Introduction

Isolation of high level nuclear waste (HLW) within ‘multiple barrier system’, is a widely accepted methodology for nuclear waste management [1, 2]. The idea involves fixing of HLW in suitable inert matrices and placing them inside deep geological repository within canisters/overpacks made up of stainless steel/copper alloy [3–15]. The metallic containers are isolated from host rock by composite layer of bentonite or other clay minerals, graphite, sand etc. [16–20]. Any openings remaining within the multi-barrier system are further blocked through backfilling with crushed rock, concrete, clay minerals etc. Both buffer and backfill materials provide cushioning effect to the metallic containers from external stresses. Countries like Finland, Sweden, Switzerland, France have already made significant progress in building geological repository system. Although such geological repositories are most likely to be constructed within dry regions but interaction between meteoric water with ‘multiple barrier system’ components cannot be ruled out. Such interactions become all the more important if one considers small local water bodies which are likely to form due to usage of water for construction and development of geological repositories. Creation of different openings such as boreholes, access shafts, galleries, disposal tunnels etc. will not only require significant amount of water but will also provide ample scopes for the construction materials (e.g. cement. cementitious backfill, plaster, grouted waste package etc.) and accumulated broken rock masses (due to excavation) to interact with flowing or stagnant water bodies. Needless to say, such natural and anthropogenic materials’ interactions with water can lead to formation of cave-like carbonate deposits such as stalactite, stalagmites, columns, flowstones and draperies (collectively called speleothems) within geological repositories as has been witnessed within various man-made underground openings such as abandoned road tunnels, water channels, mines etc. [21–33]. It is therefore understood that such carbonate precipitations within geological repository or adjacent to any nuclear facility, can modify fluid flow-paths through altering rock porosity and permeability, which may ultimately result in facilitating (creation of open spaces through dissolution) or isolating (immobilization through sorption and/or co-precipitation) radionuclides from contacting with the mobile fluid phase [34–37]. Such reduction in narrow open spaces due to precipitation can also hinder maintenance of homogeneous chemical/radiochemical/thermal/microbial environment within geological repositories.

The possibility of radionuclide migrations and redistributions within geological repository arise in the event of failure(s) of ‘Engineered Barrier Systems (EBS)’ i.e. canister, overpack and buffer materials. In such event, the HLW waste matrix can come in contact with water and through its leaching the radionuclides can get released into ‘near field region’. Of the various components, Cs (cesium) and Sr (strontium) radionuclides, particularly 137Cs (half life ~30 years) and 90Sr (half life ~28.8 years) are important as they are known to be (i) major heat producing radionuclides, (ii) abundant within HLW, (iii) relatively easily leachable from waste form matrices, and (iv) emit γ-rays and β-particles respectively. Among the other Cs radionuclides 135Cs has a long half-life (~2×106 years) also. Because of their high solubility within ground waters; as Cs+ and Sr2+ under all conditions of Eh and pH; and sorption within natural materials, radiocesium and radiostrontium can become part of food chain web which can significantly affect health and safety of biosphere [38–40]. For example, substitution of 90Sr for Ca in human bone can lead to irradiation of bone marrow thereby increasing the chance of leukaemia [41]. It may be mentioned here that calcium carbonates being one of the major constituents of Earth’s crust, play a major role in the preservation of biosphere and food-chain web through regulating ‘carbon cycle’. Calcium carbonates get precipitated from solvated calcium and carbonate ions as a result of supersaturation or as a biomineralization product. Apart from geological repository scenario, radiocesium and radiostrontium can encounter carbonate materials in the event of approved release, accidents/fall outs [42–44] or leakage of HLW-storage tanks as happened in Hanford, USA [45–47], Oak Ridge, USA [48, 49], Sellafield, UK [50] and Mayak, Russia [51, 52]. Incorporation radiocesium and radiostrontium within natural rocks (mostly of sedimentary origin), soils and minerals take place due to multi-site sorption. Experimental studies have shown that these radionuclides get preferentially fixed at ‘Frayed Edge Sites’ or sorped on ‘Regular Exchange Complex Sites’ like interplanar sites or edge lattice positions [53–59]. Needless to say, considering the prolong service period of geological repository (~one million years) it is very important to develop sorption database for important radionuclides on all possible natural materials that are likely to be encountered within multiple barrier system. This will help in ‘total system performance assessments’ as well as to carry out modeling study [60]. Unfortunately no sorption studies have been carried out for radiocesium and radiostrontium on naturally precipitated carbonate materials. In order to fill-up this gap, an attempt has been made in the present investigation to understand the sorption of Cs and Sr radionuclides on natural carbonates, using stalagmite sample collected from Dharamjali cave (Pithoragarh district) in the eastern Kumaun Lesser Himalaya [61]. Details of the experimental procedures are given below.

