Introduction

Plutonium (Pu), an anthropogenic radionuclide, has been intentionally or accidentally introduced into the environment since the first nuclear weapons tests in the mid-1940s through various human activities including: releases in accidents (e.g., Chernobyl and Fukushima [1, 2]); discharges from reprocessing plants [3,4,5]; and global fallout from above-ground nuclear weapons testing [6, 7]. Because of their high chemical toxicity, long half-lives and internal radiation threat, the determination of concentrations of Pu isotopes in the environment is of great important. Meanwhile, Pu isotopes in the environment have been widely used as tracers to study environmental processes [8], such as water circulation and scavenging efficiency in the oceans [9,10,11,12,13]. All such investigations require highly sensitive analytical techniques to measure Pu isotopic compositions [14], producing high quality data sets.

Initially, α-spectrometry was used as the main analytical tool to determine Pu concentration in the environment for the alpha emitting isotopes (238Pu, 239Pu and 240Pu) [15, 16]. However, α-spectrometry is burdened with several disadvantages such as requiring a large volume of sample, laborious pretreatment and long counting time ranging from days to several weeks. Also, α-spectrometry has no ability to distinguish 239Pu and 240Pu due to their nearly identical α radiation energies (5.16 MeV and 5.17 MeV, respectively). Thus, α-spectrometry cannot provide information on the 240Pu/239Pu atom ratio, which hampers Pu tracing studies of environmental processes. The 240Pu/239Pu atom ratio is related to the Pu contamination source because Pu isotopic ratios vary with reactor type, nuclear fuel burn-up time, neutron flux and energy, and for fallout from nuclear detonations, they vary with weapon type and yield [17]. The average 240Pu/239Pu atom ratio of global fallout is considered to be 0.180 ± 0.014 [18]. Weapons-grade Pu has a lower 240Pu/239Pu atom ratio (0.01–0.07) because of its high 239Pu purity [19, 20]. Reported 240Pu/239Pu atom ratios of reactor-grade Pu vary from 0.2 to 1.0 depending on the fuel burn-up, for example, they are 0.30–0.38 from Fukushima accident fuel [2, 21] and 0.42–0.52 from Chernobyl accident fuel [22,23,24,25].

In the past 30 years, thanks to the rapid development of mass spectrometry, various analytical methods for Pu determination have been established in environmental samples, such as thermal ionization mass spectrometry (TIMS [4, 5, 26]), accelerator mass spectrometry (AMS [27]), inductively coupled plasma mass spectrometry (ICP-MS [11, 28,29,30]) and multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS [12, 13, 31]). Compared to α-spectrometry, mass spectrometry offers advantages for Pu determination, including high sensitivity, a small sample volume requirement and a capability to provide accurate information about the 240Pu/239Pu atom ratio. Certainly, alpha spectrometry cannot be totally replaced by ICP-MS methods, for example, 238Pu can only be measure by alpha spectrometry at the present. The standards of Pu isotopes did not catch up with the developments in mass spectrometric instrumentation and measurement methodologies for the last 30 years [32]. In order to produce consistent and reliable results, laboratories must implement an appropriate quality control/quality assurance (QC/QA) program in which their performance can be monitored. The use of reference materials (RMs) is highly recommended for method development and validation. Meanwhile, confidence in comparability and reliability of measurement results in nuclear materials and environment sample analyses have been established via RMs. Isotopic ratios of Pu in the environment have been widely investigated to elucidate its source terms and to study its behavior and fate [5, 12, 13, 33].

