Introduction

Zeolites and clay minerals are successfully applied as permeable reactive barriers (PRB) in radioactive waste management due to their structure and physical properties [1, 2]. Their excellent adsorption and ion exchange properties as well as their selectivity towards several cations especially for radionuclides such as 137Cs, 152Eu, 60Co etc. have received great attention in environmental technology [3,4,5,6,7,8,9].

Cesium exists in radioactive waste mainly as two isotopes, 137Cs (T1/2 = 30.17 years) and 135Cs (T1/2 = 2.3 × 106 years) known for their high yield and long-term environmental importance. Both belong along with 99Tc, 126Sn, 79Se, 90Sr, Pu, Am, to the most long-living isotopes in nuclear reactors. Cesium mainly appears in aqueous solutions in cationic form exhibiting high mobility in soils and bioconcentration/bioaccumulation ability in the food chains. Though stable cesium isotopes are hardly affecting the human health, the uptake of its radioactive isotopes can cause, through their radiation, serious health disorders and even death [6,7,8,9,10,11,12,13].

Cobalt can be usually released in the environment as radioactive isotopes 58Co (T1/2 = 71 days) and 60Co (T1/2 = 5.27 years) from the operation of nuclear power reactors. Cobalt is a naturally occurring element that exists in three oxidation states (0, + 2 and + 3), with the most prevalent being + 2 state under environmental condition. At low concentration levels, Co(II) is beneficial for living organisms since it participates in their metabolic processes. However, the US Environmental Protection Agency (EPA) as well as the International Agency for Research on Cancer (IARC) has grouped cobalt as known human carcinogen even at low concentrations. The permissible limits of cobalt in the irrigation water and live-stock watering are 0.05 and 1.0 mg L−1, respectively. Exposure to Co(II) beyond permissible levels may damage human health and cause low blood pressure, paralysis, diarrhea, vision impairment, bone, lung, heart, thyroid problems and genetic mutations. In addition, radioactive Co(II) isotopes cause hair loss, bleeding, sterility, coma and even death. A significant level of Co(II) in the soil can also cause acute toxicity to the plant kingdom [14,15,16,17,18].

Europium is a very reactive lanthanide that exists in the Eu3+ oxidative state. 151Eu (47.8%) and 153Eu (52.2%) are the two stable, naturally occurring isotopes of the element. Europium has no known toxic effects, but it is studied as a natural analogue of trivalent actinides like americium and curium. 241Am (T1/2 = 454 years) is a decay product of 241Pu (T1/2 = 14.4 years) and is an integral part of radioactive contamination. According to Seaborg's theory, the chemical structure of americium is similar to that of europium. Therefore, europium can be used as a chemical analog of americium for laboratory studies [19,20,21,22,23].

Sorption or ion-exchange using zeolites and clay minerals are well established techniques used for many years for the removal of cesium-137 and other radionuclides from aqueous solutions or the retardation of their migration in the environment. Regardless of their high cesium removal capacity, they also consist of the main parts of the engineered barrier systems (EBS) isolating the radioactive waste [4,5,6,7,8,9, 12,13,14, 19].

Concerning the Cs and Co-sorption by bentonites a significant number of articles has appeared and is still appearing in the literature [3,4,5,6,7,8,9, 14,15,16,17,18]. The sorption of europium onto bentonites and zeolites is less investigated [19,20,21,22,23]. Moreover, little is known about the effect of irradiation on the sorption of the above ions especially using low irradiation doses [24,25,26,27,28]. Akhalbedashvili et al., Galambos et al. and Keheyan et al. have referred to small doses of materials irradiation. Akhalbedashvili et al., reported that at doses around 5.50 kGy an ordering of structure showed while Galamboš et al. didn’t observe changes in the structure of Slovak bentonites at doses of 10 and 100 kGy γ-rays [2, 26, 28].

In this work the sorption properties of Greek bentonites and zeolites were investigated for repository conditions of low-level nuclear waste. Raw and modified material with low irradiation dose were tested for removal of Eu, Cs and Co from aqueous solutions using 152Eu, 137Cs and 60Co tracers and γ-ray spectroscopy. The sorption experiments were undertaken using the batch system and considering different parameters such as pH, metal concentration, competitive ions, and temperature. The structural changes of the sorbents due to irradiation and metal sorption were examined through XRD, FTIR and SEM/EDS. Sorption isotherms were reproduced by mathematical models and the thermodynamic parameters along with the results of the desorption tests gave clues about the sorption mechanism.