Experimental

For the present study, stalagmite samples were collected from Dharamjali cave occurring within Thalkedar limestone of Kumaun Lesser Himalaya. A representative stalagmite sample from Dharamjali cave and its cross-section is shown in (Fig. 1). The sample has been characterized for its physical and chemical properties before it was taken for radiochemical sorption studies.

Fig. 1
figure 1

Cross-section of a the natural stalagmite is shown in b. Note the optical images of three major domains a coarse columnar calcite crystals within translucent layer, b microcrystalline calcite interlayered with unidentified brownish/greyish phase(s) within grey layers, and c botryoidal fibrous aragonite crystals defining the milky white layers. (Color figure online)

Full size image

Physical characterization

Phase analysis of the stalagmite sample has been carried out using optical microscopic technique. Thin slices of the stalagmite were mounted on glass slides, ground on different grades of emery papers and finally polished on lapping wheel using 0.5 µm diamond paste.

Same polished samples were used for phase analyses (at ambient temperature and pressure) using micro-Raman spectroscopy. The analyses spots were excited at 532 nm (Nd-YAG laser; power ~20 mW at sample position) using a 10× objective lens. The scattered light was collected by the same objective lens and passed through an edge filter to separate the Stokes signal from the Rayleigh and anti-Stokes scattered signals. The Stokes signal after the edge filter was passed through a fiber-coupled spectrograph (Acton series SP 2300i, 1800 groove/mm) and detected by a thermo-electric cooled (−75 °C) charge-coupled device (CCD). The resolution of instrument was 1 cm−1.

Chemical characterization

In line with the focus of present study, trace elemental analysis of stalagmite samples has been carried out using a non-destructive spectroscopic method called ‘laser induced breakdown spectrometry (LIBS)’. It involved ablating a small quantity of the sample and monitoring the subsequent emission from atomic species. To this end, a 50 fs Ti–Sapphire laser system (λ = 800 nm) operating at 1 kHz was employed to generate plasma by focusing 1 mJ of energy onto the sample surface using a BK7 planoconvex lens (f = 10 cm) to a spot diameter of approximately 14 μm. The samples were mounted on a motorized Y–Z translational stage so as to provide a fresh surface for each laser shot and thereby obtain compositional profiles across different domains. An Echelle grating spectrograph in conjunction with an ICDD (International Centre for Diffraction Data) is used for acquisition of spectra.

Radionuclide sorption study

For present experimental study, 137Cs and 85+89Sr radiotracers were extracted from HLW arising from PUREX process and purified by cation exchange route [62]. To ensure better quality of data, experimental parameters were optimized following the procedure outlined in Sanwal et al. [33].