According to the International Organization for Standardization (ISO), RMs have property values certified by technical validations from worldwide inter-comparison exercises traceable to SI units. Pu-RMs, supplied by the National Institute of Standards and Technology (NIST), European Commission Directorate General Joint Research Centre and International Atomic Energy Agency (IAEA), are used in the assay and isotopic standard as the spike in the analysis of Pu by isotope dilution mass spectrometry. The Pu-RMs are designed for use in developing accurate methods of analysis, calibrating measurement systems including determining performance characteristics or measuring a property at the state-of-the-art limit and ensuring long-term adequacy and integrity of measurement quality assurance programs [34]. The following should be considered when Pu-RMs are chosen: (1) the RMs should satisfy the user’s target; (2) the matrix and the concentration of Pu-RMs should be as close as possible to the test values of sample: (3) handling and storage should be done according to the certification document; and (4) the Pu-RMs should be used within the validity period. Many matrices, such as sediment, seawater, biota and fallout, of Pu-RMs have been presented in the literature. The activity levels of Pu-RMs are usually certified by the IAEA or NIST. For any certified value and/or information value, one should refer to the original certification of the related RMs. However, the values of Pu isotopic ratio are not always simultaneously certified, for example, the certified 239+240Pu activities of IAEA-384 (Fangataufa Lagoon Sediment) and IAEA-385 (Irish Sea Sediment) were given as corresponding to 103–110 Bq kg−1 and 2.89–3.00 Bq kg−1 at the 95% confidence interval by the IAEA, but their 240Pu/239Pu atom ratios were not provided by the agency. Nevertheless, the missing 240Pu/239Pu atom ratios of IAEA-384 and IAEA-385 are usually able to be found in the literatures [31, 35,36,37,38].

Our objective here is to review the Pu-RMs covering sediment/soil, seawater, biota, and fallout and provide their statistical values of Pu isotopic ratios after statistical analysis of Pu datasets. The certified activity of Pu and accurate information on isotopic ratio of Pu in RMs is important for QC/QA analyses and method validations of environmental Pu analysis, and for instrument calibration using these materials. There is a strong need for matrix matched isotopic standards for use as quality control materials [32]. Especially in mass spectrometric analyses, where the reproducibility on Pu isotope ratio measurements is found to be excellent, the pedigree of the materials used as standards has large implications on the accuracy of the result. With modern mass spectrometers precision of < 0.01% in the isotope ratio measurement is achievable for actinides [39,40,41]. Both accuracy and precision would have to be combined to appropriately estimate realistic uncertainties obtained in isotope ratio measurements [39,40,41]. Meanwhile, the certified Pu activity and statistical value of Pu isotopic ratio will also help in quickly establishing a suitable selection of RMs for Pu determination. Finally, we point out that the concurrent information on certified activity provided by RMs producer and statistical value of Pu isotopic ratio will also strengthen data quality in establishing a baseline for environmental Pu analysis and environmental risk assessment related to nuclear power plant operations that are seeing a dramatically rise in some countries.

Data sources and methods

Data sources

We collected over 600 data of Pu-RMs from more than 100 publications (the original data see in Supporting Information: spreadsheet S1), in which the matrices of the RMs included sediment/soil, seawater, biota and fallout. These Pu datasets were mainly extracted from data base websites such as the Web of Science, Google Scholar and Scopus when we input the keywords “plutonium and 240Pu/239Pu atom ratio” combined with the primary keyword “reference material and/or certified reference material”. Data of Pu-RMs from geological and environmental reference materials (GeoReM) database are also included, which is download from http://georem.mpch-mainz.gwdg.de.

Data treatment

Assuming a non-parametric distribution of data for which distribution-free statistics are applicable, we recalculated the collected Pu data by statistical analyses. Each set of Pu data collected from the literature is assumed to have the same weight (i.e., weighted mean = arithmetic mean). The statistical values of Pu isotopic ratios were the arithmetic means with corresponding standard deviations when more than two results were reported or were weighted means with weighted uncertainties in the case of only two results being reported. The median values of Pu were calculated from all reported Pu results passing the following criterion of Z-score, rounded off to the most significant number of the uncertainty. Meanwhile, following the guidelines of the ISO [42] and NIST [43], we also calculated the expanded uncertainty with a coverage factor of k = 1.96, corresponding to a level of confidence of about 95%. Confidence intervals were taken from a non-parametric sample population representing a two-sided interval at 95% confidence limits.

