Low-sulphate water sample preparation for LSC detection of 35S avoiding sulphate precipitation
Introduction
The knowledge of groundwater residence times is mandatory for the sustainable management of groundwater resources. The data can be used (i) for recommending groundwater abstraction rates that ensure sustainable aquifer use, (ii) for assessing groundwater travel times and related matter (and contaminant) transport, and (iii) for evaluating aquifer vulnerabilities regarding anthropogenic contamination.
A suitable tool for investigating groundwater residence times is the application of environmental tracers, i.e., of naturally occurring substances that are generally suitable for studying water related physical and/or chemical processes. Powerful in this regard are, besides stable isotopes, environmental radionuclides. Ideally, their half-lives should be in the same time range as the investigated processes. Rather long-lived radionuclides (such as 3H, 14C, 36Cl, 39Ar, 81Kr, and 85Kr) have proven suitable for studying long-term processes. However, using radionuclides for covering shorter timespans, e.g. for the investigation of groundwater residence times of less than one year, is only rarely discussed in the literature. This is mainly due to the fact that only a rather small number of generally applicable natural radionuclides show adequately short half-lives (such as the frequently applied 222Rn (e.g., Treutler et al., 2007; Schmidt et al., 2010; Petermann et al., 2018) or 224Ra (e.g., Moore and de Oliveira, 2008; Rocha et al., 2015)).
A promising novel approach for covering the sub-yearly timescale is based on the application of radioactive sulphur (35S). 35S is continuously produced in the stratosphere by cosmic ray spallation of 40Ar. After its production 35S rapidly oxidizes to sulphate, gets dissolved in the meteoric water and is finally transferred with the rain to the groundwater (Tanaka and Turekian, 1991). 35SO42− activities in precipitation range generally between ca. 5 and 100 mBq/l (Oh et al., 2019; Urióstegui et al., 2015; Cho et al., 2011; Hong and Kim, 2005; Osaki et al., 1999; own data). Since there is no natural 35S source in the subsurface the 35S activity concentration in any freshly recharged groundwater starts to decrease by decay with an 87.4 day half-life as soon as the rainwater enters the ground. This makes 35S a potential residence time tracer suitable for investigating groundwater ages between about three to nine months (i.e., between one and three 35S half-lives).
The idea of using 35S as residence time tracer was first introduced nearly two decades ago (Michel et al., 2000; Sueker et al., 1999). However, the published case studies were all limited to high geographical elevations where snowmelt is the dominant hydrological recharge event, thus simplifying the annual 35S input function to the peak snowmelt. Since rainfall is likely to show a substantial variation in 35S activity (e.g., Tanaka and Turekian, 1991; Plummer et al., 2001; own data) the experiences reported from these alpine/subalpine watersheds are of only limited applicability in non-alpine climates.
The drivers of the variable 35S activity in rain have not been systematically investigated yet and call for further studies. Still, in any case the varying 35S activities in rainfall require analysing a rather large number of rain samples in order to set up reasonable 35S input functions. That in turn makes a sample preparation procedure desirable that is as straightforward as possible.
35S detection by liquid scintillation counting (LSC) requires pre-concentration of 35SO42− from large water samples (generally up to 20 L; an approach that requires smaller sample volumes (5 L) that was suggested by Oh et al. (2019) requires large-volume ultra-low-level LSC measurement). A related state-of-the-art approach was suggested by Urióstegui et al. (2015) and improved by Schubert et al. (2019). It entails (i) sulphate extraction from the water with an anion-exchange resin (Amberlite IRA400/Cl-form), (ii) its elution from the resin with a NaCl solution, and (iii) its precipitation from the eluate as fine-grained BaSO4 by addition of BaCl2. The precipitate is finally (iv) homogeneously suspended in the gel-forming Insta-Gel Plus® scintillation cocktail and measured by LSC (Urióstegui et al., 2015; Schubert et al., 2019). The approach was developed for water samples that contain sulphate with rather low 35S/32SO42− ratios as they are typical for groundwater. Although the approach allows measuring samples that contain sulphate loads of up to 1500 mg it has four major disadvantages: (i) the BaSO4 precipitation step is rather labour-intensive, (ii) the added BaCl2 may contain substantial amounts of 226Ra thus increasing the LSC background, (iii) measuring gel suspensions requires very careful sample preparation since very fine-grained precipitates are mandatory in order to prevent inaccurate measurements due to self-absorption, and (iv) LSC measurement of heterogeneous gel suspensions of low energy β-emitters (such as 35S) always yields lower counting efficiencies than counting of mono-phase emulsions. Therefore our study aimed at simplifying and improving the sample preparation procedure by avoiding the BaSO4 precipitation step and producing a mono-phase emulsion of sample and LSC cocktail instead of a heterogeneous gel suspension. The improved approach is applicable if waters with relatively high 35S/32SO42− ratios (as they are typical for rainwater) are to be measured.
