Comparison of beta (LSC) and gamma (HPGe) spectrometric methods for lead-210 in chronological study

²¹⁰Pb is a natural geogenic radioisotope that represents part of the uranium decay series and originates from the decay of ²²⁶Ra, which is found in most soils and rocks and produces short-lived gaseous ²²²Rn as its daughter. Most of this ²²²Rn decays to ²¹⁰Pb within the soil, producing supported ²¹⁰Pb, which is essentially in equilibrium with the parent ²²⁶Ra. However, some of the ²²²Rn diffuses upwards into the atmosphere, where it decays to ²¹⁰Pb (Mabit et al., 2008; Barlas, Simsek and Cagatay, 2014). This ²¹⁰Pb is deposited as fallout with aerosol particles by washout and sedimentation and commonly termed unsupported or excess (²¹⁰Pb_ex) to distinguish it from the supported ²¹⁰Pb in the soil. Goldberg (1963) was the first who outlined the ²¹⁰Pb geochronology, Krishnaswami et al. (1971) applied it to the dating of lacustrine sediments and Koide et al. (1972) later applied it to marine sediments. ²¹⁰Pb excess is extensively used for dating sediments in paleoenvironmental studies (Appleby and Oldfield, 1978; Jia et al., 2006; Zaborska et al., 2007; Kim et al., 2008; Begy et al., 2011).

Depending on different physical and chemical principles, several analytical techniques are available for the measurement of ²¹⁰Pb. There are several methods for analysing the total activity of ²¹⁰Pb in the environmental samples based on alpha, beta and gamma spectrometry (Lehto and Hou, 2011; Jia and Torri, 2007). They differ concerning the reachable detection limit, selectivity, analytical error, reproducibility and stability against different chemical composition and levels of the natural radionuclides (Ebaid and Khater, 2006).

Gamma ray spectrometry is a direct counting method of the low-energy (46.5 keV) photon emitted during the decay of ²¹⁰Pb. However, there are some challenges for biological and environmental samples (Jia et al., 2006). The 46.5 keV photons are emitted in 4% of ²¹⁰Pb decays. The self-absorption of 46.5 keV line strongly depends on the chemical composition of the sample. Therefore, the high correction factors with associated high relative uncertainties are unavoidable. The minimum detectable activity (MDA) of this technique is higher than that required for the most environmental applications, when maximum sensitivity and accuracy are required, the chemical separation method still remains the first choice.

Alpha and beta techniques for determination of total ²¹⁰Pb after radiochemical separation are more sensitive and can provide improved analytical precision (Mabit et al., 2008). The main advantage is that the evaluation of ²¹⁰Pb through the determination of ²¹⁰Po using alpha spectrometry and ²¹⁰Bi by beta particles spectrometry has a much lower limit of detection in comparison with gamma spectrometry and requires only gram-size samples (Barlas Simsek and Cagatay, 2014; Zaborska at al, 2007). On the other hand, methods involving chemical separation steps require secular equilibrium to be established between ²¹⁰Pb and ²¹⁰Po or ²¹⁰Bi. This introduces long delays between separation and the counting time (Villa et al., 2005, 2007; Jia and Torri, 2007); as a result these methods become more laborious and expensive compared to gamma ray spectrometry (Mabit et al., 2008).

Determination of ²¹⁰Pb via its beta emitting daughter ²¹⁰Bi using liquid scintillation counting (LSC) technique could be an alternative to alpha spectrometry, as it eliminates long wait for secular equilibrium, and provides a comparable detection limit with alpha spectrometry (Villa et al., 2007). The radiochemical Pb separation procedure by precipitation has been recommended by many authors (Goldberg, 1963; Polikarpov, 1966; Dušauskienė-Duž, 1997; Jia et al., 2001; Mažeika et al., 2004) due to its quite easy operation and short preparation time. Using this technique, Pb is separated by precipitation as lead sulphide (PbS) or lead sulphate (PbSO₄) and measured with liquid scintillation or proportional counter. Before precipitation, interfering bismuth is separated from the solution by an anion exchange resin column. The advantage made with the help of ²¹⁰Bi is that the measurement does not need to be made immediately after the separation of ²¹⁰Pb (Lechto and Hou, 2011). This method has a wide applicability for measuring low level ²¹⁰Pb concentrations in most environmental samples (Jia and Torri, 2007).