Following batch equilibration, 137Cs and 85+89Sr sorption experiments were carried out with vortex shaker. Powdered samples (0.1 g) were added to solutions (3 mL), each of which were spiked separately with 137Cs and 85+89Sr and maintained at fixed pH (1–6). Equilibration time was maintained at 60 min for each study. After the experiments, solutions were filtered and counted using single channel gamma analyzer equipped with NaI(Tl) detector. For better accuracy, separate examinations were done with blank solutions and tracer absorption on the walls of extraction vials were found negligible. K d (distribution coefficients), were calculated using following equation:

$$K_{{\text{d}}} = {{\left[ {V\left( {C_{o} - C_{e} } \right)} \right]} \mathord{\left/ {\vphantom {{\left[ {V\left( {C_{o} - C_{e} } \right)} \right]} {C_{e} \times M}}} \right. \kern-\nulldelimiterspace} {C_{e} \times M}}$$
(1)

where ‘C o’ and ‘C e’ are feed and effluent concentrations respectively; ‘V’ is volume (mL) of solution, and ‘M’ is amount of ion exchanger/adsorbent used.

It may be mentioned here that the requirement for carrying out the present sorption study under acidic pH stems out for two reasons. (i) Presence of localized acidic environment, within repository or any other geological environment with nuclear facilities, due to presence of fulvic acids and humic substances cannot be ruled out [63–75], and (ii) earlier experimental studies show natural carbonates (calcite and aragonite) may not totally dissolve even under strong acidic environment [32, 33]. This observation is contradictory to common belief but the possible reasons of enhancement of natural carbonate stability even under acidic conditions are being evaluated separately. One possible reason could be natural incorporation of wide range of trace elements within carbonate lattices [76–79].

Results and discussion

Physical characterization

Longitudinal cross section of the natural stalagmite sample (Fig. 1a) is shown in Fig. 1. It is evident from Fig. 1b that the sample has three basic components, namely (i) porous brown colored domains, (ii) massive white domain, and (iii) translucent grey colored domains. Representative optical images of the three components are given in Fig. 1c–e. As evident from Fig. 1c, the porous domain is found to be very thinly layered, fine grained and highly pitted. High magnification optical images identify the domain as made up of microcrystalline calcite. On the other hand, white domain is found to be constituted of botryoidal aragonite needles (Fig. 1d). The remaining translucent domain is constituted of elongated calcite phases, commonly known as ‘palisadic’ type.

The interpretations of the optical microscopic images (Fig. 1) to different microcrystalline phases have been verified with micro-Raman data (Fig. 2). Although the calcite, aragonite and ikaite have overlapping Raman bands, their specific occurrences were indicated by certain characteristic peaks. For example, in the low frequency region, the Raman band at 207 cm−1 is characteristics of aragonite/ikaite structures whereas that at 282 cm−1 is the manifestation of calcite as well as aragonite structures [80]. Moreover, the in-plane bending mode of aragonite appears at 705 cm−1 and shift to higher frequency (at 712 cm−1) for calcite [81]. Figure 2 shows the Raman spectra of the stalagmite sample corresponding to the three distinctly coloured domains, (i) white (ii) brown and (iii) grey. As shown in the top panel of Fig. 2, the Raman spectrum of the brown and grey colored domains are almost identical to each other while that at the white domain is different. The spectral bands corresponding to different vibrational modes of the stalagmite sample are shown in expanded form in the bottom panels of Fig. 2. The white domain of stalagmite sample shows multiple bands in the low frequency region (154, 183, 207 and 260 cm−1) but the brown and grey domains show only two bands (158 and 283 cm−1; bottom left panel in Fig. 2). Thus, the white domain contains aragonite and/or ikaite structures, and the brown/grey domains contain calcite and/or aragonite structures [80, 81]. Expanded view of the second higher energy band (in-plane bend of CO3 2−) shows that the band appears at 704 cm−1 for the ‘white domain’ and shifts to 712 cm−1 for the brown/grey domains’ (bottom middle panel in Fig. 2), which is a clear signature that the white domain is rich in aragonite and the brown/grey domain in calcite. Moreover, the position of the symmetric stretch band of CO3 2−, which can distinguish between ikaite (1070 cm−1) and aragonite/calcite (1086 cm−1) structures, appears at 1087 ± 2 cm−1 for all the three domains (white, brown and grey) of the stalagmite sample. Thus, the position of the CO3 2−-symmetric stretch band, in agreement with the position of the in-plane bending of CO3 2−, conclusively proves that white domain of the sample is rich in aragonite and the brown/grey regions in calcite.

