1 Introduction

Plutonium (Pu) is an anthropogenic element, ubiquitous in the environment as a result of fallout from nuclear weapons testing in the 1950s and 60s, nuclear power plant accidents and marine discharges of reprocessing waste [1, 2]. However, distribution of Pu from the latter is more localised, owing to 239+240Pu being contained within the non-volatile fraction of nuclear fuel debris [3, 4]. A total of 520 atmospheric nuclear weapons tests were conducted worldwide between 1945–1980 [5]. Due to their high radiotoxicity and long retention times, 239Pu and 240Pu isotopes are considered important transuranic nuclides in the environment with half-lives of 24,110 and 6561 years, respectively [65]. Another benefit of this strip** process is that the levels of U purification prior to analysis are lower than other MS techniques, allowing for the simplification of the radiochemical procedures prior to analysis [66].

2.3.2 Thermal Ionisation Mass Spectrometry and Resonance Ionisation Mass Spectrometry

In addition to AMS there are also some alternative mass spectroscopy techniques that are highly sensitive for the detection of Pu isotopes including Thermal Ionisation Mass Spectrometry (TIMS) and Resonance Ionisation Mass Spectrometry (RIMS). The TIMS method has a higher sensitivity for 239Pu and 240Pu than ICP-MS and interferences due to UH and UH2 are less significant. This means that TIMS has become the method of choice for measuring isotope ratios with precision as low as 0.002% [67]. However, TIMS is limited by the relatively high cost of analytical facilities and the extensive sample preparation prior to analysis to produce a thin filament source, taking days to weeks of dissolution and separation steps [68]. The method used in RIMS, employs tuned laser beams for the selective excitation of the Pu atoms. It is both highly sensitive and selective for the measurement of 239Pu with detection of 239Pu activities as low as 100 atoms per sample, equivalent to of 0.1 nBq. This method however, is only available at specialist laboratories worldwide [4].

2.3.3 Inductively Coupled Plasma Mass Spectrometry

An alternative to mass spectrometry methods described above is Inductively Coupled Plasma Mass Spectrometry (ICP-MS). This method has grown in popularity over the past 10 years which is shown by the increasing number of publications using the method and it has become a widely used technique for the detection of Pu isotopes due to its high sensitivity, short analytical times, and relatively simple operation. However, the method can be hindered by the formation of interferences due to polyatomic species, formed from matrix elements and plasma gases. These polyatomic interferences require removal as they have the same integer mass-to-charge ratios as the analyte of interest, leading to false detection or overestimation of results [14, 69]. For Pu isotopes, the major interfering ions are a consequence of the presence of 238U which is ubiquitous in the environment. Uranium hydrides 238UH+ and 238UH2+ cannot be resolved from 239Pu+ and 240Pu+ making analysis of these isotopes a challenge [65]. With the concentration of 238U in environmental samples being up to 6–9 orders of magnitude higher than that of Pu, another issue for the analysis of 239Pu as a result, in quadrupole ICP-MS is the peak tailing from 238U [70]. Therefore, low-resolution ICP-MS cannot always reliably determine 239Pu, relying heavily on the chemical purification steps prior to analysis, which are used to remove U isotopes from the matrix [71]. However, these procedures also bring additional U into the final sample solutions through atmospheric contamination of the glassware, and reagents [72]. There are also other minor polyatomic interferences which need to be taken into consideration such as plasma gas induced Hg and Pb interferences (Table 4).

Table 4 Polyatomic interferences for Pu isotopes using ICP-MS [7].

$${\mathrm{U}}^{+}+{\mathrm{CO}}_{2}\to \mathrm{U}{{\mathrm{O}}_{n}}^{+}$$
$${\mathrm{UH}}^{+}+{\mathrm{CO}}_{2}\to \mathrm{U}{{\mathrm{O}}_{n}}^{+} \quad (n = 1-2)$$

