Introduction

Liquid scintillation Counting is a well-established technique for the detection and quantitative measurement of soft as well as hard beta emitters. Simultaneous measurement of alpha and beta activities in liquid scintillation counters (LSC) based on pulse shape analysis (PSA) technique has been recognized for many years [1–3]. Unlike conventional LSCs where only particle energy in terms of pulse height is measured, when pulse shape is added as one more dimension, the energy spectrum turns into pulse height-pulse shape spectrum that enables alpha beta discrimination. Alpha and beta events in liquid scintillation cocktail are distinguished by examining the electronic pulses that are made up of prompt and delayed components. The difference between the delayed components of their fluorescent decay forms the basis for discrimination of alpha and beta pulses [4].

The purpose of this work was to standardize the calibration procedure for the discrimination of alpha/beta activities in a sample with minimum deviation (<10%) from the true activity levels and least variation from the optimized PSA setting.

Experimental

Activity measurements were performed with LKB Wallac Quantulus 1220 LSC equipped with Pulse Shape Analyzer (PSA), which enables the discrimination of alpha events from that of beta. In the present study, 241Am (alpha energy (MeV)—5.49(85%) and 5.44 (12%)) and 90Sr/90Y (beta energy (MeV) −0.546(100%)/2.28(100%)) were used as an alpha and beta standards respectively [5]. Although in some studies on gross alpha beta analysis natural uranium and 226Ra in equilibrium with its daughters were employed as alpha standards and 137Cs and 90Sr/90Y were employed as beta standards [6], 241Am and 90Sr/90Y standards were particularly chosen for the calibration studies as most of the emitted alphas are in the energy range 4–6 meV, while the betas typically have E max values below 2.5 meV. Moreover these standards remove uncertainty associated with in-growth of decay progeny and cover a useful energy range. 241Am standard (0.2245 μCi/g) used in the present study was received from BIPM, France. 90Sr/90Y standard (15.54 μCi/g) procured from Board of Radiation and Isotope Technology (BRIT) was standardized by CIEMAT/NIST standardization technique to obtain absolute disintegration rate [7]. These standards were diluted with 0.1 N HCl and 0.1 N HNO3 carrier solutions respectively.

Optimization of PSA setting

LKB Quantulus 1220 LSC incorporates two dual programmable Multi Channel Analyzers. This enables simultaneous measurement of four spectra, each with 1024 channel resolution. Setting the PSA level by means of user friendly software, it is possible to route alpha events into one half of the MCA (SP12) namely α-MCA and beta events into other half (SP11) namely β-MCA. Pulse shape analyzer has to be set for each and every sample to achieve the optimum separation of alpha events from that of beta. The optimum PSA level is a numeric parameter that is represented by a line, the angle of which is user controlled, divides the dual plane into short (beta) and long (alpha) pulse categories. When dividing line (PSA setting) is correctly set, counts from these two categories are stored as two separate spectra. The PSA levels can be scanned from 1 to 256 [2]. When NO PSA setting is used, all events are stored in α-MCA (SP12). As PSA values are scanned, some events from α-MCA start registering into the β-MCA. At an optimum PSA setting, alpha and beta events can be discriminated with minimum interference of beta events in α-MCA and alpha events in β-MCA.

Calibration procedure

Quantulus 1220 LSC was calibrated using pure 241Am and 90Sr/90Y standard solutions of activity concentrations 11502 and 11098 dpm (α/β~1) respectively. For precise analysis of gross alpha and beta activities in low active samples (activity concentration <50 dpm), the LSC system was recalibrated with another set of 241Am and 90Sr/90Y standards with activity concentration 22.6 and 28 dpm respectively. Additionally pure 241Am standard of 49.25 dpm and pure 90Sr/90Y standard of 930.7 dpm (α/β~1:20) were also used to calibrate LSC. At each calibration stage, the procedure followed was same as described by Sanchez-Cabeza, J.A [8].

PSA levels were scanned for 241Am and 90Sr/90Y standards and the percentage interference for each PSA setting was calculated. When a pure 241Am standard was counted, alpha interference was calculated as the fraction of counts observed in β-MCA (SP11) to the counts observed in α and β-MCA (SP12 + SP11). Similarly when a pure 90Sr/90Y source was counted, beta interference was calculated as the fraction of counts observed in α-MCA (SP12) to the counts observed in α and β-MCA (SP12 + SP11). A plot of percentage spillover against the PSA setting is illustrated in Fig. 1. The two spillover curves plotted are referred as crossover plots [9]. The PSA setting at which the two spillover curves cross was selected as optimized PSA setting for which minimum spillover of alpha events in β-MCA and beta events in α-MCA was observed.

