXS). Data obtained using XPS was used to determine the Fe environment and TOF-SIMS was used to determine the distribution of Fe and Bi throughout the films. It was found that templated bismuth ferrite materials were highly active for the photocatalytic degradation of rhodamine B. Thin films of modified bismuth niobate materials (Bi2MNbO7 where M = Al, In, Ga or Fe) were analysed for their photocatalytic activity toward the degradation of methyl orange.429 Crystal structure and elemental analysis were investigated using XRD and EDXRF. The band gap energy was estimated using UV-VIS to be M = Fe(2.47eV), Ga(2.67eV), Al(2.79eV), In(3.01eV) and the reaction time t(1/2) were M = Ga(239min), Al(278min), In(296min), Fe(319min). These results indicated that a slight band gap narrowing had a positive effect on catalytic properties but those further than 2.5eV have a detrimental effect.
Mesoporous materials continued to be a popular research area because of their large surface area and structural support. Transition metal loaded materials have been reported for a wide range of oxidative reactions, including gas phase oxidation of toluene and ethyl acetate using Co-loaded zeolite sieve ZSM-5,433 solvent free selective oxidation of cyclohexane over transition metal incorporated hexagonal mesoporous silica molecular sieve HMS434 and carbon monoxide oxidation by the mesoporous silica SBA-15 supported nano-gold catalysts (in Chinese).435 A range of analytical techniques, namely XRD, TEM, SEM, XPS, BET, XRF and ICP-OES were typically used for characterisation purposes. Schiff base catalysts are conventionally supported on polymeric materials such as resin, however, these can be vulnerable to attack from some chemicals and solvents. To overcome this, Yang et al. proposed SBA-15 supported Cu(II) and V(IV) Schiff base catalysts for the oxidation of styrene.436 The authors used XRD and TEM to confirm that the support structure remained intact after introduction of the metal complexes. Bulk elemental analysis by ICP-OES estimated that the Cu content was 0.18 mmol g−1, whilst XPS estimated the surface concentration to be 0.1 mmol g−1, suggesting the majority of the Cu was on the surface. The V analogue, on the other hand, had a 0.312 mmol g−1 bulk concentration with a negligible surface concentration, suggesting the V filled the internal structure. It was found that the V catalyst had superior selectivity for styrene oxidation. The same authors also reported a similar study using Cu-modified salicylaldimine immobilised on SBA-15.437 Styrene oxidation was also achieved using a La-doped KIT-6 mesoporous material.438 Using XRD it was shown that La had been substituted into the crystal structure, rather than grafted onto the surface. The use of XPS established the La oxidation state. A biporous (micro- and mesoporous) composite material was prepared by He and colleagues by mixing ZSM-5 with colloidal MCM-48.439 Studies of this material using XRD and TEM showed that MCM-48 crystals grew onto the ZSM-5 support, giving rise to a biporous material. This was then loaded with Pd and the resulting material showed superior activity for benzene oxidation. Increased activity was thought to occur because of increased support activity and Pd dispersion. The authors also thought that these materials could become important materials for eliminating volatile organic compounds in air. The removal of volatile organic compounds has been attracting increased attention because of their toxicity and involvement in the formation of smog. A Au-modified ceria catalyst440 and a Mn–Ce mixed oxide catalyst441 have also been reported for the oxidation of these compounds. In both papers, the usual suite of characterisation techniques was used.