Following the recommendations from the ISO [44] and International Union of Pure and Applied Chemistry (IUPAC) [45], we calculated the statistical value along with the application of an appropriate target standard deviation, namely, a Z-score as suggested by the following equation:

$$ Z = \frac{{X_{i} - X_{\text{a}} }}{{S_{\text{b}} }} $$
(1)

where Xi is the reported Pu isotopic ratios or activities from the literature i, Xa is a mean value of combined results and Sb indicates the target standard deviation. In general, the relative bias in radioactive analysis is below 20% (Sb< 10%). We included the uncertainty of the assigned value (Stu) in the target value for bias [46].

$$ Z = \frac{{X_{i} - X_{\text{a}} }}{{\sqrt {S_{\text{tu}}^{2} + S_{\text{b}}^{2} } }} $$
(2)

The accuracy of the Pu isotopic ratio and the activity from different publications was expressed by the Z-score for each RM. We thought a result was acceptable when this relative difference between the average value and the assigned value was below or equal to 2, indicating about 95% of the data points were between a Z-score of − 2 and + 2. Otherwise, the acceptable result from the extreme of the distribution had about a 1 in 20 chance when the Z-score was outside |Z| > 2. Furthermore, the chance that the result was acceptable was only about 1 in 300 when the Z-score was outside |Z| > 3 [47]. In addition, the median absolute deviation (MAD) is a measure of statistical dispersion, which is in detail described in elsewhere [48]. In brief, assuming the data follows the Gaussian distribution, the outliers fall in 50% of the area on both sides and the normal values fall in 50% of the area in the middle [48]. The MAD is simultaneously used to test statistical dispersion and to get rid of the outliers in a data set in this study. The results show that the two methods are consistent, which is potentially due to the results of the same RM measured in each laboratory are relatively concentrated and they have no big deviation.

Results and discussion

We summarized 30 RMs for the environmental Pu analysis, thoroughly covering the matrices of sediment/soil, seawater, biota and fallout. We tested the performance of the Pu dataset for each RM through the Z-score distribution. Here, taking IAEA-135 and IAEA-368 as typical examples, we discuss in detail their Z-score distributions for 239+240Pu activity and 240Pu/239Pu atom ratio. As shown in Fig. 1, the Z-scores of 239+240Pu activity and 240Pu/239Pu atom ratio for IAEA-135 and IAEA-368 were between − 2 and + 2, and that indicated the Pu datasets from different publications were satisfactory. We excluded the data point if the Z-score outliers |Z| > 2 and recalculated until a satisfactory result was obtained. Finally, we calculated the statistical values for each RM and tabulated them (Table 1).

Fig. 1
figure 1

Distributions of Z-scores of 239+240Pu activity and 240Pu/239Pu atom ratio for IAEA-135 (a) and IAEA-368 (b)

Full size image
Table 1 Activity levels and isotopic ratios of Pu-RMs
Full size table

Pu-RMs in sediment and soil

Activity level of Pu

The evaluation results in order of ascending activities, the distribution of medians and corresponding standard deviation for four typical RMs (IAEA-135, IAEA-368, IAEA-384 and NIST-4357) are shown in Fig. 2.

Fig. 2
figure 2

Data evaluation for 239+240Pu activity in IAEA-135 (a), IAEA-368 (b), IAEA-384 (c) and NIST-4357 (d). The median (black solid line) and corresponding standard deviation (red dashed lines) are shown. The error bars correspond to the combined uncertainty reported in the literature. (Color figure online)

Full size image

IAEA-135 (Irish Sea Sediment): the 239+240Pu activities showed a consistent result although they were sourced from the determination results of different instruments. For example, the 239+240Pu activity measured by ICP-MS ranged from 222 to 238 Bq kg−1 (average = 224 ± 10 Bq kg−1) [49] and that analyzed by semiconductor alpha-spectrometry (SAS) was in the range of 216–228 Bq kg−1 (average = 222 ± 7 Bq kg−1) [49]. The 239+240Pu concentration determined by AMS was also in good agreement with the values of ICP-MS and SAS, being 221 ± 16 Bq kg−1 [49]. Through analyzing the published 239+240Pu activity of IAEA-135 [31, 49,50,51,52,53,54,55,56,57,58,59,60,61,62,63], we found these Pu data showed good homogeneity, falling less than two standard deviations from the distribution mean (Fig. 2a). The statistical value of 239+240Pu activities for IAEA-135 was 218.2 ± 6.8 Bq kg−1 (n = 25, the 95% confidence interval is 215.9–220.5 Bq kg−1). This statistical value was comparable to those certified by IAEA (205–226 Bq kg−1), indicating these measured values are highly reliable. The median given as the statistical value is 219.9 Bq kg−1. Meanwhile, the statistical values of 238Pu and 241Pu activities were 34.6 ± 1.7 Bq kg−1 (n = 12) and 890 ± 141 Bq kg−1 (n = 14) (corrected to January 1, 2020 for the convenience of reference and comparison in the future study), respectively. The 95% confidence level interval corresponds to be 33.8–35.4 Bq kg−1 and 828–952 Bq kg−1.