Fig. 1
figure 1

Spillover plots of % interference of alpha events in beta MCA and beta events in alpha MCA corresponding to SQP(E)—812 for 241Am and 90Sr/90Y standards of ~104 dpm

Full size image

It is a known fact that counting efficiencies and the optimized PSA settings are affected by the degree of quenching in the sample [5]. Quantification of quench level with respect to detection (E α and E β) and spillover (E αf and E βf) efficiencies of α and β radiation is accomplished by constructing quench curve using set of quenched standards. Amount of nitro methane added to the standard vials was sequentially increased from 0 to 0.25 mL representing Spectral Quench Parameter of External standard (SQP(E)) values in the range 905–540. This SQP(E) range covers the quench levels which are usually associated with the samples. PSA settings were optimized for various quench levels and plotted against subsequent SQP(E) values as illustrated in Figs. 2 and 3. Figure 2 displays the correlation between the external quench parameter SQP(E) and optimized PSA setting for Quicksafe-400 scintillation cocktail using 241Am and 90Sr/90Y standards of respective activity concentrations ~104 dpm and ~25 dpm. Figure 3 illustrates the similar correlation for LαHβ calibration. These plots were used to determine PSA setting for a particular quench level. Quench correction curves as shown in Figs. 4 and 5 were plotted using detection (E α and E β) and spillover efficiencies (E αf and E βf) against SQP(E) and used to correct the measured activity. Detection efficiencies E α and E β decreased while spillover efficiencies E αf and E βf increased with quench. It was also observed that higher degree of quench resulted in increase in the spillover of α event into β-MCA and β event into α-MCA. These results are consistent with those of Pujol, LI. and DeVol, T.A [5, 10].

Fig. 2
figure 2

Calibration plots prepared using 241Am as an alpha standard and 90Sr/90Y as an beta standard, both having same activity concentrations ~104 dpm and ~25 dpm in Quicksafe-400 scintillation cocktail with α/β activity ratio 1

Full size image
Fig. 3
figure 3

Calibration plot prepared using 241Am of 49.25 dpm and 90Sr/90Y of 930.7 dpm, in Quicksafe-400 scintillation cocktail with α/β activity concentration ratio of ~1/20

Full size image
Fig. 4
figure 4

Detection efficiencies plotted against SQP(E) for 241Am and 90Sr/90Y of ~10000 dpm activity concentration with α/β activity ratio 1

Full size image
Fig. 5
figure 5

Spillover efficiencies plotted against SQP(E) for 241Am and 90Sr/90Y of ~104 dpm activity concentration with α/β activity ratio 1

Full size image

Sample preparation

Test samples were prepared by adding the required amount (in g) of standard solutions of 241Am and 90Sr/90Y to 12 mL of Quicksafe-400 (Zinsser Analytic) scintillation cocktail in polyethylene vials (Wheaton). The total sample and scintillation cocktail volume in the vial was maintained at 20 ml by adding 0.1 N acid carrier solution to the counting vial. Test samples were initially counted to know the quench level (SQP(E)) and then at subsequent optimized PSA setting that was obtained from the calibration plot. Quench studies were carried out using nitro-methane (Merck) as a chemical quencher.

Blank samples were prepared by adding 8 mL of 0.1 N acid carrier solution to 12 mL of Quicksafe-400 scintillation cocktail in a polyethylene counting vial to simulate sample preparation chemistry and counted along with each test sample. Prior to counting, quench level of blank sample was adjusted by adding Nitro methane so that the test sample and blank will have same degree of quench (identical SQP(E)).

Results and discussion

While analyzing the samples for gross alpha/beta activities, it is important to know the approximate activity level (or α/β activity ratio), source and chemical nature of the sample. These factors would lead to proper selection of calibration plot to be followed for the quantification of gross alpha and beta activities. Gomez Escobar et al. [11] described the anomalies observed in determining alpha activity with time and due to added chemicals using liquid scintillation calibration curve. The authors also compared accuracy and utility of quenching calibration curves with internal standard method and showed the incompetence of the constructed calibration curves for the test samples.