The selective oxidation of carbon monoxide in the hydrogen streams of PEMFCs is a popular way of preventing carbon monoxide poisoning of the anode material, which is one of the main causes of deactivation in fuel cell technology. A Ru/alumina catalyst was found to be effective for this reaction, even at low temperatures.442 The study involved the testing of various Ru-loaded alumina phases for carbon monoxide oxidation activity and characterisation of the resulting materials using XRD, TPD, temperature programmed reduction (TPR) and temperature programmed oxidation (TPO). The Ru/α-Al2O3 was found to be the most active catalyst tested, because it had a zero pore structure, allowing a high dispersion of Ru metal over the surface. Silver-modified OMS-2 molecular sieves443 and Fe or Au modified ceria444 were also found to be effective for this application. The catalytic wet oxidation of cosmetic wastewaters using Fe/γ-Al2O3 was evaluated by Bautista and colleagues.445 X-ray diffraction analysis of the catalyst showed two crystalline phases, corresponding to γ-Al2O3 and hematite (Fe2O3). Spectra obtained by using XPS confirmed that the main Fe species on the surface was Fe2O3 and total Fe content was measured using TXRF. Mossbauer analysis of samples calcined at different temperatures revealed a higher content of Fe2O3 nanoparticles at lower temperatures which corresponded with higher volatile organic compound oxidation activity.
Taking into account the volume in circulation throughout the world, and the significant concentrations in them, three way catalysts are seen as an important secondary raw material in the production of platinum group metals (PGM). Owing to the high price of PGM, analytical control of the extraction process is critical. Dal'nova and colleagues presented two investigations for the accurate determination of PGM on three way catalysts by ETAAS. The first highlighted the matrix effects (spectral and non-spectral) caused by the common non-precious metals Al, Ce, Fe, Mg, Ti and Zr.450 Direct analysis of spent three way catalyst was not possible, because of a decrease in analyte signal with increasing concentration of matrix elements. In addition, spectral interferences still remained a problem. It was thought that the formation of temperature-stable, binary alloys was the cause. A sample digestion and dilution method was proposed as an alternative. The second paper, by the same authors, proposed the use of a complex-forming thio-amide sorbent material.451 The selectivity of the sorbent material allowed for the extraction of analytes from the matrix elements discussed previously, hence, eliminating spectral interferences.
The consumption of petroleum products for energy and chemicals is at a peak rate. Although new sources continue to emerge, diminishing petroleum reserves necessitates the development of fuels from renewable resources. Biodiesel is considered by many to be an attractive alternative, hence, there has been an increase in the development of catalysts for its production. The use of hetropolyacids immobilised on mesoporous silica for esterification of fatty acids to biodiesel was reported by Tropecelo and colleagues.452 The study focused on tungstophosphoric, molybdophosphoric and tungstosilicic acids immobilised on SBA-15. The loadings of Mo and W were determined using ICP-OES, after digestion of the materials with a mixture of a 1:1 mixture of sulfuric and hydrofluoric acids. Confirmation that the SBA-15 mesostructure remained intact after immobilisation was achieved using XRD. Potentiometric titration curves, with n-butylamine, were obtained to measure catalyst acidity. It was found that tungstophosphoric acid had the strongest acidity, which increased with loading levels. Catalytic studies for the esterification of palmitic acid with methanol revealed that this was also the best catalyst. Macario et al. investigated a wide range of acid, weak acid and acid–base materials for the transesterification of triglycerides to biodiesel fuel.453 They found that microporous acid catalysts were not suitable because of diffusion limitations of reactants inside the micropores. However, catalysts with acid and base sites, such as potassium modified ITQ-6, were found to have high conversion rates. Characterisation was performed using the usual suite of techniques. Response surface methodology was reported in a number of papers to determine optimum reaction conditions.88,454 However, characterisation studies were brief.