IAEA-368 (Ocean Sediment-Mururoa Atoll): the 239+240Pu dataset was evaluated using more than ten international reports [17, 28, 31, 51, 55, 57, 58, 64,65,66,67,68,69,70,71,72,73,74,75,76], and they showed good homogeneity as suggested by the Z-scores (all below 2). As shown in Fig. 2b, the reported 239+240Pu activities had a narrow range, falling within two standard deviations from the distribution mean. Our given statistical value of 239+240Pu activities was 31.2 ± 1.4 Bq kg−1 (n = 22, the 95% confidence interval is 30.7–31.7 Bq kg−1). This statistical value is highly consistent with those certified by IAEA (29–34 Bq kg−1), indicating these measured values are highly reliable. The median given as the statistical value is 31.3 Bq kg−1. The statistical value of 238Pu and 241Pu activity was 7.0 ± 0.4 Bq kg−1 (n = 2) and 4.21 ± 0.38 Bq kg−1 (n = 4) (the 95% confidence level interval corresponds to be 6.6–7.4 Bq kg−1 and 3.90–4.52 Bq kg−1), respectively.

IAEA-384 (Fangataufa Lagoon Sediment): the combined 239+240Pu dataset from different international laboratories showed good homogeneity as the Z-scores were less than 2 [12, 13, 31, 35, 37, 49, 53, 67, 76,77,78,79,80,81,82,83,84]. The reported Pu activity had a narrow range, falling within two standard deviations from the distribution mean (Fig. 2c). Our statistical value of 239+240Pu activities was 108.4 ± 2.7 Bq kg−1 (n = 19, the 95% confidence interval is 107.0–109.8 Bq kg−1). This statistical value is highly consistent with those certified by IAEA (103–110 Bq kg−1). The median given as the statistical value is 108.7 Bq kg−1. Meanwhile, we recommended the statistical values of 238Pu and 241Pu activities to be 33.8± 1.7 Bq kg−1 (n = 8) and 21.5 ± 1.9 Bq kg−1 (n = 7) (the 95% confidence level interval corresponds to be 32.8–34.8 Bq kg−1 and 20.3–22.7 Bq kg−1), respectively.

NIST-4357 (Ocean Sediment): the 239+240Pu data representing means obtained by different international laboratories were used in the calculation process [17, 50, 61, 70, 73, 74, 85,86,87,88,89,90,91]. These Pu data fell within less than two standard deviations from the distribution mean (Fig. 2d). The Z-scores were also between − 2 and + 2, showing good consistency in the Pu dataset. The statistical value of 239+240Pu activity was 9.96 ± 0.47 Bq kg−1 (n = 13, the 95% confidence interval is 9.74–10.18 Bq kg−1). This statistical value is highly consistent with the certified values from IAEA (9.2–13.3 Bq kg−1), indicating these measured values are highly reliable. The median given as the statistical value is 9.90 Bq kg−1. Meanwhile, the statistical value of 238Pu and 241Pu activity was 1.93 ± 0.01 Bq kg−1 (n = 2) and 45.0 ± 2.6 Bq kg−1 (n = 4) (the 95% confidence level interval corresponds to be 1.92–1.94 Bq kg−1 and 42.9–47.1 Bq kg−1), respectively.

Using the above method, we also analyzed 18 other Pu-RMs, namely, IAEA-300, IAEA-306, IAEA-307, IAEA-315, IAEA-326, IAEA-327, IAEA-367, IAEA-375, IAEA-385, IAEA-410, IAEA-412, IAEA-447, IAEA-soil-6, NIST-1646a, NIST-2702, NIST-4350b, NIST-4353 and NIST-4354. The statistical values of 238−241Pu activities are presented in Table 1, which were comparable to their certified values. The Pu isotopic ratios of RMs are usually not certified by the approving agency such as the IAEA and NIST. However, these isotopic ratios of Pu are very important to trace Pu sources and calibrate measurement methods. Therefore, we further analyzed isotopic ratios of Pu-RMs and reported their statistical values.

