Introduction

Despite prevention measures, anthropogenic radionuclides are still released into the environment resulting in contamination of soils and water bodies [1]. Among the main sources of water bodies pollution by artificial radionuclides are nuclear weapon tests [2, 3], daily work of enterprises treating irradiated nuclear fuel [4] as well as nuclear accidents [5]. Isotopes of cesium and strontium play the leading role in pollution of natural water bodies and 90Sr is one of the most dangerous one due to its relatively long half-life (29.12 years), high affinity to bone tissue and high beta energy of its progeny 90Y. Presence of 90Sr in liquid radioactive waste as well as its release into natural waters makes it necessary to search for methods of effective separation of 90Sr from various aqueous solutions with the aim of radiochemical analysis or deactivation. This process is complicated because of interference from alkaline earth cations presenting in both radioactively contaminated natural waters and liquid radioactive waste. Concentration of calcium in natural waters varies within 1–100 mg/L in surface waters, up to 10–250 mg/L in underground waters and up to 600 mg/L in salt lakes and seawaters [6,7,8,9,10,11,12].

A number of sorbents based on various sparingly soluble compounds are developed for elimination of strontium from natural waters and liquid radioactive waste. Among sparingly soluble compounds used in these sorbents, there are granulated and modified manganese dioxide including mesoporous manganese dioxide and birnessite [13,14,15,16,17], titanium hydroxide [18], calcium and magnesium phosphates and hydrophosphates [19], zirconium phosphate [20, 21] as well as titanosilicates [14, 22,23,24]. Possibility of using the sorbents for strontium preconcentration in methods of radiochemical analysis of aqueous samples is evaluated for some of these sorbents [25,26,27,28,29].

Granulated inorganic sorbents seem to be the most prospective for decontamination of radioactively contaminated waters. These sorbents provide distribution coefficients of strontium within 102–104 mL/g, and their static exchange capacity may reach several hundred mg/g. Possibility of their use should be determined by their cost, availability (commercial or lab sample) and enough mechanical strength for exploitation in column conditions. Extraction chromatographic resins highly selective for strontium have found application in analysis of aqueous samples. Bezhin et al. [29] reported a possibility of using various granulated inorganic sorbents for determination of radiostrontium in seawater samples. A sorbent Modix (SPE Eksorb Ltd) is chosen as the most promising one, however the development of an analytical method is still in progress.

Strontium carbonate is another one sparingly soluble salt of strontium. Despite is has a higher solubility than strontium phosphate, the study of strontium sorption by carbonate-containing sorbents is also interesting. Solubility product of strontium carbonate is intermediate between the respective values for strontium sulfate and phosphate: SrSO4 > SrCO3 > Sr3(PO4)2. This work is aimed to a study of physicochemical and sorption properties of a carbonate-containing zirconium hydroxide (T-3K sorbent produced by JSC Inorganic Sorbents, Russia), selectivity of strontium sorption as well as a possibility of using this sorbent for preconcentration of 90Sr from fresh water and seawater and for decontamination of aqueous media.

Experimental

The T-3K sorbent was an inorganic polymeric material with amorphous structure, which was synthesized via sol–gel method as white spherical granules. The sorbent can be easily used in column conditions due to its spherical form, chemical stability and mechanical strength. The grain size of the sorbent was 0.4–1.0 mm, bulk density was 1.05 g/mL and the mechanical strength was at least 5 MPa.

The method of synthesis of the T-3K sorbent included treatment of gel spheres of the hydrated zirconium dioxide (T-3 sorbent) by sodium hydrocarbonate solution [30]. According to the producer’s data, chemical reactions of the surface modification can be described by the following Eqs. (1), (2):

$${\equiv} {\text{Zr-OH}} + {\text{H}}_{2} {\text{CO}}_{3} \to \equiv {\text{Zr-CO}}_{3} {\text{H}} + {\text{H}}_{2} {\text{O}}$$
(1)
$${\equiv} {\text{Zr-OH}} + {\text{NaHCO}}_{3} \to \equiv {\text{Zr-CO}}_{3} {\text{Na}} + {\text{H}}_{2} {\text{O}}$$
(2)

According to the suggestions of the developers, its chemical composition is surface-modified hydrated zirconium dioxide, Zr(OH)3.5(NaCO3)0.5 [30].

The study of physicochemical properties of the T-3K sorbent

Physicochemical properties of the T-3K sorbent were studied in comparison with the respective properties of the T-3 sorbent. Analysis of element composition of the sorbent was performed using an ARL QUANT’X (USA) energy dispersive X-ray fluorescence spectrometer (the detection limit is 10 ppm, relative error is 10–20%). Infrared spectrum of the sorbent was obtained using Vertex-70 IR-Raman spectrometer with RAM-II device (Bruker, Germany). To make interpretation of the IR spectrum easier, we have additionally obtained the IR spectrum of the hydrated zirconium dioxide (a T-3 sorbent), which is the raw material for synthesis of the T-3K sorbent.

Surface area of the sorbents was determined by the method of low temperature nitrogen sorption using the Nova 1200e high-speed surface area analyzer (Quantachrome, USA). Temperature of degassing was 100 °C, the degassing length of time was 2 h, and the sorbent weight was 250 ± 5 mg.