Hou et al. [49] reported that the optimal conditions to eliminate U interferences was 1.2 mL min−1 CO2 and 8 mL min−1 He, which reduced overall interference on 239Pu to < 1 × 10–8. However, it was reported that although this high flow rate is optimal for the removal of interfering ions, increasing flow rates above 1.2 mL min−1 of CO2 results in declining intensity of PuO+ signal. This was attributed to the increased production of PuO2+ within the collision cell [7]. Similar results were reported by Childs et al. [77] where significant U interferences were observed when comparing a U spiked Pu standard with an un-spiked standard; therefore it was deemed that Pu quantification was not possible using high flow rates of CO2 [77]. As the m/z ratio for PuO2+  > 271 is beyond the mass range for older ICP-MS/MS instruments, the loss of Pu signal measured at the Pu+ mass rather than shifted to an oxide form can have a negative impact on the measurement sensitivity. User requirements, particularly the nuclear industry, for analysing heavy elements in mass shift modes has meant that new instruments such as the Agilent 8900 have an extended m/z detection range up to 275, allowing for the detection of the mass shifted Pu isotope [73]. This highlights a need for manufacturers to extend the m/z range in future instruments to improve reaction cell chemistry and therefore allow for greater research into elements with complex interferences, further reducing detection limits and increasing sensitivity.

In addition to NH3 and CO2, O2 has also been used as a reaction gas with the Pu+ ion readily converted to both PuO+ and PuO2+ [73]. Of these two ions the favoured one for analysis is PuO2+ as it is subject to lesser interference than PuO+ which experiences dominant interference from uranium oxides, 238U16O+, 238U16O1H+ and 238U16O1H2+ for the measurement of 239Pu16O+ and 240Pu16O+ ions, causing a less efficient elimination of the uranium interference [7].

$${\mathrm{U}}^{+}+{\mathrm{O}}_{2}\to \mathrm{U}{{\mathrm{O}}_{n}}^{+}$$
$${\mathrm{UH}}^{+}+{\mathrm{O}}_{2}\to \mathrm{U}{{\mathrm{O}}_{n}}^{+} \quad (n = 1-2)$$

Zhang et al. [28] found that both 238U+ and 238U1H+ preferably reacted with O2 to form 238U16O2+ and therefore the interference was significantly reduced. The optimal conditions in order to observe maximal sensitivity of 242Pu+ (880 Mcps (mg L−1)−1) at m/z 274 (PuO2+) was obtained using 0.09 mL min−1 O2 as a reaction gas and 12 mL min−1 He [73]. The Pu+ signal decreases exponentially by more than 600 times when using O2/He gas mode as opposed to He only mode and this can be attributed to the formation of PuO2+ when subjected to relatively high O2 levels in the reaction cell. The use of this reaction gas is however limited to detectors with m/z reaching > 271. The reaction mechanism for the removal of uranium interferences can be seen in Fig. 1. It should be noted that there may still be some tailing of the 238U16O2+ on to 239Pu16O2+.

Table 5 summarises the detection limits achieved for ICP-MS/MS analysis using reaction gasses. With older ICP-MS/MS instruments being limited to detect m/z ratios no greater than > 271 one of the most commonly used reaction gases reported in ICP-MS/MS has been NH3. However, with the need for safe gas handling due to the corrosive nature of NH3 and the availability of quadrupole systems capable of > 271 amu, alternative methods are beginning to be favoured. With recent advancements in ICP-MS/MS technology allowing for m/z ratios > 271 to be detected, oxygen gas presents an exciting development in the detection of Pu isotopes in the presence of U in samples with detection limits exceeding that of NH3 and CO2.

2.3.6 Sector Field Inductively Coupled Plasma Mass Spectrometry

Limitations in the mass resolution of quadrupole ICP-MS has led to the development of high-resolution mass spectrometers. Sector field ICP-MS (SF-ICP-MS) which is based on the magnetic field approach and uses double focusing, to improve the mass resolution of ion peaks [81, 82]. This is achieved using an electrostatic analyser (ESA) before or after the magnetic field before passing the sample through an exit slit to filter the isotopes. Consequently, compared to a quadrupole system either an improvement in selectivity in high resolution mode or an improvement in the sensitivity as well as a reduction of the noise level can be achieved in low resolution mode (similar to quadrupole); this results in low achievable detection limits in the pg kg−1 range [83,84,85]. Another advantage of SF-ICP-MS over traditional ICP-MS is the ability to measure the signals on flat-topped peaks at lower resolutions. This offers an improvement in the measurement of isotope ratio precision over quadrupole based ICP-MS; however, it is important to note that precision is reduced with increasing resolution due to the deterioration of peak shape and is still poorer than that of MC-ICP-MS, where true simultaneous ratio measurements are made. Similarly to traditional quadrupole ICP-MS, SF-ICP-MS requires a high level of decontamination prior to analysis to remove interferences from UH+ as even high resolution mode is insufficient to fully remove this interference [86, 87]. This alongside the relatively higher cost of instrumentation and subsequently less common availability, make SF-ICP-MS a less attractive method for the determination of Pu in soil erosion studies. However, SF-ICP-MS is an more appropriate option in cases where a higher degree of specificity is required (e.g. forensic identification of Pu source using isotope ratios) compared to the requirement for soil erosion studies [83, 88].