Isotopic ratio of Pu

The evaluation results in order of 240Pu/239Pu atom ratio, the distribution of medians and corresponding standard deviation for four typical RMs (namely, IAEA-135, IAEA-368, IAEA-384 and NIST-4357) are presented in Fig. 3.

Fig. 3
figure 3

Data evaluation for 240Pu/239Pu atom ratio in IAEA-135 (a), IAEA-368 (b), IAEA-384 (c) and NIST-4357 (d). The median (black solid line) and corresponding standard deviation (red dashed lines) are shown. The error bars correspond to the combined uncertainty reported in the literature. (Color figure online)

Full size image

IAEA-135 (Irish Sea Sediment): the 240Pu/239Pu atom ratio showed a consistent result although they were sourced from the determination results of different instruments. For example, the 240Pu/239Pu atom ratio measured by ICP-MS was 0.207 ± 0.006, which was consistent with the AMS result (0.219 ± 0.022) [49]. As shown in Fig. 3a, the 240Pu/239Pu atom ratio showed good homogeneity, falling less than two standard deviations from the distribution mean. After analyzing the published 240Pu/239Pu atom ratios [31, 49, 51, 54, 58, 61, 92,93,94,95,96,97,98], we recommended the statistical value of the 240Pu/239Pu atom ratio for IAEA-135 to be 0.209 ± 0.008 (n = 18, the 95% confidence interval is 0.206–0.212), with the variation of 240Pu/239Pu atom ratio below 4%. The median given as the statistical value is 0.210. Such a statistical value was higher than that of global fallout [18] because low-level liquid radioactive waste from BNFL Sellafield was discharged to the Irish Sea from 1952 to 1992 [99]. Meanwhile, the statistical values of 238Pu/239+240Pu activity ratios and 241Pu/239Pu atom ratios were recommended to be 0.153 ± 0.008 (n = 5) and 0.0047 ± 0.0007 (n = 7) (the 95% confidence level interval corresponds to be 0.147–0.159 and 0.0043–0.0051), respectively.

IAEA-368 (Ocean Sediment-Mururoa Atoll): we evaluated the 240Pu/239Pu atom ratios using findings from more than ten international laboratories [17, 28, 31, 51, 58, 64, 65, 68, 70,71,72,73,74,75, 86, 92, 97, 98, 100, 101]. The 240Pu/239Pu atom ratios showed good homogeneity because the Z-scores were below 2. As shown in Fig. 3b, the 240Pu/239Pu atom ratios fell within two standard deviations from the distribution mean. The statistical value of the 240Pu/239Pu atom ratios for IAEA-368 was reported to be 0.0326 ± 0.0013 (n = 20, the 95% confidence interval is 0.0321–0.0331), which was comparable to that obtained in weapons-grade Pu [20] and was significantly lower than that of global fallout [18]. The median given as the statistical value is 0.0326. Meanwhile, the 238Pu/239+240Pu activity ratios and 241Pu/239Pu atom ratios were recommended to be 0.232 ± 0.047 (n = 1) and 0.000170 ± 0.000121 (n = 7), respectively.

IAEA-384 (Fangataufa Lagoon Sediment): we summarized the published 240Pu/239Pu atom ratios from the literature [12, 13, 31, 35, 37, 49, 76, 81, 82, 102], and they showed good homogeneity as the Z-scores were below 2. As shown in Fig. 3c, the 240Pu/239Pu atom ratios had a narrow range, falling within two standard deviations from the distribution mean. Our statistical value of 240Pu/239Pu atom ratio for IAEA-384 was 0.051 ± 0.001 (n = 13, the 95% confidence interval is 0.050–0.052), with the variation of 240Pu/239Pu atom ratio below 2%. The median given as the statistical value is 0.050. Such a statistical value was comparable to that obtained in weapons-grade Pu [20] and was reasonable considering the 193 French nuclear weapon tests (equivalent to 13.2 Mt TNT) conducted in the Mururoa and Fangataufa Atolls [103] resulted in 239Pu enrichment of IAEA-384. The statistical value of 238Pu/239+240Pu activity ratios was 0.310 ± 0.009 (n = 5) (the 95% confidence interval is 0.303–0.317).