Experimental

Materials and methods

The bentonite used in this study was commercial (BENTOMINΕ KIMOLOS) from Kimolos island, Greece and the Zeolite from Petrota region in Thrace, Greece [9, 19, 29]. Both were in the form of grains with size less than 250 μm. The materials were γ-irradiated with a total dose of 3 and 6 kGy, using a 137Cs source.

The mineralogical composition of the Kimolos bentonite and Petrota zeolite used for the experiment work was determined through Powder X-ray Diffraction (pXRD) using a Philips PW 1710 diffractometer with Ni-filtered CuKα radiation. The pulverized material was scanned with a step size of 0.0131° 2θ in the 2θ interval 5–70°. The material was additionally characterized, before and after the sorption tests by Scanning Electron Microscopy/Energy Dispersive Spectroscopy (SEM/EDS) utilizing a JEOL JSM-840 equipped with an Oxford ISIS300 micro-analyzer and by Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR) in the range of 4000–400 cm−1 using a Thermo Scientific Nicolet iS20 spectrometer with a diamond reflectance crystal.

Sorption experiments

Cesium aqueous solutions (concentration range 50 to 1500 mg L−1) were prepared by dissolution of CsNO3 in bi-distilled water and spiked with a small activity of 137Cs radioactive tracer. Europium solutions (concentration range 50–500 mg L–1) were prepared by dissolving Eu(NO3)3 spiked with 152Eu in distilled water [9, 19]. Stock solutions of Co (concentration range 10–500 mg/L spiked with 60Co were also prepared from (CH3COO)2Co·4H2O. All the reagents were of analytical grade (Merck, Darmstadt, D). A water purification system (Millipore) with Elix and Milli-Q was used to provide ultra-pure water.

As it is also derived by the code MEDUSA, all three ions are in the form of positively charged species in a large pH region (Cs+ up to pH 12, Co2+ and Eu3+ up to pH 8) [30]. After preliminary experiments in different pH values, it was decided to perform the sorption study at pH 4 for Cs and Eu and pH 5 and 8 for Co as optimum pH values in order to avoid precipitation and to obtain the maximum sorption capacity. Batch experiments were carried out by contacting for 24 h 10 mL of the individual solutions with the adsorbent (dosage 1.2 g L−1) at ambient temperature (ca. 298 Κ). Preliminary experiments showed that the contact time was sufficient for the establishment of equilibrium. At the end of experiments and after the separation of the solid from the liquid phase by centrifugation (4000 rpm for 10 min), the equilibrium pH of the supernatant solutions was measured. The concentration of europium, cesium and cobalt was determined by gamma-ray spectroscopy using the 121.8, 661.6, and 1173.2 keV energy peaks for 152Eu, 137Cs and 60Co, respectively with 5% uncertainty. A CANBERRA ReGe detector (efficiency 23%, resolution 1.8 keV for the 1332.5 keV 60Co gamma-radiation) connected with a standard data acquisition and analysis system (EG&G Maestro) was used for the measurements.

The obtained data were used to calculate the metal uptake and construct the corresponding isotherms. The uptake data were modelled using the Langmuir and Freundlich linear isotherm Eqs. (1–2) [31, 32]. The Langmuir model assumes a localized monolayer adsorption on a fixed number of adsorption sites of equal energy (homogenous sorption). In contrast, the Freundlich model considers a surface heterogeneity not reaching a limited sorption capacity.

$${\mathbf{Langmuir}} \, \left( {\mathbf{L}} \right) \,\,\,\,\,\,\,\,\, \frac{{C_{e} }}{{q_{e} }} = \frac{1}{{K_{L} q_{\max } }} + \frac{1}{{q_{\max } }}Ce$$
(1)
$${\mathbf{Freundlich}} \, \left( {\mathbf{F}} \right) \,\,\,\,\,\,\,\,\, In\,q_{e} = In\,K_{L} + \frac{1}{n}In\,Ce$$
(2)

qe and Ce in the above equations are the equilibrium metal concentrations in the solid (in mg g−1) and liquid (in mg L−1) phase, respectively, qmax the maximum sorption capacity (in mg g−1), KL and KF the Langmuir and Freundlich equilibrium constants and n, in the case of Freundlich equation, a parameter characterizing the system heterogeneity.