The most commonly used models for ²¹⁰Pb chronology are based on radioactive decay equations for converting ²¹⁰Pb_ex data from a sediment profile into a sediment chronology and include: the Constant Flux: Constant Sedimentation (CF:CS), Constant Rate of Supply (CRS) and Constant Initial Concentration (CIC). Descriptions of the models and the associated equations have been reported in (Krishnaswamy et al., 1971; Pennington et al., 1976 and Appleby, 2001). Model selection depends on environmental conditions, sediment processes, such as bioturbation, erosion, deposition.

The CIC model is appropriate when initial activity of ²¹⁰Pb_ex is constant and there is no mixing of surface sediments (MacKenzie et al., 2011 and Mabit et al., 2014). The CRS model assumes that the supply of ²¹⁰Pb_ex to the accreting material is constant in time (Appleby and Olfield, 1978), the initial ²¹⁰Pb concentration in the sediments is variable, and the influx rate of sediment is variable (Goldberg, 1963 and Appleby and Oldfield, 1978). When ²¹⁰Pb_ex fluxes and sediment mass accumulation rate (MAR) are both constant, the CIC and CRS models converge to CF:CS model and the ²¹⁰Pb_ex concentration follows an exponential decrease with mass depth (Mabit et al., 2014).

²¹⁰Pb dating depends on the accurate determination of the level of unsupported ²¹⁰Pb in a series of sediment samples, what is quite problematic because of sampling and analytical limitations (MacKenzie et al., 2011). Loss of surface material during sediment sampling can generate inaccuracies in chronology as described by Farmer et al. (2006). The validation of chronology based on ²¹⁰Pb dating models for the second half of the 20th century is often achieved by the use of artificial fallout radionuclides (e.g., ¹³⁷Cs and ²⁴¹Am) from the atmospheric testing of nuclear weapons and/or the Chernobyl accident as independent chronostratigraphic markers (O’Reilly et al., 2011).

The aim of the present study was, firstly, to compare two nuclear analytical techniques for ²¹⁰Pb analysis in sediment samples using LSC and gamma-ray spectrometer with well type HPGE detector and, secondly, to perform ²¹⁰Pb dating of sediment cores attributed to two lakes located in the Baltic Uplands.

The study was carried out on sediment cores from two lakes (Fig. 1) located in the Baltic Uplands. Lake Karackiai is a little kettle-hole type lake located in eastern part of Lithuania, near the border with Belarus. The lake covers area of 17 ha with maximal depth of 10 m. Forests occupy a prevailing part of Lake Karackiai catchment. Lake Rajgrod is located in a glacial tunnel valley in the northeastern part of Poland. The lake covers an area of 1500 ha, its greatest depth is 52 m and average is 9.4 m (Krzywicki et al., 2007). The sampling area was attributed to the eastern part of lake. The Lake Karackiai is insignificantly impacted by anthropogenic activity. The Lake Rajgrod is impacted by agriculture, also there are water level regulation by dam constructed on out flowing stream (in 50s XX c.). Based on visual evaluation Lake Rajgrod is more eutrophic compared with Lake Karackiai.

Fig. 1

Radiochemical procedure

The first step of pre-treatment involved the drying of the sample at 105°C temperature, crushing, grinding and sieving through 2 mm sieve mesh size. For ashing, 10 g of the dried sample was taken and ashed at 550°C for 6 hours. Finally, the sample ash was ground and homogenized.