A high 35S/32SO42− ratio of the water (i.e., a higher relative 35S activity) reduces the total SO42− load of the sample that is necessary for attaining a countable 35S sample activity. Nevertheless, even low SO42− concentrations in aqueous samples complicate LSC measurement with most commercial LSC cocktails since divalent anions (such as SO42−) generally trigger phase separation in clear cocktail emulsions. Therefore, we decided to use the LSC cocktail Hionic-Fluor® (PerkinElmer), which is reported to be specifically applicable for aqueous samples with elevated ionic strengths. Still, the capacity of Hionic-Fluor® to hold elevated total salt loads is also limited. In preliminary experiments we found that the cocktail (18 ml) is suitable for a total sulphate load of the processed aqueous sample (2 ml) of up to 100 mg. Hence the cocktail accepts sulphate loads of up to 20 L rainwater (with an assumed SO42− concentration of 5 mg/l).
Still, the ionic strength limitation requires a water sample preparation procedure that avoids the addition of any ions that would increase the given sample intrinsic ion strength. Consequently, our study also aimed to develop a sample preparation approach, in which the high salt concentration that is required for the elution of sulphate from the ion exchange resin is achieved with ions that are volatile, thus allowing their easy removal from the sample before LSC measurement.
Section snippets
Material and methods
A Quantulus GCT 6220 liquid scintillation counter was used for all 35S activity measurements. The measurements were done in the “Normal” counting mode with GCT correction either for 60 min or until the previously set 2 Sigma threshold of statistical uncertainty (0.5%) was reached (Table 1). The detection background was counted and subtracted from the sample counts by measuring a35S-dead background vial, which was treated in the same way as all 35SO42− containing lab-made “standards” and natural
Experimental
The following three sets of experiments, each aiming at an individual objective, were executed:
- (i)
The first set of experiments was carried out for developing and fine-tuning the Hionic-Fluor® based sample preparation procedure. The experiments were carried out with five defined aqueous 35SO42− standard solutions with increasing total sulphate loads.
- (ii)
The second set of experiments was carried out for optimizing (i.e., minimizing) the load of anion-exchange resin. Six experiments with identical 35SO42−
Optimization of the resin load
Aim of the experiments described in sect. 3.2 was to find the optimum load of Amberlite IRA67 (OH-form) for extracting 100 mg sulphate from an aqueous solution (rainwater). As mentioned above, the sulphate recovery was detected in two ways, by LSC (35S) and by ion chromatography (sulphate). The results of the LSC measurements are displayed in Table 2.
Fig. 1 compares the 35S activities that remained in the solutions (calculated from the detected 35S recoveries shown in Table 2) with the sulphate
Conclusion
The introduced approach is recommended as alternative to the established method for 35S detection in natural water samples which necessitates BaSO4 precipitation. The major advantages of avoiding the sulphate precipitation step are (i) that the sample processing is less labour intensive, (ii) that it prevents potential addition of 226Ra to the sample with the BaCl2 that is added for the sake of BaSO4 precipitation (Urióstegui et al., 2015), (iii) that it minimizes self-absorption by the
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
There are no conflicts of interest.
References (17)
- et al.
Atmospheric depositional fluxes of cosmogenic 35S and 7Be: implications for the turnover rate of sulfur through the biosphere
Atmos. Environ.
(2011) - et al.
Determination of residence time and mixing processes of the Ubatuba, Brazil, inner shelf waters using natural Ra isotopes
Estuar. Coast Shelf Sci.
(2008) - et al.
Groundwater residence times in Shenandoah national park, blue ridge mountains, Virginia, USA: a multi-tracer approach
Chem. Geol.
(2001) - et al.
Improved approach for LSC detection of 35S aiming at its application as tracer for short groundwater residence times
J. Environ. Radioact.
(2019) - et al.
Use of cosmogenic 35S for comparing ages of water from three alpine/subalpine basins in the Colorado Front Range
Geomorphology
(1999) - et al.
Measurement of cosmogenic 35S activity in rainwater and lake water
Anal. Chem.
(2005) - et al.
Detection of deep stratospheric intrusions by cosmogenic 35S
PNAS
(2016) - et al.
Timescales for migration of atmospherically derived sulphate through an alpine/subalpine watershed, Loch Vale, Colorado
Water Resour. Res.
(2000)