The effect of competitive ions on metal sorption was investigated using solutions of NaNO3 (0.1 mol L−1).

Thermodynamic studies

Thermodynamic data were obtained for 293, 298, 308 and 318 K. The measurements were undertaken during 180 min using 250 mL of a 250 mg L−1 cobalt solution. At predetermined time intervals, 5 mL of the solutions were withdrawn by a syringe, and their metal concentration was determined, as previously mentioned, by gamma-ray spectrometry. The Gibbs free energy change, ΔG°, determines the spontaneous nature of the adsorption process and is given by the Eq. (3) where R (J mol−1 K−1) is the universal gas constant, T (K) is the absolute temperature and KD = qe/Ce is the distribution coefficient [33, 34].

$$\Delta G^{^\circ } = - RT\,In\,K_{D}$$
(3)

ΔH° and ΔS° are associated with ΔG° with Eq. (4).

$$\Delta G^{^\circ } = \Delta H^{^\circ } - T\Delta S^{^\circ }$$
(4)

The negative values of ΔG° found, confirm the feasibility of the process and its spontaneous nature at a given temperature.

The value of ΔH° determines the nature of the sorption: 2.1–20.9 kJ mol−1 for physical and 20.9–418.4 kJ mol−1 for chemical sorption. It was obtained from van’t Hoff Eq. (5) by combining Eqs. (3) and (4):

$$In\,K_{D} = \frac{\Delta S}{R}^{^\circ } - \frac{{\Delta H^{^\circ } }}{R}\frac{1}{T}$$
(5)

Desorption experiments

The environmental compatibility of the sorbents after loading was assessed by the Toxicity Characteristic Leaching Procedure (TCLP, U.S. EPA Method 1311). Desorption studies were conducted using Extraction Fluid #1 (5.7 mL glacial CH3COOH and 64.3 mL of 1 mol L−1 NaOH successively added to distilled water subsequently diluted to a total volume of 1 L) for zeolite, and Extraction Fluid #2 (5.7 mL glacial CH3COOH diluted with distilled water to a total volume of 1 L) for bentonite. Both fluids were in contact with 20 mg of the appropriate adsorbent, after loading it with 1000 and 250 mg L−1 solution of Cs+ and Co2+ respectively and shaking for 18 h [35].

Results and discussion

Sorption studies

The effect of the initial metal concentration was investigated by constructing the sorption isotherms, which describe the equilibrium relationship between the sorbent and the sorbate. The determined isotherms for the cesium sorption, before and after irradiation onto the Kimolos bentonite are given in Fig. 1 while the corresponding to Petrota zeolite in Fig. 2.

Fig. 1
figure 1

Sorption isotherms of Cs onto raw bentonite, 3 kGy and 6 kGy dose irradiated bentonite (pH 4, T:293 K)

Full size image
Fig. 2
figure 2

Sorption isotherms of Cs onto raw zeolite, 3 kGy and 6 kGy dose irradiated zeolite (pH 4, T:293 K)

Full size image

In the case of bentonite with 6 kGy irradiation dose the adsorption capacity of Cs increased after the concentration of 250 mg L−1, compared to the raw material while the 3 kGy irradiation dose resulted in a slightly lowered uptake. In most literature studies decrease or absence of effect was observed in Cs adsorption capacity after application of irradiation to bentonite [27, 28, 36] which is attributed to partial loss of the ion exchange sites (hydrated cations Na+, K+ and Ca2+ present in the interlayers). An increase in adsorption capacity of Cs was observed in a study of Belkhiri et al. after application of 1000 kGy to Algerian bentonite, which was attributed to a probable expansion of the bentonite’s interlayer space [37]. Also, in Allard & Calas, it is mentioned that in experiments where gamma irradiation was applied to smectites with doses 100 MGy and 1.0 GGy the cation exchange capacity was increased which can result in higher adsorption capacity [25]. In the current study the irradiation doses used are lower and the increase in adsorption of irradiated material could be probably attributed to the expansion of the bentonite’s interlayer space after irradiation.