Leaching

There are used several leaching methods as reported in Jia et al., 2006: (i) leaching with aqua regia; (ii) successive leaching with HNO₃ + HF, HClO₄ and HCl, etc. To remove nearly all organic matter and silicates from the samples, which can prevent the recovery of ²¹⁰Pb and decrease the radiochemical yield, samples oxidation and leaching were carried out followed by wet ashing in our study. After the sample pre-treatment, the chemical separation was started by adding to the beaker containing 3 g of the ashed sample 1 ml of stable Pb²⁺ carrier (30 mg/ml) as the chemical yield tracer. For the further procedure, 5 ml of HNO₃ and 15 ml of HCl were added to samples and heated at 250°C to evaporate the solution to dryness. The remaining material was treated with 10 ml of HCl (1M) to finally destroy the organic matter. Solution was evaporated to dryness again. The residue was finally dissolved with 30 ml of 1M HCl, filtered through a 0.1 µm Millipore filter paper, and transferred to a 50 ml volumetric beaker.

Separation

Pb was separated from the leachate solution using anion exchange resin column in chloride form (Eichrom 1 ×8, 100–200 mesh). Ion-exchange column was approx. 130 mm long and 11 mm of inner diameter. The anion exchange resin was sequentially treated with 20 ml of 2M H₂SO₄, 20 ml of 6M HNO₃ and 100 ml of distilled water to remove fine particles as well as other components after each separation procedure.

The scheme of procedure for the ²¹⁰Pb separation from leachate solution is shown in Fig. 2.

Fig. 2

The available leachate solution from the sample was loaded onto an anion exchange resin column preconditioned with 20 ml of 2M HCl at a flow rate 0.5 ml min^–1. After rinsing with 10 ml of 1M HCl supernatant was discarded. Under this condition (10 ml of 1M HCl) Bi should be fixed on the resin in contrast to Pb (Happel et al., 2006). From the loaded anion exchange resin column Pb could be separated by elution with 0.05M or 0.5M HCl (Figgins, 1961; Gibson, 1961 and Bhatki, 1977). H₂O as the Pb elution liquid to separate Pb from Bi can be used as well (Jia and Torri, 2007). In our study Pb was eluted with 50 ml of H₂O at the same flow rate. The separation time of Pb from Bi was measured.

For determination of the radiochemical yield gravimetrically 4 ml of H₂SO₄ were added to the eluate in order to precipitate lead as sulphate (PbSO₄). After evaporation of solution till volume of 20–30 ml the precipitate clearly occurred. Then volumetric beaker with content was cooled and precipitate was filtered through ashless Whatman filter. Filter paper with precipitate was put in a weighed crucible, ashed at 550°C for 4 hours, cooled until constant weight and weighed again to calculate the Pb radiochemical yield based on Pb²⁺ carrier added.

Dissolution

According to Jia et al. (2006) the PbSO₄ precipitate can be dissolved in sufficient volume of 6M ammonium acetate by heating. Nearly in the same way Letho and Hou (2011) recommended to dissolve the lead sulphate in 10 ml of 6M ammonium acetate. Other solutions, including EDTA/NaOH and ammonium citrate, were investigated for PbSO₄ dissolution (Kim et al., 2008 and Lozano et al., 2012). In our case, PbSO₄ was distributed in 2 ml of 3M ammonium acetate, mixed with 18 ml of scintillation cocktail (Optiphase HiSafe 3, Perkin Elmer) in a 20 ml plastic vial and stored for over 30 days in a refrigerator for consequent LSC assay.

The results of the radiochemical yield (gravimetrical and by ²¹⁰Pb tracer) obtained from model samples and other methodical features (solubility and homogeneity of measuring form for LSC) are given in Table 1.

Table 1

Comparison of investigated radiochemical procedures.