NIST-4357 (Ocean Sediment): we collected means from different international laboratories concerning the 240Pu/239Pu atom ratios for our calculation [17, 61, 70, 73, 74, 86,87,88,89,90,91]. The 240Pu/239Pu atom ratios fell within less than two standard deviations from the distribution mean (Fig. 3d). The Z-scores were also between − 2 and + 2, showing good consistency in the 240Pu/239Pu atom ratios. Our statistical value of 240Pu/239Pu atom ratios for NIST-4357 was 0.233 ± 0.004 (n = 10, the 95% confidence interval is 0.231–0.235), with the variation of 240Pu/239Pu atom ratio below 2%. The median given as the statistical value is 0.233. The statistical values of 238Pu/239+240Pu activity ratios and 241Pu/239Pu atom ratios were recommended to be 0.192 ± 0.010 (n = 2) and 0.00507 ± 0.00008 (n = 3) (the 95% confidence level interval corresponds to be 0.180–0.204 and 0.00499–0.00515), respectively.

Following the above method, the information values of Pu isotopic ratio from 18 other RMs (same as mentioned above) were also obtained and given in Table 1. Since the stocks of the existing primary standard are declining, the Pu standard with a certified isotopic ratio is not unable to meet the growing demands of Pu measurement. And the transport/import of Pu standard is a greater hurdle compared to the RMs. These statistical values of Pu isotopic ratios will thus help in calibrating the mass bias, by quickly establishing a suitable selection of RMs for Pu determination.

Pu-RMs in water

Activity levels of Pu

Two typical Pu-RMs in seawater, namely, IAEA-381 and IAEA-443, were analyzed using the Z-score and statistical evaluation.

IAEA-381 (Irish Sea Water): 239+240Pu activities used in our analyses were findings from multiple independent international laboratories [49, 104,105,106,107,108,109,110,111,112]. As shown in Fig. 4a, the Z-scores were less than 2, showing good consistency of the Pu dataset derived from findings of different laboratories. The 239+240Pu activities fell within less than two standard deviations from the distribution mean (Fig. 5a). The statistical value of 239+240Pu activities for IAEA-381 was calculated to be 14.7 ± 0.7 mBq kg−1 (n = 9, the 95% confidence interval is 14.4–15.0 mBq kg−1). This statistical value is highly consistent with those certified by IAEA (13.2–15.2 mBq kg−1), indicating these measured values are highly reliable. The median given as the statistical value is 14.8 Bq kg−1. Meanwhile, the statistical values of 238Pu and 241Pu activities were calculated to be 2.61 ± 0.12 mBq kg−1 (n = 6) and 83.3 ± 7.0 mBq kg−1 (n = 3) (the 95% confidence level interval corresponds to be 2.53–2.69 mBq kg−1 and 78.1–88.5 mBq kg−1), respectively.

Fig. 4
figure 4

Distributions of Z-scores of 239+240Pu activity and 240Pu/239Pu atom ratio for IAEA-381 (a) and IAEA-443 (b)

Full size image
Fig. 5
figure 5

Data evaluation for 239+240Pu activity and 240Pu/239Pu atom ratio in IAEA-381 (a, b) and IAEA-443 (c, d). The median (black solid line) and corresponding standard deviation (red dashed lines) are shown. The error bars correspond to the combined uncertainty reported in the literature. (Color figure online)

Full size image

IAEA-443 (Irish Sea Water): This RM has been recently provided by IAEA. We prepared our Pu dataset from the findings of multiple international laboratories [27, 111,112,113,114,115]. As shown in Fig. 4b, the Z-scores were below 2, showing good consistency of the Pu dataset from the findings of different laboratories. The 239+240Pu activities fell within less than two standard deviations from the distribution mean (Fig. 5c). Therefore, the statistical value of 239+240Pu activities for IAEA-443 was calculated to be 15.2 ± 0.7 mBq kg−1 (n = 7, the 95% confidence interval is 14.7–15.7 mBq kg−1). This statistical value was comparable to those certified by IAEA (14.3–15.0 mBq kg−1), indicating these measured values are highly reliable. The median given as the statistical value is 14.7 mBq kg−1. Meanwhile, the statistical values of 238Pu and 241Pu activities were calculated to be 2.80 ± 0.09 mBq kg−1 (n = 1) and 95.9 ± 13.3 mBq kg−1 (n = 4), respectively.