Competing Na+ ions found to decrease significantly (25–28%) the adsorption capacity of Cs in all cases showing the importance of studying the simultaneous sorption of two or more ions to determine the performance of a sorbent material in real radioactive waste conditions [19].

In the case of Cs sorption on zeolite, the changes in adsorption capacity are not significant after irradiation. Other researchers have reported decrease in adsorption capacity but using high irradiation doses. For example, Belkhiri et al. studied the effect of irradiation on Cs adsorption onto a synthetic NaX Zeolite and found that after irradiation with doses 100, 500, 1000 kGy the pore volume and surface area were decreased compared to the raw material, thus adsorption was inhibited. The decrease was attributed to partial loss of crystallinity due to a dealumination process causing enhanced removal of water inducing itself a partial collapse of the framework [37]. Akhalbedashvili et al. studied the effect of gamma rays at doses 200, 700 kGy on Cs adsorption onto Armenian clinoptilolite and found that ion exchange sorption decreased. It was also mentioned that in small doses (5.5 kGy) the structure gets ordered but at higher doses the crystal lattice gets disordered [26].

The determined isotherms for the cobalt sorption, before and after irradiation by the Kimolos bentonite and Petrota zeolite are given in Figs. 3 and 4 respectively. Figure 3 also shows the effect of competitive ions on the adsorption capacity of raw bentonite. In the case of bentonite after irradiation with 6 kGy dose, the adsorption capacity decreased. Decrease in Co sorption after irradiation of bentonite was also observed by Holmboe et al. [27]. In addition, competing Na+ decreased the adsorption capacity for Co compared to the raw bentonite due to antagonistic effects.

Fig. 3
figure 3

Sorption isotherms of Co onto raw bentonite, 6 kGy dose irradiated bentonite and raw bentonite with competitive ions (pH 5, T:293 K)

Full size image
Fig. 4
figure 4

Sorption isotherms of Co onto raw Zeolite and 6 kGy dose irradiated zeolite (pH 5 and 8, T:293 K)

Full size image

Concerning the effect of 6 kGy irradiation dose on Co adsorption for zeolite, slight changes can be observed leading to the conclusion that the irradiation dose used in the current study was not enough to affect zeolite’s sorption properties.

Adjustment of the experimental results to Langmuir and Freundlich models showed that in the case of bentonite the predominant model is the Langmuir indicating monolayer coverage. For Co-sorption onto zeolite the best fitting observed to the Freundlich model (Table 1) indicating surface heterogeneity and different sorption mechanism especially when the initial pH is set to 8, where the uptake was higher [38].

Table 1 Parameters of the Langmuir and Freundlich models for cobalt sorption
Full size table

Figure 5 shows a comparative diagram of the adsorption capacities for Eu, Cs, Co on raw and 6 kGy irradiated bentonite and zeolite. As can be seen raw bentonite performs better than zeolite especially for Eu and Cs and after irradiation its Eu- and Cs-adsorption capacity enhanced while for Co, reduced.

Fig. 5
figure 5

Comparative histogram for adsorption capacity of Eu, Cs, Co of raw and irradiated bentonite, zeolite (Eu: pH 4, C 500 mg L−1; Cs: pH 4, C 1500 mg L−1; Co: pH 5 and 8, C 500 mg L.−1)

Full size image

Alkhalbedashvili et al. compared the effect of electron and gamma irradiation on ion exchange capacity of Sr2+ and Cs+ for clinoptilolite and observed that although in both cases it was decreased, the adsorption capacity depends more on the ionic radius than the dose of irradiation [26]. From the results in the current study, it was concluded that irradiation affects differently the adsorption capacity of different metals for the same materials, supporting the above theory. Tables 2 and 3 contain literature results for Cs-, Co- and Eu-sorption onto bentonites and zeolites for comparison with the findings of this work where it is shown that Greek minerals exhibit significant sorption efficiency.