2.3.7 Multi Collector Inductively Coupled Plasma Mass Spectrometry

Multi-collector ICP-MS (MC-ICP-MS) is based on the simultaneous detection of isotopes, eliminating classical sources of uncertainty from the sequential scanning used in ICP-MS [89]. Typically, MC-ICP-MS instruments will have up to nine faraday cages making up the detection assembly and newer instruments make use of ion counting systems to improve the abundance selectivity. Therefore, MC-ICP-MS can be used for measuring isotope compositions with both high precision and accuracy, and has the advantage of a high ionisation efficiency in comparison to the TIMS, allowing for a larger theoretical mass range of isotopes to be measured [81]. Similarly, to SF-ICP-MS, this analysis method has a requirement for the removal of UH+ via extensive separation prior to analysis and one of the challenges which must be overcome using MC-ICP-MS is the limited ‘practical’ mass range—needing repeat analyses to cover broad mass range, hence longer analysis time compared to ICP-MS/MS, limiting sample throughput. A consequence of this is the need to select an internal standard which falls into the mass range which is usually limited between 10% and 30% [90, 91]. Similarly to the SF-ICP-MS instrumentation, MC-ICP-MS is relatively more expensive than ICP-MS/MS, therefore with less availability and slower sample throughput is not suitable to soil erosion studies which need quick analysis of large quantity of samples.

2.3.8 Time of Flight Inductively Coupled Plasma Mass Spectrometry

An alternative analysis method is time of flight ICP-MS (ICP-TOF–MS). This technique pushes a packet of sample ions from the ICP into a ion flight tube, accelerates them and then separates the ions of different mass to charge ratio by their drift time [92]. Counting of the ions proceeds in a temporal succession on a microsecond time scale and because the packet of ions was sampled at the same time from the ICP this method of detection is essentially simultaneous [93]. This gives it an advantage of requiring a low sample volume and quick analysis time. However, prior to analysis a high degree of separation is required, and selectivity is similar to that of traditional quadrupole ICP-MS. Although this method does not offer advantages over ICP-MS/MS in terms of analysis for the purpose of soil erosion measurement, it does have a high potential to be used alongside laser ablation for high resolution analysis of impurities in nuclear fuels, nanomaterials and biological matrices [94,95,96,97].

3 Discussion

This review summarised the advancements of Pu isotope analysis over the past 20 years, by identifying common methods reported for the determination of Pu isotopes in environmental samples and comparing these methods for their respective advantages for the measurement of soil redistribution rates. A future challenge which must be addressed is the need for ultra-trace analysis of Pu isotopes in soils, so that Pu can be used as an effective tool for the quantification of soil erosion in areas where global fallout is minimal (tropics). Wilken et al. [39] demonstrated the applicability of using 239+240Pu in tropical Africa for the determination of soil erosion for study sites along the East African Rift Valley system. Despite lower global fallout in the tropics, a relatively high 239+240Pu baseline inventory was found at the reference sites. Cultivated sites showed signs of substantial soil erosion and sedimentation that exceeded 40 cm over 55 years. However, half of the slope sites at the cropland site in DR Congo fell below the detection limit of ICP-MS analysis, which makes the drawing of conclusions from data generated by traditional techniques very difficult if not impossible. This challenge could be addressed using ICP-MS/MS, through which the removal of UH+ interferences greater selectivity can be achieved (Table 4). The observation of extensive soil erosion, yet inability to determine measurable quantities of Pu emphasised the value of Pu isotopes measured by ICP-MS/MS to study the impact of erosion in tropical Africa where the baseline Pu signal is likely to relatively much lower than in other global regions. With the advancements of Pu analysis using reaction cell technology, analysis challenges such as limited sensitivity and cost of analysis associated with traditional methods of analysis can be overcome. The improved detection limits using ICP-MS/MS can be seen in Table 6 and the use of Pu isotopes to determine soil redistribution rates in challenging environments (low signal), increasing its viability for use in geochemical surveys associated with soil erosion studies.