Isotopic ratio of Pu

Information on isotopic ratio of Pu in the RMs was not certified by the IAEA. Here, we combined Pu isotopic data from the literature and derived their recommended values by statistical analyses.

IAEA-381 (Irish Sea Water): 240Pu/239Pu atom ratios of IAEA-381 have been extensively reported [49, 86, 104, 106, 108,109,110, 112, 116]. As shown in Fig. 4a, the Z-scores were below 2, showing good consistency of 240Pu/239Pu atom ratios provided by different laboratories. The 240Pu/239Pu atom ratios fell within less than two standard deviations from the distribution mean (Fig. 5b). The statistical value of 240Pu/239Pu atom ratios for IAEA-381 was calculated to be 0.236 ± 0.005 (n = 8, the 95% confidence interval is 0.233–0.239), a variation in the results of ± 3.4%. The median given as the statistical value is 0.238. This statistical value was higher than that of global fallout because the radioactive waste from BNFL Sellafield was discharged to the Irish Sea from 1952 to 1992 [99]. In addition, the recommended value of 238Pu/239+240Pu activity ratios was calculated to be 0.182 ± 0.002 (n = 3) (the 95% confidence interval is 0.181–0.183).

IAEA-443 (Irish Sea Water): We summarized the reported 240Pu/239Pu atom ratios from several international laboratories [12, 13, 111,112,113, 115, 117, 118]. As shown in Fig. 4b, the Z-scores were below 2, showing good consistency of 240Pu/239Pu atom ratios from different laboratories. The 240Pu/239Pu atom ratios fell within less than two standard deviations from the distribution mean (Fig. 5d). We recommended the statistical value of 240Pu/239Pu atom ratio for IAEA-443 to be 0.232 ± 0.003 (n = 7, the 95% confidence interval is 0.230–0.234), a variation in the results of ± 1.3%. The median given as the statistical value is 0.233. This statistical value was comparable to those obtained for IAEA-381 and IAEA-135, which were also sampled from the Irish Sea. Meanwhile, the statistical value of 238Pu/239+240Pu activity ratio was recommended to be 0.192 ± 0.008 (n = 1).

It is well known that the mass bias correction is necessary to obtain high quality data in the determination of Pu using ICP-MS [119]. However, the transport/import of Pu standard with a certified isotopic ratio presents the greatest hurdle at present for many laboratories working on environmental Pu analysis. Hence, calibration of the mass bias in the determination of Pu ratio by mass spectrometry is a difficult issue. In the absent of any Pu isotope ratio standard, the purification of Pu from high Pu concentration seawater obtained from the IAEA could be regarded as an alternative approach for the mass bias correction in the determination of Pu isotopic ratio considering the precision required for environmental Pu analysis. Of course, for safeguards analysis, which requires much higher precision for Pu isotope ratio analysis, certified Pu isotope ratio standards should be used, but this issue is beyond the scope of this study. Here, our statistical values of Pu isotopic ratios after the statistical analysis could be recommended to calibrate the mass bias (Table 1), which is very important for the validation of Pu measurement.