Table 2 Literature data for cesium sorption
Full size table
Table 3 Literature data for cobalt and europium sorption
Full size table

Table 4 presents the thermodynamic parameters (enthalpy, entropy, Gibbs energy) and distribution coefficients for Co-sorption onto zeolite, calculated for the temperatures 293, 298, 308 and 318 K. The results indicate that the adsorption process for Co is spontaneous and exothermic in nature and the magnitude of ΔH° suggests ion exchange and chemical sorption.

Table 4 Distribution coefficients for the sorption of Co onto Zeolite
Full size table

The negative value of ΔS° in both cases suggests the decrease of disorder at the sorbent to aqueous phase interface. The values though are too close to zero suggesting that the modifications to the sorbent’s structure are minimal and the phenomena are stable and non-reversible. In the literature the Co-sorption onto bentonites and zeolites, investigated at different pH values (5 and 6) and it is reported as endothermic or exothermic process [17, 18, 42,43,44,45,45]. The Cs- and Eu-sorption onto raw Kimolos bentonite found to be slightly affected by the temperature and thermodynamic studies revealed a spontaneous process exothermic in the case of Cs, and endothermic in the case of Eu [9, 19]. In most of the articles ion exchange is dominated on Cs-sorption while in the case of Eu, many researchers observed complex mechanism combining cation exchange and surface complexation [20,21,22,23].

Characterization of the materials

The pXRD patterns indicated that the raw bentonite used for the experimental work mainly consisted of (86%) montmorillonite Ca-rich as can be seen in Fig. 6a, and it is semicrystalline. Small amounts of quartz and feldspars were also present. The main diffraction peaks at 2θ = 5.84, 17.18, 19.86, 29.5, 34.84, 54.38, 61.94° corresponded to (001), (003), (100), (005), (006), (210), (0010) planes of montmorillonite, respectively [46]. The main reflection of d (001) spacing in the region of 5° < 2θ < 8° corresponds to the inter lamellar distance of Mt at 15.2 Å [47]. After irradiation (Fig. 6b) slight differences are observed, concerning the peak position or diffraction intensity, indicating that the main components did not change. Concerning zeolite, the XRD-patterns indicated that it mainly consisted of (89%) HEU-type Zeolite as can be seen in Fig. 7a. Small amounts of mica, quartz, clay minerals, cristobalite and feldspars were also present. Slight differences are observed in the case of zeolite after γ-irradiation (Fig. 7b), indicating, as in the case of bentonite, partial expansion of the interlayer (or channels) space which affects the sorption capacity.

Fig. 6
figure 6

pXRD spectra of a raw bentonite, b irradiated bentonite (6 kGy)

Full size image
Fig. 7
figure 7

pXRD spectra of a raw zeolite, b irradiated zeolite (6 kGy)

Full size image

The characteristic microstructure of montmorillonite with microaggregates and small particles irregularly distributed is shown in SEM image (Fig. 8a) [9]. EDS results showed that the raw bentonite was composed averagely of 60.85% O, 2.59% Mg, 10% Al, 24.74% Si, 1.2% K, 0.53% Ca, 0.85% Fe. After irradiation (Fig. 8b) no important morphological changes can be observed. After Cs adsorption onto raw (Fig. 8c) and irradiated bentonite (Fig. 8e) inside the grain-like morphology appeared plate-like, compact and almost flat sheets. The same observation in the case of Cs-sorption onto raw bentonite was made in SEM images by Muslim et al. [48]. Similar images can also be seen in Zhang et al. [49], where the sorption of Pb(II) onto alkaline Ca-bentonite was investigated. After irradiation the fragments are increased. The EDS spectrum proved the existence of Cs in both cases.

Fig. 8
figure 8

SEM images for: a Raw bentonite, b Bentonite with 6 kGy irradiation dose, c Raw bentonite after Cs sorption, d Raw bentonite after Co sorption, e Irradiated bentonite after Cs sorption, f Irradiated bentonite after Co sorption

Full size image

In the case after Co adsorption onto raw (Fig. 8d) and irradiated (Fig. 8f) bentonite, the initial grain like morphology changed completely to a plate like, compact, smooth, flat morphology and the existence of Co was confirmed in the EDS spectrum.

SEM image of the raw zeolite (Fig. 9a) shows small and large crystals having monoclinic symmetry of blades and laths with few of them having a coffin shape [50]. After cobalt adsorption (Fig. 9b), similar to the case of raw and irradiated bentonite, the morphology changed to plate-like, compact, smooth and flat.