The radiochemical yield of Pb evaluated by both methods was as follows: 72–77% for gravimetric method and 70–73% for ²¹⁰Pb tracer method. All three solvents used to dissolve PbSO₄ precipitates were suitable. The long-term stability of dissolved PbSO₄ mixture with 18 ml scintillation cocktail Optiphase HiSafe 3 was confirmed by clear and homogenous measuring form with 2 ml of 6M ammonium acetate (#1) and 2 ml of 3M ammonium acetate (#2). Whereas PbSO₄ dissolved in 2 ml of 0.5 EDTA/NaOH and mixed with 18 ml of scintillation cocktail resulted a precipitate formation on the bottom of the vial after 1 day. The counting efficiency based on standard and blank counting rates in the ²¹⁰Pb– ²¹⁰Bi window (30–770 channels) for PSA level 125 varied between 88% and 91%. Based on experimental results (Table 1) 2 ml of 3M ammonium acetate solution was used further in radioanalytical procedure for ²¹⁰Pb determination in lakes sediment samples by LSC method.

A set of samples from two cores of both lakes was prepared and measured. The data obtained from direct measurements of ²¹⁰Pb (gamma spectrometry) and from indirect measurements (LSC) was used to compare the results of different measurement techniques. The results of ²¹⁰Pb activities versus core depth for both methods are shown in Fig. 3 (Lake Karackiai) and Fig. 4 (Lake Rajgrod).

Fig. 3

Fig. 4

It is evident that, except for a few samples from Lake Rajgrod core (Fig. 4), the²¹⁰Pb activity concentrations obtained using both methods are similar within the range of uncertainties. The observed shift between results for particular samples are likely to be related to an incomplete homogenization of those samples. In fact, analysis of the sediments between three methods, alpha and gamma spectrometry, and LSC (Villa et al., 2007) has shown the similar discrepancies, which can be attributed to insufficient processing as well. Also, a trend of higher activities of ²¹⁰Pb measured by gamma compared to LSC for the upper depth intervals of Lake Rajgrod core can be observed. The relative uncertainty associated with gamma spectrometry measurements of ²¹⁰Pb is quite large for samples of Lake Rajrod (Fig. 4). At low ²¹⁰Pb activity (20 Bq kg^–1) uncertainly associated with gamma spectrometry measurements can reach ±70%, although for higher activities (>60 Bq kg^–1) uncertainties become acceptable (±20%) and when activities exceed 110 Bq kg^–1, uncertainty levels are close to those associated with LSC measurements (±15%). A similar systematic uncertainty between measuring techniques were also noticed by Tanner et al. (2000), Zaborska et al. (2007) and Mabit et al. (2008).

The ²¹⁰Pb excess (unsupported), which is used for age calculation, was determined subtracting supported ²¹⁰Pb (²¹⁴Pb) from total ²¹⁰Pb. The total ²¹⁰Pb activities were not determined for each sediment slice by LSC. Further for age calculation ²¹⁰Pb_ex activity extrapolation from the exponential function was applied for comparison of two age data sets.

Activity of unsupported ²¹⁰Pb decreases with depth for the both cores, however some differences from the exponential function can be observed for surface layers. ²¹⁰Pb_ex profile for Lake Karackiai exhibits surface mixing with depth of approx. 2 cm (Fig. 5), while mixing at Lake Rajgrod profile was obtained to approx. 11 cm depth (Fig. 6). The ²¹⁰Pb equilibrium depth reached 12 cm for Lake Karackiai core (Fig. 5) and 30 cm for Lake Rajgrod core (Fig. 6). Figs. 5 and 6 illustrate only the ²¹⁰Pb_ex results which were used for age-depth model.

Fig. 5

Fig. 6

The inventories of ²¹⁰Pb_ex (Bq/m²) were calculated based on cumulative sum of the individual ²¹⁰Pb_ex activity values (Bq/kg) determined by both beta and gamma methods over the entire profiles, multiplied by the mass depth of each sediment layer (g/cm²). The results of both sites included in this study were found to be comparable. In fact, the calculated ²¹⁰Pb_ex inventories results derived by beta and gamma methods for Lake Karackiai are 2579 and 2783 Bq/m² while for Lake Rajgrod are 1918 Bq/m² and 2650 Bq/m², respectively.