AAS | Atomic absorption spectrometry |
AED | Atomic emission detection |
AES | Atomic emission spectrometry |
AFM | Atomic force microscopy |
AFS | Atomic fluorescence spectrometry |
AMS | Accelerator mass spectrometry |
ANOVA | Analysis of variance |
ANSI | American National Standards Institute |
APDC | Ammonium pyrrolidine dithiocarbamate |
ARXPS | Angle resolved X-ray photoelectron spectroscopy |
ATR | Attenuated total reflection |
BCR | Bureau of Community Reference |
BET | Brunauer, Emmett and Teller |
CA | Cluster analysis |
CCA | Chromated copper arsenate |
CCD | Charge coupled device |
CE-ESI-MS | Capillary electrophoresis-electrospray ionisation-mass spectrometry |
CFP | Circulating fluidised bed |
CMOS | Complementary metal oxide semiconductor |
CPE | Cloud point extraction |
CRM | Certified reference material |
CVAAS | Cold vapour atomic absorption spectrometry |
CV-AFS | Cold vapour atomic fluorescence spectrometry |
CWADP | Chemical warfare agent degradation product |
CZE | Capillary zone electrophoresis |
DCP | Direct current plasma |
DESI | Desorption electrospray ionisation |
DMFC | Direct methanol fuel cells |
d-DIHEN | Demountable direct injection high efficiency nebuliser |
DPT | Diphenyl tin |
DRIFTS | Diffuse reflectance infra-red Fourier transform spectroscopy |
DRS | Diffuse reflectance spectroscopy |
DTA | Differential thermal analysis |
DTPA | Diethylenetriaminepentaacetic acid |
ECNI | Electron capture negative ionisation |
EDS | Energy dispersive spectrometry |
EDX | Energy dispersive X-ray analysis |
EDXRF | Energy dispersive X-ray fluorescence |
EELS | Electron energy loss spectrometry |
EI | Electron ionisation |
ELCAD | Electrolyte cathode discharge |
EPMA | Electron probe microanalysis |
ERDA | Elastic recoil detection analysis |
ESI | Electrospray ionisation |
ESI-MS | Electrospray ionisation-mass spectrometry |
ESP | Electrostatic precipitator |
ESR | Electron spin resonance |
ETAAS | Electrothermal atomic absorption spectrometry |
ETV | Electrothermal vaporisation |
EXAFS | Extended X-ray absorption fine structure |
FAAS | Flame atomic absorption spectrometry |
FF | Fast flow |
FFF | Field flow fractionation |
FI | Flow injection |
FIB | Focused ion beam |
FLIM | Fluorescence lifetime imaging |
FTA | Fission track analysis |
FTIR | Fourier transform-infra red |
FTS | Fischer–Tropsch synthesis |
FYXAS | Fluorescence yield X-ray absorption spectroscopy |
GC | Gas chromatography |
GC-AED | Gas chromatography atomic emission detection |
GD | Glow discharge |
GEXRF | Grazing exit X-ray fluorescence |
GFAAS | Graphite furnace atomic absorption spectrometry |
GISAXS | Grazing incidence small angle X-ray spectrometry |
GIXRD | Grazing incidence XRD |
GIXRF | Grazing incident XRF |
GI-XSW | Grazing incidence X-ray standing waves |
GSIMS | Gentle secondary ion mass spectrometry |
GXRD | Glancing angle X-ray diffraction |
HGAAS | Hydride generation atomic absorption spectrometry |
HPLC | High performance liquid chromatography |
HRTEM | High resolution transmission electron microscopy |
HR-CS-AAS | High resolution continuum source atomic absorption spectrometry |
HRXRD | High resolution X-ray diffraction |
IBMK | Isobutylmethyl ketone |
ICCD | Intensified charge coupled device |
ICP | Inductively coupled plasma |
ID | Isotope dilution |
IDMS | Isotope dilution mass spectrometry |
INAA | Instrumental neutron activation analysis |
IRMS | Isotope ratio mass spectrometry |
ISO | International organisation for standardisation |
LA | Laser ablation |
LC-MS | Liquid chromatography-mass spectrometry |
LDA | Linear discriminant analysis |
LED | Light emitting diode |
LEIS | Low energy ion scattering |
LEXES | Low energy electron induced X-ray emission spectroscopy |