Table 6 Comparison of analytical techniques for the determination of Pu activity concentrations
Full size table

Both radiometric and mass spectrometry techniques require extensive and time-consuming sample preparation steps prior to analysis which consist of the digestion of soil samples and radiochemical separation from the matrix elements. Radiometric measurements using both alpha-particle spectrometry and LSC are simple and cost-effective techniques for the determination of 238Pu, 239 + 240Pu and 241Pu. However, these methods do not have the ability to detect isotopes 239Pu and 240Pu individually. In addition, they require relatively long counting times compared to mass spectrometry methods for accurate quantification of Pu at environmental levels summarised in Table 6. Not all of the papers in Table 6 provided sufficient details to provide an in-depth comparison between methods, with many papers missing crucial details about operating conditions. An ideal format for the comparison of methods and to guide future studies would follow the presentation of experimental details by Kazi et al. [98] and Wang et al. [99]. In contrast, mass spectroscopy techniques can provide shorter analysis times and are highly sensitive with detection limits as low as 10–3 mBq g−1. Furthermore, these methods have the capability to provide individual isotopic concentrations of 239Pu and 240Pu. However, the availability and cost of some mass spectroscopy techniques is a limiting factor. In some cases depending on the intended purpose of the analysis a combination of both radiometric and mass spectroscopy techniques may be used [16]. Although AMS can be considered the gold standard for ICP-MS analysis, the cost of instrumentation set up (approximately $4 million for the set-up of each facility) and therefore availability of analytical facilities is a major limiting factor, making this method for the determination of Pu in environmental samples unattractive. Alternative mass spectrometric techniques such as SF-ICP-MS and MC-ICP-MS, have the advantage of increased resolution for the determination of isotopic ratios compared to traditional quadrupole ICP-MS, however, they require a comparable level of decontamination prior to analysis to remove interferences from UH+. This challenge can be overcome using reaction cell technology via developments in recent years of ICP-MS/MS to selectively mass shift interferences during analysis, taking advantage of the high throughput capabilities of this instrumentation over other instruments, enabling its broader application to survey scale studies on soil redistribution rates. Although not reported in the literature at this point in time, exciting developments in the field of analytical chemistry using reaction cell technology paired with high resolution SF-ICP-MS and MC-ICP-MS show promise for the future detection of isotopic ratios. However, for the purpose of soil erosion studies the additional costs associated with the setup of these analysis methods and the surplus ability to determine accurate ratios to the requirement of soil erosion measurement, makes it unlikely these methods will be used for this purpose in the future.

4 Conclusion

The development of ICP-MS/MS has opened many novel fields of research involving the analysis of Pu isotopes in soils where ultra-trace detection is required, including as a soil erosion tracer. The developments of reaction cell technology clearly demonstrates that ICP-MS/MS can be a routine tool to support Pu analysis in areas of research such as nuclear decommissioning and soil erosion tracing. The advantages that ICP-MS/MS analysis can offer relative to other instrumentation is the increased rate of analysis and subsequent lower costs per sample, meaning that the method has better availability and can be deployed for survey scale research. However, to improve the detection limits of Pu isotopes, developments in mass spectroscopy measurements using oxygen as the reaction gas are necessary in order to detect high end m/z ratios (> 271) to further enhance the selectivity for Pu through removal of polyatomic interferences. Additionally, there is a need to refine the separation process prior to analysis to allow for the effective pre-concentration of ultra-trace Pu. This has the potential to increase Pu’s applicability to be used as a soil redistribution tracer in challenging environments, such as tropical Africa, where Pu concentrations will be present in soils at ultra-trace levels. This data has the potential to inform land management practices via the better understanding of the rate of soil losses in the tropics.