The CRS model could be used to derive ²¹⁰Pb chronology of sediments as a ²¹⁰Pb_ex activity with depth is highly variable (Goldberg, 1963; Appleby and Oldfield, 1978; Barlas Simsek and Cagatay, 2014). Therefore, the sediment age, sedimentation rate (SR) and mass accumulation rate (MAR) values for studied cores were determined by the CRS model (Appleby et al., 1991). The ²¹⁰Pb age-depth relationships for CRS model of two cores are shown in Figs. 7 and 8.

Fig. 7

Fig. 8

The artificial fallout radionuclide ¹³⁷Cs data representing independent chronomarker are included into those graphs for ²¹⁰Pb chronology validation (Pennington et al, 1973; Smith, 2001 and Walling et al, 2002). ¹³⁷Cs inventory in core of Lake Karackiai was estimated 1470 Bq/m², while in core of Lake Rajgrod was estimated 3100 Bq/m². The Lake Rajgrod ¹³⁷Cs inventory data significantly high imply the Chernobyl accident fallout impact on catchment.

The ¹³⁷Cs peak was found at depth of 6–7 cm (359±19 Bq/kg) in core of Lake Karackiai and at depth of 12–15 cm (114±6 Bq/kg) in core of Lake Rajgrod and can be attributed to Chernobyl accident. Those sediment layers were dated by ²¹⁰Pb CRS model as 1973–1978 and 1985–1989, respectively. Some discrepancies between ²¹⁰Pb CRS model dates and ¹³⁷Cs peak position could be explained as ¹³⁷Cs penetration to greater depth than expected from diffusion (Mabit et al., 2014). It can be influenced by combination of physical and biological mixing of newly accumulated sediments (O’Reilly et al., 2011), as well as by postdepositional migration in sediments (Appleby et al., 1991).

The comparing of sedimentation parameters evaluated by ²¹⁰Pb_ex data obtained by two analytical methods is presented in Table 2.

Table 2

Sediment mass accumulation and sedimentations rates (mean values) based on ²¹⁰Pb_ex evaluated by gamma and beta methods.

From ²¹⁰Pb chronology derived mean values of sediment mass accumulation rate (MAR) were following: 0.008 g/cm²/yr based on beta method and 0.009 g/cm²/yr based on gamma method for Lake Karackiai; 0.055 g/cm²/yr based on beta method and 0.060 g/cm²/yr based on gamma method for Lake Rajgrod. The mean values of linear sedimentation rate were: 0.14 cm/yr using gamma data and 0.12 cm/yr using beta data for Lake Karackiai; 0.41 cm/yr using gamma data and 0.33 cm/yr using beta data for Lake Rajgrod. The differences in accumulation rate in these two lakes might be due to the damming of the Lake Rajgrod in mid 50s of XX c. The water level of the lake was raised some 1–1.5 m. After the II world war the town developed and due to the lack of environmental awareness the lake waters were polluted what increased the rate sedimentation (Krzywicki et al., 2007).

²¹⁰Pb activity concentration in lacustrine sediments was determined using two nuclear analytical techniques including beta LSC and HPGe gamma spectrometry. Both techniques in the range of uncertainties gave similar ¹³⁷Cs, ²¹⁰Pb and ²²⁶Pb measurement results. The large relative uncertainty associated with gamma analysis of samples with low ²¹⁰Pb activity makes beta LSC technique attractive in some circumstances even if this method is labour consuming. The values of ²¹⁰Pb_ex inventory derived using data from beta and gamma measurements were found to be similar with for Lake Karackiai (2579 and 2783 Bq/m²) and slightly variable for Lake Rajgrod (1918 and 2650 Bq/m²). Among two lakes more eutrophic Lake Rajgrod exhibited the higher sediment mass accumulation rate (0.055 and 0.060 g/cm²/yr) compared to less eutrophic Lake Karackiai (0.008 and 0.009 g/cm²/yr). The ²¹⁰Pb chronologies with some limitations were validated using artificial fallout radionuclide ¹³⁷Cs as independent chronomarker.