Fig. 2
figure 2

Top panel: Raman spectrum of polished stalagmite in three different regions (white, brown and grey). The picture of the sample is shown in the inset. Bottom panels: expanded view of the different vibrational bands: lattice vibration (left), CO3 2− in-plane bend regions (middle) and CO3 2− symmetric stretch (right). (Color figure online)

Full size image

Chemical characterization

Certain elements (e.g. Sr, Mg, Rb etc.) which are present within HLW are also known to occur within natural fluids which lead to precipitation of natural carbonates. It is therefore always interesting to know of this gamut of available cations and anions, which are the ones which get fixed within the carbonate lattices and how their respective compositional variation relates to phase changes which happen within speleothem in micrometer to few millimetre scales mostly due to fluctuations in climate. This information is important to explain sorption experimental data as well. Representative LIBS data are shown in Fig. 3. It is noted that the brown domain constituted of microcrystalline calcite contained significant amount of K (potassium) whereas Mg (magnesium) was preferentially locked inside columnar calcite rich grey domains. Sr (Strontium) exhibited higher preference for aragonite containing white domains. Similar trace elements distribution trends, in terms of Sr/Ca and Mg/Ca ratios, are observed across the multiple layers of the stalagmite (Fig. 4). What becomes very clear from this study are, in case of an open system where different calcium carbonates are allowed to precipitate from natural liquid, (a) among the calcium carbonates Sr has a preference for aragonite over calcite, and (b) aragonite can host Sr over long time scale.

Fig. 3
figure 3

LIBS data from different domains of stalagmite

Full size image
Fig. 4
figure 4

Elemental distributions in different domains of stalagmite

Full size image

Radionuclide sorption study

Sequestration of radionuclides within geological repositories by natural carbonates will be primarily done by sorption mechanisms especially at the initial stages. The sorption mechanism may involve various processes in operation, either individually or collectively like

  • adsorption (building up of chemical component at the fluid/solid interfaces without formation of a three dimensional molecular arrangement),

  • absorption (uptake of a (fluid) chemical component within micro/nanoporous solid (e.g. in zeolites) through diffusion),

  • precipitation (incorporation of an aqueous chemical component within a given solid phase through growing (homogeneous/heterogeneous nucleation) a three dimensional crystalline molecular arrangement,

  • co-precipitation (precipitation process involving specific trace element iso-structural with one of the host component), and

  • recrystallization (reconstitution of host’s crystalline structure to a similar one but with different trace element concentrations as a function of intensive/extensive variables of the system).

Over prolonged time-scale of interaction, all these processes can operate in the following sequence (arranged in order of increasing reaction time durations), adsorption → absorption → precipitation → co precipitation → recrystallization. The present experimental study however corresponds to the faster sorption mechanism only i.e. adsorption process.

Detail experimental results for sorption studies are given in Tables 1 and 2. To optimise ‘equilibration time’, experiments with powdered carbonate domains were carried out over a time interval ranging from 5 to 60 min [32, 33]. In order to remove any possible perturbations in sorption kinetics due to ‘nucleation and crystal growth mechanisms [82]’, higher contact time was avoided. As sorption was maximum at 60 min so this was considered as equilibration time duration. Keeping equilibration (contact) time fixed at 60 min, centrifugation speed was optimized and high %sorption values were obtained for 7000–8000 rpm [32, 33]. Hence, all other experimental studies were done maintaining centrifugation speed at 8000 rpm. Similarly, in ‘volume efficiency’ experiments (with 5–20 mL feed solutions) maximum K d value was obtained for 5 mL solution [32, 33].