Performance of the 104 dpm, 25 dpm and LαHβ calibration plots over a range of quench levels (550 < SQP(E) < 900) was tested for a number of spiked samples with different α/β activity ratios. Utility of these calibration plots was also checked for the range (minimum to maximum) of activity concentration values.

Gross alpha and beta activities calculation

Count rate observed in α and β-MCA is a function of both alpha and beta disintegrations plus their interference in β and α-MCA respectively. Equations 1 and 2 were used to calculate the gross alpha and beta activities [9]. Efficiency of alpha particles in alpha MCA is defined as E α and its Spillover efficiency in beta MCA is defined as Eαf. Similarly efficiencies of beta particles in beta and alpha MCAs are defined as E β and E βf respectively. These four efficiencies were used to determine the gross alpha (A α) and gross beta (A β) activities by solving the following equations:

$$ Z_{\alpha } = A_{\alpha } E_{\alpha } + A_{\beta } E_{{\beta {\text{f}}}} $$
(1)
$$ Z_{\beta } = A_{\beta } E_{\beta } + A_{\alpha } E_{{\alpha {\text{f}}}} $$
(2)

Equation 1 gives alpha count rate (Z α) which is mainly because of alpha disintegrations (A α E α) in alpha MCA (SP12) and beta spillover (A β E βf) in the same MCA. Likewise Eq. 2 gives beta count rate (Z β) that is a combination of both beta disintegrations (A β E β) and alpha spillover (A α E αf) in beta MCA (SP11).

Solving simultaneous Eqs. 1 and 2, we get

$$ A_{\alpha } = {\frac{{Z_{\alpha } E_{\beta } - Z_{\beta } E_{{\beta {\text{f}}}} }}{{E_{\alpha } E_{\beta } - E_{{\alpha {\text{f}}}} E_{{\beta {\text{f}}}} }}}\;{\text{ and }}\;A_{\beta } = {\frac{{Z_{\beta } E_{\alpha } - Z_{\alpha } E_{{\alpha {\text{f}}}} }}{{E_{\alpha } E_{\beta } - E_{{\alpha {\text{f}}}} E_{{\beta {\text{f}}}} }}} $$

where Z α and Z β are the count rates (cpm) of alpha and beta particles respectively.

Ratio of the difference between true dpm and calculated dpm to the true dpm gives the relative percentage deviation with respect to true activity concentration values of alpha and beta standards as derived from the equation.

$$ \% deviation = {\frac{(Truedpm - calculateddpm)}{Truedpm}} \times 100 $$

At the optimized PSA settings, alpha and beta activities were calculated along with percentage deviations with respect to the true activities (dpm). While deriving alpha and beta activities if the inaccuracies were found to be more than 10%, the test samples were counted at different PSA settings with ±10 unit variations from the optimized PSA setting. The difference between the PSA setting at which test sample showed minimum % deviation and the optimized PSA setting was used as a parameter to judge the practicability of the calibration plot. The minimum difference validates the efficacy of the calibration plot.

For alpha and beta events, full energy window corresponding to the entire spectrum (channels 1–1024) was utilized. Quantulus produces separate alpha and beta spectra in a single measurement. Figures 6, 7, 8 and 9 illustrate liquid scintillation spectra of discriminated 241Am alphas and 90Sr/90Y betas in test samples with different quench levels and activity ratios.

Fig. 6
figure 6

Liquid scintillation spectra of discriminated alpha and beta in test sample (SQP(E)—835) spiked with ~10 dpm of 241Am and 90Sr/90Y with alpha/beta activity concentration ratio 1

Full size image
Fig. 7
figure 7

Liquid scintillation spectra of discriminated alpha and beta in test sample (SQP(E)—685) spiked with 241Am of 55 dpm and 90Sr/90Y of 2752 dpm with alpha/beta activity concentration ratio 1/50

Full size image
Fig. 8
figure 8

Liquid scintillation spectra of discriminated alpha and beta in test sample (SQP(E)—817) spiked with 241Am of 11072 dpm and 90Sr/90Y of 1561 dpm with alpha/beta activity concentration ratio 7

Full size image
Fig. 9
figure 9

Liquid scintillation spectra of discriminated alpha and beta in test sample (SQP(E)—719) spiked with ~14000 dpm of 241Am and 90Sr/90Y with alpha/beta activity concentration ratio 1

Full size image

Minimum detectable activity (MDA) values achieved for gross alpha and beta activities were 12.54 dpm/L (0.209 Bq/L) and 56.0 dpm/L (0.933 Bq/L) respectively for a counting time of 200 min. (SQP(E)—870).