Fig. 9
figure 9

SEM images for: a Raw Zeolite, b Zeolite after Co2+ adsorption

Full size image

The investigation by ATR-FTIR didn’t reveal any difference in the samples before and after sorption except for zeolite after Co-sorption at pH 8. Figure 10 shows the FTIR spectra of Petrota zeolite before and after cobalt sorption at initial pH 5 and 8. The spectrum of raw zeolite exhibits three characteristic bands at 1029, 791 and 588 cm−1, that are attributed to the internal asymmetric T–O–T stretching vibration, the Si–O–Si vibration of the SiO2 moiety and the T–O–T bending vibration or the vibrations of 4- and 6-membered rings, respectively. The broad band at around 3500 cm−1 and the band at 1620 cm−1 are due to the stretching and bending vibrations of the OH groups of water molecules present in the zeolitic framework [51,52,53].

Fig. 10
figure 10

FTIR spectra for: a Raw zeolite and Co-loaded zeolite at initial pH = 5, b Raw zeolite and Co-loaded zeolite at initial pH = 8 (fingerprint region)

Full size image

Co-sorption at pH 5 does not alter the characteristic bands in the spectrum (Fig. 8a). On the contrary, when the pH is set to 8 (Fig. 8b) some changes can be observed. A new band appears at 1380 cm−1, most probably attributed to the lattice vibrations as suggested by Nyberg et al. and to the excess alumina in the pores [54]. The band at 588 cm−1 is shifted to higher wavenumbers, while the main band at 1029 cm−1 shows a tendency to split into two new peaks. With regards to the latter observation, Sobalík et al. noted similar results when investigating the presence of transition metal ions in zeolites [55]. The splitting of the main band can be attributed to the binding of Co2+ ions to framework oxygen atoms which causes a perturbation and deformation of the T–O–T bonds, leading to a different vibration frequency. The splitting into two separate bands is indicative of different cobalt-oxygen bonding strengths, or different cobalt-binding sites in the framework [56]. These observations indicate that chemisorption dominates for the Co-sorption at pH 8. The same mechanism was reported by other researchers who found that inner-sphere complexation for Co-sorption is strengthened with increasing pH [57].

Desorption tests

The study of the environmental compatibility (TCLP method) showed safe disposal of the loaded materials. For Cs-loading the percentage of the metal detected in the leachate was around 20% while for Eu was undetectable. For Co-loading the obtained results were 14% for bentonite and 12.5% for zeolite for sorption at pH 5, while at pH 8, 11.5 and 9.5% for bentonite and zeolite respectively, demonstrating in this case also, the dominance of a different mechanism for the Co-sorption at pH 8. As it is known the desorption of cations adsorbed by ion exchange becomes easier than those bound by inner-sphere complexation and therefore desorption implies the stability and irreversibility of the adsorption process and can provide mechanism information [57].

Conclusions

  • Sorption tests of Cs, Eu and Co onto Greek bentonite and zeolite as well as leaching studies showed that they can efficiently remove the above hazardous cations from aqueous solutions.

  • The presence of Na+ competing ions decrease the adsorption capacity of the tested materials.

  • Irradiation affected differently the adsorption capacity of Cs, Eu, Co for the same materials showing that its effect may depend on the ion’s size. In the case of Eu and Cs the small dose of irradiation applied resulted in an increase of the adsorption capacity, in contrast to Co.

  • Generally, zeolite was less affected by irradiation compared to bentonite, but bentonite performed better as sorbent material in most cases.

  • XRD investigation showed that the dose of irradiation used did not affect the main components of the studied sorbent materials and led to slight structural changes.

  • SEM investigations showed that irradiation did not affect the morphological characteristics of the materials, whereas the Cs- and Co-loaded adsorbents exhibited changes in morphology.

  • The adjustment of the sorption isotherms to Langmuir and Freundlich models as well as the thermodynamic data and desorption results and furthermore the FTIR investigation suggest ion exchange in the case of Cs-sorption. Co-sorption seems to be governed by chemisorption which is strengthened with increasing pH while in the case of Eu a complex mechanism seems to occur following both cation exchange and surface complexation.