Table 1 137Cs sorption studies (5 mL of feed solution with initial concentration, C o = 280.20 mCi/L; equilibration time: 60 min with vortex shaker) [%Sorption = (C o − C e) × 100/Co, K d = (C o − C e) × V / (C e × M)]) on 0.05 g (=M) powdered aragonite, columnar calcite and microcrystalline calcite domains over pH range of 1–6. Δ aragonite, ○ columnar calcite, □ microcrystalline calcite domains
Full size table
Table 2 85+89Sr sorption studies (5 mL of feed solution with initial concentration, C o = 980.29 µCi/L; equilibration time: 60 min with vortex shaker) [%Sorption = (C o − C e) × 100/C o, K d = (C o − C e) × V/(C e × M)]) on 0.05 g (=M) powdered aragonite, columnar calcite and microcrystalline calcite domains over pH range of 1–6. Δ aragonite, ○ columnar calcite, □ microcrystalline calcite domains
Full size table

For all studied radionuclides, sorption coefficients within different domains of stalagmite are found to be from moderate to low. 137Cs sorption was found higher in calcite (both columnar and microcrystalline varieties) compared to aragonite at low pH (1–2) and it reversed with increasing pH (Fig. 5a). Similar was the trend for 85+89Sr (Fig. 5b) also, but the K d value was much less compared to those for Cs-sorption. This observation comes as a contradiction to the general perception that sorption of Sr radionuclides should be more on calcium carbonate surfaces than Cs due to similarity in ionic charges. The reason behind such contradictory observations is being explored in the light of present simulation studies. One possible reason behind lower sorption K d of radiostrontium, in comparison with that of radiocesium, may be due to already existence of Sr within active sorption sites of speleothems domains (incorporated during its formation through natural process), which might have acted as barrier towards further sorption of similar elements. In fact the present LIBS data show higher abundance of Sr within aragonite domains compared to calcite ones. But it may also be mentioned here that in actual scenario radionuclide sorption depends on various factors including crystal chemistry of substrate [83], fluctuations in solution chemistry (steady state vs non-steady state [84], dominant speciation on carbonate surfaces [85] etc. In fact recent leaching studies with thermodynamically most stable surface plane of calcite (104) show existence of equilibrium dynamics between dissolution and re-precipitation with strong control over recrystallization by pH, CO2 partial pressure, and presence of ions in the aqueous phase [86].

Fig. 5
figure 5

Variations in log(K d) values of 137Cs, and 90Sr within different phases of cave deposits (Δ aragonite, ○ columnar calcite, □ microcrystalline calcite) as a function of pH

Full size image

Conclusions

Detailed microstructural studies of stalagmite sample revealed presence of three parts, namely (i) porous brown colored domains, (ii) massive white domains, and (iii) translucent grey colored domains. Optical microscopic analyses of the domains identified them to made up of microcrystalline calcite (brown domain), botryoidal aragonite (white domain) and palisadic calcite (grey domain) respectively. Phase identifications of the domains were further confirmed by micro-Raman analyses. Since calcite, aragonite and ikaite have overlapping Raman bands, so their presence were identified by certain characteristic peaks/spectral bands occurring in the low and high frequency regions, in-plane bending modes, vibrational modes and symmetric stretch band of CO3 2−.

Trace-elemental analyses using LIBS technique showed that microcrystalline calcite domains were rich in potassium (K) whereas magnesium (Mg) was more in palisadic calcite domains. Strontium exhibited higher preference for aragonite domains. It became apparent from this study that Sr has a preference for aragonite over calcite.

Detail experimental results from sorption studies show that both the radionuclides exhibit moderate to low sorption coefficients within all the different domains of stalagmite. Sorption of 137Cs and 85+89Sr was more in calcite (both columnar and microcrystalline varieties) compared to aragonite at low pH (1–2) and it reversed with increasing pH. Interestingly K d values for 85+89Sr was much less compared to those for Cs-sorption. One possible reason behind this could be pre-existence of Sr within active sorption sites (as evidenced from LIBS data) in course of natural growth of speleothem that inhibited further incorporation of similar element within carbonate lattices under acidic environment.