Analysis of results obtained using ~104 dpm and ~25 dpm calibration plots

Practicability of calibration plot prepared employing 241Am and 90Sr/90Y standards of activity concentration ~104 dpm (104 dpm calibration plot) was verified for the three sets of spiked samples as illustrated in Tables 1 and 2. First set comprises test samples with standardized 241Am and 90Sr/90Y activity concentrations of equal value (α/β~1) in the range from 60 to 15000 dpm. Derived alpha and beta activities were within ±5.5% with at most 3 unit variations on the higher side (+3) of the optimized PSA settings.

Table 1 Validation results for different α:β activity ratios using ~104dpm calibration curve: calculated and true activity concentration values along with percentage deviation and PSA variation
Full size table
Table 2 Comparison of results obtained by 104 dpm and LαLβ (~25 dpm) calibration plots for test samples with α/β activity ratio ~1
Full size table

Test samples (LαHβ) in second set were spiked with 241Am of low activity concentration and 90Sr/90Y of high activity concentration in order to cover a range of α/β activity ratios from 1:2 to 1:20. In this study for α/β activity ratios 1:2 to 1:19.7, deviation was found to be within ±7% with at most 5 unit variations on the lower side (−5) of optimized PSA setting. In case of α/β activity ratio 1:21.5, inaccuracy observed was −32.6% with respect to alpha which is not acceptable. For α/β activity ratio 1:15, deviation with respect to alpha was found to be 12.6% because of low activity concentration of 241Am (~32 dpm) in test sample. Hence it was concluded that 104dpm calibration plot worked satisfactory which resulted in deviation ≤±7% for a range of α/β activity ratios 1:2 to 1:20 with alpha activity concentration above 60 dpm. Test samples (HαLβ) in the third set were spiked with 241Am of high activity concentration and 90Sr/90Y of low activity concentration to cover a range of α/β activity ratios from ~3:1 to ~20:1. For the test samples with activity ratios 3:1 to 19:1, derived alpha and beta activity concentrations were found to be within ±8% with −8 unit (maximum) variation from optimized PSA settings. In case of α/β activity ratio ~20:1, deviation of −21.3% was observed with respect to gross beta with −10 unit variation from optimized PSA setting, mainly due to low activity concentration (~55 dpm) of 90Sr/90Y.

To validate practicability of the 104 dpm and LαLβ calibration plots, for the range of activity concentration level, test samples spiked with equal activity concentrations of 241Am and 90Sr/90Y (α/β~1) covering an activity range of 8 to 200 dpm were analyzed using both the calibration plots as shown in Table 2. This exercise was necessary in order to find out, up to which activity concentration level these calibration plots provide agreeable (deviation < ±10%) results.

When the test samples spiked with equal activity concentrations of 241Am and 90Sr/90Y (α/β~1) less than 60 dpm, were analyzed using LαLβ calibration plot, calculated alpha and beta activities were within ±5.6% at the optimized PSA settings. Whereas, when the same samples were analyzed using 104 dpm calibration plot, ±4.3% deviation was observed at the cost of −8 unit variation (maximum) from optimized PSA setting. Test samples spiked with 241Am and 90Sr/90Y (α/β~1) standards of activity concentrations above 60 dpm; however, when analyzed by 104 dpm calibration plot, derived alpha and beta activities were within ±5.5% as compared to ±8.5% deviation obtained with LαLβ calibration plot. It is obvious from the above studies that LαLβ calibration plot gave precise results as compared to 104 dpm calibration plot for samples containing 241Am and 90Sr/90Y (α/β~1) less than 60 dpm.

Analysis of results obtained by LαHβ calibration plot

Almost all real samples such as drinking water, environmental and effluent samples show high beta and low alpha activity concentrations [12]. To simulate these activity levels, different test samples were spiked with 241Am of low activity concentration (of range 31.7–331.1 dpm) and 90Sr/90Y of high activity concentration (of range 160.9–5514.6 dpm). Practicability of LαHβ calibration plot was studied for these test samples that were grouped in three categories depending on their alpha activity concentration viz. 31.7, 55.5 and 145.7 dpm respectively. Test samples in each category were spiked with the same activity concentration of 241Am and different activity concentrations of 90Sr/90Y in order to cover a range of α/β activity ratios from 1:5 to 1:50. These activity proportions were considered so as to cover most environmental situations [13].