LIBS | Laser induced breakdown spectroscopy |
LIDAR | Light detection and ranging |
LIF | Laser induced fluorescence |
LIF-ASE | laser induced fluorescence-amplified spontaneous emission |
LIPS | Laser induced plasma spectroscopy |
LOD | Limit of detection |
LOQ | Limit of quantitation |
LTE | Local thermodynamic equilibrium |
MALDI | Matrix assisted laser desorption/ionisation |
MAS | Magic angle spinning |
MC-ICP-MS | Multiple collector ICP-MS |
MEEKC | Microemulsion electrokinetic chromatography |
MFM | Magnetic force microscopy |
MIC | Microwave induced combustion |
MIM | Metal insulator metal |
MOCVD | Metal organic chemical vapour deposition |
MPT | Monophenyl tin |
MS | Mass spectrometry |
MWDS | Microwave desolvation system |
Nd:YAG | Neodymium: yttrium aluminium garnet |
NEXAFS | Near edge X-ray absorption fine structure |
NIR | Near infra red |
NIST | National Institute of Standards and Technology |
NMR | Nuclear magnetic resonance |
NRRA | Nuclear resonant reaction analysis |
OES | Optical emission spectrometry |
OSL | Optical stimulated luminescence |
PB/HC-OES | Particle beam/hollow cathode optical emission spectrometry |
PBDE | Polybrominated diphenyl ethers |
PCA | Principal component analysis |
PD-MS | particle desorption mass spectrometry |
PEMFC | Proton exchange membrane fuel cells |
PET | Polyethyleneterephthalate |
PGM | Platinum group metals |
PIGE | Proton induced gamma ray emission |
PIXE | Particle induced X-ray emission |
PLS | Partial Least Squares |
PLS-DA | Partial least squares discriminant analysis |
PTFE | Polytetrafluoroethylene |
PVC | Polyvinyl chloride |
PVD | Physical vapour deposition |
RBS | Rutherford backscattering spectrometry |
REE | Rare earth elements |
RF | Radio frequency |
RI | Refractive index |
RMM | Relative molecular mass |
RoHS | Restriction of hazardous substances |
RSD | Relative standard deviation |
S | Solid |
SAXS | Small angle X-ray spectroscopy |
SEM | Scanning electron microscopy |
SBR | Signal to background ratio |
SF | Sector field |
SIBS | Spark induced breakdown spectrometry |
SIMCA | Soft independent modelling of class analogy |
SIMS | Secondary ion mass spectrometry |
SNMS | Secondary neutral mass spectrometry |
SPE | Solid phase extraction |
SPM | Scanning probe microscopy |
SPME | Solid phase micro-extraction |
SRD | Synchrotron radiation diffraction |
SRM | Standard reference material |
SRXRF | Synchrotron radiation induced X-ray fluorescence |
TEM | Transmission electron microscopy |
TGA | Thermogravimetric analysis |
TIMS | Thermal ionisation mass spectrometry |
TMAH | Tetramethyl ammonium hydroxide |
TMPP | Tris(2,4,6-trimethoxyphenyl) phosphonium propylamine |
TOC | Total Organic Carbon |
TOF | Time of flight |
TPD | Temperature programmed desorption |
TPO | Temperature programmed oxidation |
TPR | Temperature programmed reduction |
TPT | Triphenyl tin |
TSL | Thermally stimulated luminescence |
TRXPS | Total reflection X-ray photoelectric spectroscopy |
TXRF | Total reflection X-ray fluorescence spectrometry |
USN | Ultrasonic nebuliser |
UV | Ultra violet |
VIS | Visible |
VPD-DC | Vapour phase decomposition-droplet collection |
VUV | Vacuum ultra violet |
WAXS | Wide angle X-ray spectroscopy |
WDS | Wavelength dispersive spectrometry |
WDXRF | Wavelength dispersive X-ray fluorescence |
WEEE | Waste from Electrical and Electronic Equipment |
XAFS | X-ray absorption fine structure |
XANES | X-ray absorption near edge spectrometry |
XAS | X-ray absorption spectroscopy |
XBIC | X-ray beam induced current |
XPS | X-ray photoelectron spectroscopy |
XRD | X-ray diffraction |
XRF | X-ray fluorescence |
XRR | X-ray reflectometry |
XSW | X-ray standing waves |
This journal is © The Royal Society of Chemistry 2011 |