Pu-RMs in biota

We discussed in detail the RM in biota for IAEA-134 (sample matrix: fish flesh in Irish Sea) through combining the Pu data from literatures [31, 49,50,51, 92]. As shown in Fig. 6a, the Z-score values of 239+240Pu activity and 240Pu/239Pu atom ratio were less than 2, showing good consistency of Pu dataset from the different laboratory. The 239+240Pu activities and 240Pu/239Pu atom ratios fell within less than two standard deviations from the distribution means (Fig. 6b, c). The statistical values of 238Pu and 239+240Pu activities for IAEA-134 were calculated to be 2.53 ± 0.18 Bq kg−1 d.w. (n = 2) and 15.6 ± 0.9 Bq kg−1 d.w. (n = 5) (the 95% confidence level interval corresponds to be 2.32–2.74 Bq kg−1 d.w. and 14.9–16.3 Bq kg−1 d.w.), respectively. The statistical values of 238Pu/239+240Pu activity ratios and 240Pu/239Pu atom ratios were also calculated to be 0.153 ± 0.008 (n = 2) and 0.208 ± 0.005 (n = 4) (the 95% confidence level interval corresponds to be 0.144–0.162 and 0.203–0.213), respectively. The median given as the statistical value of 240Pu/239Pu atom ratio is 0.210. In addition, the statistical values of Pu activities and isotopic ratios for IAEA-414 (sample matrix: fish flesh) and IAEA-446 (sample matrix: seaweed) are also presented in Table 1.

Fig. 6
figure 6

Distributions of Z-scores of 239+240Pu activity and 240Pu/239Pu atom ratio (a) and data evaluation for 239+240Pu activity (b) and 240Pu/239Pu atom ratio (c) in IAEA-134. The black solid lines and red dashed lines showed the median value and the corresponding standard deviation, respectively. The error bars correspond to the combined uncertainty reported in the literature. (Color figure online)

Full size image

Pu reference fallout material

Reference fallout material of Pu is very important for QC of fallout Pu analysis in atmospheric studies. However, scarcely any reference fallout material has been certified by NIST or IAEA. Otsuji-Hatori et al. [120] tried to prepare the reference fallout material using deposition samples collected at 14 Japanese meteorological stations during 1963–1979 and they measured the 238Pu activities (0.218 ± 0.089 Bq kg−1 d.w.) and 239+240Pu activities (6.61 ± 0.42 Bq kg−1 d.w.) by α-spectrometry. The 238Pu/239+240Pu activity ratios were calculated to be 0.033 ± 0.012. Subsequently, Zhang et al. [70] measured the 241Pu activities (4.62 ± 0.84 Bq kg−1 d.w.) and 239+240Pu activities (6.68 ± 0.16 Bq kg−1 d.w.) of reference fallout material in Japan using SF-ICP-MS. Their measured 239+240Pu activities were in good agreement with the reference value provided by Otsuji-Hatori et al. [120]. Meanwhile, the 240Pu/239Pu atom ratios and 241Pu/239Pu atom ratios were calculated to be 0.1919 ± 0.0005 and 0.00104 ± 0.00021, respectively. Recently, Ohtsuka et al. [101] measured the 240Pu/239Pu atom ratio of Japan fallout material to be 0.191 ± 0.003, which was in good agreement with Zhang et al.’s result [70]. Therefore, we analyzed these Pu datasets of reference fallout material in Japan and listed statistical values in Table 1. At present, only limited reference fallout materials have been referenced. It is a great challenge to collect enough fallout material to prepare the reference material since the activity levels are becoming extremely low and there is no long-term observation station to provide a collection place.

Conclusions

We first reviewed over 30 Pu-RMs and prepared many Pu datasets. After statistical analyses, we compiled the statistical values of 238−241Pu activities and isotopic ratios of Pu (238Pu/239+240Pu activity ratios, 240Pu/239Pu atom ratios and 241Pu/239Pu atom ratios) for the RMs. Especially, we expect the statistical values of the Pu isotopic ratios will come to fill the gaps in information provided for the RMs by their approval agencies such as IAEA and NIST. The statistical values of 239+240Pu activity are highly consistent with the certified values from the approval agencies (R2 = 0.9999, Fig. 7), indicating these measured values are highly reliable. Therefore, we expect that through further independent work at different laboratories, the statistical values of Pu isotopic ratios have high reliability. The statistical value of Pu isotopic ratio will also help in establishing a baseline for environmental Pu analysis and provide a way to suitably select RMs. Finally, we point out that Pu isotopic ratios measured in the reference materials will not have the same credentials as the certified values referring to ISO test method or ASTM test method. These statistical values can, at best, be consensus values. Therefore, it is very necessary to produce the matrix matched Pu isotopic standards for quality control of environmental Pu analysis.

Fig. 7
figure 7

Relationship between statistical value of 239+240Pu activity of reference material and its certified 239+240Pu activity from the approval agencies

Full size image