For the first category of test samples as illustrated in Table 3, where alpha activity concentration was lowest (31.7 dpm), results were found to be in good agreement with the true alpha and beta activity concentrations for α/β activity ratios 1:5–1:15. Deviations observed for these activity ratios were within ±4.1% with merely 2 unit variations from optimized PSA setting. For the activity ratio 1:20 (α/β), deviation with respect to alpha increased to ~8%, higher deviation which amounts to −25.2% for 1:25(α/β) activity ratio was insignificant and hence rejected.

Table 3 Calculated and true activity concentrations along with percentage deviation and PSA variation for test samples containing 31.7a dpm of 241Am and different activity concentrations of 90Sr/90Y
Full size table

As illustrated in Table 4, test samples spiked with 55.5 dpm of 241Am with different activity concentrations of 90Sr/90Y, were analyzed using LαHβ calibration plot. For test samples with α/β activity ratios ~1:10 to 1:25, inaccuracy of 5.2% was observed in deriving alpha and beta activities with merely 1 unit variation from the optimized PSA setting. Test sample with α/β activity ratio ~1:30, showed deviation of 6.3% with respect to beta with −5 unit variation from the optimized PSA setting. For α/β activity ratios 1:35, 1:43.1 and 1:50 where deviations observed were more than 10%, highlighted the non-reliability of LαHβ calibration plot for these activity ratios.

Table 4 Calculated and true activity concentrations along with percentage deviation and PSA variation for test samples containing 55.5a dpm of 241Am and different activity concentrations of 90Sr/90Y
Full size table

Test samples spiked with 145.7 dpm of 241Am and different activity concentrations of 90Sr/90Y with a range of α/β activity ratios ~1:5 to ~1:45 were also studied using LαHβ calibration plot as illustrated in Table 5. Inaccuracies observed in estimating alpha and beta activities were within ±7.2% with ±6 unit variation from the optimized PSA setting. This study emphasizes the usefulness of the above calibration plot for a wide range of α/β activity proportions with alpha activity concentration ~150 dpm.

Table 5 Determination of gross alpha/beta activities by LαHβ calibration plot for test samples containing 145.7a dpm of 241Am along with different activity concentrations of 90Sr/90Y
Full size table

The test samples with alpha/beta activity ratios such as 1:5, 1:10, 1:15, 1:18 and 1:20 were analyzed using ~104 dpm as well as LαHβ calibration plots and the results were compared as shown in Table 6. Inaccuracies in estimating respective alpha and beta activities in these test samples were found to be within ±6.5% (except 1:15 (α:β) activity ratio) with ±6 unit variation from optimized PSA setting when analyzed by 104dpm calibration plot. This is in contrast to LαHβ calibration plot where ±5% deviation (except 1:18 (α:β) activity ratio) with merely ±2 unit variation from optimized PSA setting was observed. The test sample with α/β activity ratio 1:18 and alpha activity concentration 331.1 dpm, when analyzed by LαHβ calibration plot, deviation was found to be 28.7% with respect to alpha. On the contrary when the same test sample was analyzed by 104 dpm calibration plot, the results were found to be within 6% which signifies the efficacy of 104 dpm calibration plot with alpha activity more than 300 dpm.

Table 6 Comparison of results obtained by ~104dpm and LαHβ calibration plots for test samples with different α:β activity ratios
Full size table

Practicability of the 104 dpm and LαHβ calibration plots was studied for a range of alpha activity concentrations as illustrated in Table 7. Table 7 confirms the usefulness of LαHβ calibration plot over 104 dpm calibration plot for the wide range of α/β activity ratios with the alpha activity concentration less than 300 dpm whereas 104 dpm calibration plot worked satisfactory above 300 dpm.

Table 7 Performance validation of ~104dpm and LαHβ calibration plots with respect to alpha activity concentration in LαHβ test samples
Full size table

Conclusion

Liquid scintillation counting technique offers simultaneous gross alpha and gross beta measurement with ~100% counting efficiency. In conclusion, all the three calibration plots viz. 104 dpm, LαLβ and LαHβ have certain limitations over the activity concentration level and α/β activity proportions. These limitations should be taken into account for the accurate measurement of gross alpha and gross beta activities. When practicability of these three calibration plots were compared, it was found that the choice of proper calibration plot plays a vital role in determining gross alpha and gross beta activities in a given sample. The validation trials with artificially spiked samples confirmed that the calibration plots compensate for varying quench in sample.