The effect of pH on strontium sorption by the T-3K sorbent

The dependence of strontium sorption on pH was studied using a weakly mineralized water (tap water) with addition of 1 mg/L of stable strontium and spiked with 90Sr as a radiotracer (the activity concentration was 10 kBq/L). Sorbent weight/solution volume ratio was 50 mg per 30 mL. Time of the phases contact was 1 week. Tap water was stored and filtered before the experiments to eliminate iron hydroxide colloids. Concentration of stable strontium was adjusted by addition of SrCl2 solution with known concentration. The pH was adjusted by addition of HCl or NaOH solutions. Measurements of 90Sr were performed using a semiconductor alpha/beta radiometer UMF-2000 (Doza, Russia) after equilibrium 90Y ingrowth. The degree of strontium sorption S was calculated in accordance with the Eq. (3):

$$S = \frac{{I_{{{\mathrm{in}}}} - I_{{{\mathrm{end}}}} }}{{I_{{{\mathrm{in}}}} - I_{{\mathrm{b}}} }},$$
(3)

where \(I_{{{\mathrm{in}}}}\), \(I_{{{\mathrm{end}}}}\) are initial and final count rates of the solution, cpm;\(I_{{\mathrm{b}}}\) is background count rate, cpm.

The effect of Ca2+ and Na+ concentrations on strontium sorption

In the experiments, distilled water with addition of various concentrations of Ca2+ and Na+ ions was used. Concentrations of ions were adjusted to the concentration range typical for natural surface and ground waters and soil solutions by addition of CaCl2 and NaCl solutions. Concentrations of calcium varied from 1 to 300 mg/L, pH values in solutions before sorption were 5.6 ± 0.1. Concentrations of sodium varied from 0.02 to 230 mg/L, pH values in solutions before sorption were 6.5 ± 0.1. Strontium concentration was adjusted to 1 mg/L in all solutions by addition of SrCl2 solution with a known concentration. Temperature of the solution was 22 ± 1 °C. The sorbent weight was 50 mg and the solution volume was 30 mL. The experiments were performed in three replicates. Time of phase contact was 1 week. Chemical composition of natural waters as well as strontium and calcium concentrations in the solutions before and after sorption were determined using the NexION 350X mass spectrometer with inductively coupled plasma. Data treatment was performed using TotalQuant software. Distribution coefficients of strontium and calcium Kd (mL/g) were calculated according to the Eq. (4):

$$K_{{\mathrm{d}}} = \frac{S}{1 - S} \cdot \frac{V}{m},$$
(4)

where S is degree of sorption, dimensionless; V is volume of the solution, mL; m is weight of the sorbent, g.

Separation factor R (Sr/Ca) was calculated according to the Eq. (5):

$$R = \frac{{K_{{\mathrm{d}}} ({\text{Sr}})}}{{K_{{\mathrm{d}}} ({\text{Ca}})}},$$
(5)

where Kd (Sr) is the distribution coefficient of strontium, mL/g; Kd (Ca) is distribution coefficient of calcium, mL/g.

Study of selectivity of strontium sorption by the T-3K sorbent over alkaline and alkaline earth ions in weakly mineralized water and seawater

Sorption of strontium from weakly mineralized (tap) water and seawater (Mediterranean Sea). Sorbent weight/solution volume ratio was 500 mg per 100 mL. Time of phase contact was 1 week. Concentrations of alkaline and alkaline earth elements were determined by ICP-MS. Distribution coefficients of the alkaline and alkaline earth elements as well as Sr/Metal separation factors were calculated as described above.

Sorption of 137Cs was studied using low mineralized water with concentration of stable cesium of 0.01 mg/L, which was adjusted by addition of CsCl solution with known concentration. Sorbent weight/solution volume ratio was 50 mg per 30 mL. Time of phase contact was 1 week. Activity concentration of 137Cs radiotracer was 20 kBq/L. Measurements of 137Cs were performed using a semiconductor alpha/beta radiometer UMF-2000 (Doza, Russia).

The isotherms of strontium sorption

The isotherms of strontium sorption were obtained using tap water and seawater (Mediterranean Sea) with addition of 10–9–3 g/L of stable strontium; 90Sr was used as a radiotracer. Time of phase contact was 1 week to determine equilibrium sorption characteristics. The sorbent weight was 50 mg and the solution volume was 30 mL.

Measurements of 90Sr were performed using a semiconductor alpha/beta radiometer UMF-2000. Mathematical treatment of the linear sections of the isotherms performed using the least squares method; the slope of the linear sections and distribution coefficients of strontium were calculated.

Study of 90Sr desorption from the T-3K sorbent

Desorption of 90Sr from the T-3K sorbent nu acidic solutions was studied under batch conditions. The effects of acid type (HCl or HNO3), acid concentration and time of phases contact were studied. The sorbent was loaded by 90Sr under batch conditions for 1 week contacting 500 mg with 100 mL of a solution. After loading, the sorbent was washed with distilled water, dried at a room temperature and then contacted with 10 bed volumes (BV) of the stripping solutions. Samples aliquots were collected after time intervals within 24 h, analyzed, and the degree of desorption was calculated.

Results and discussion

Chemical composition and porosity of the T-3K sorbent

Table 1 shows chemical composition (as element oxides) of the T-3 and T-3K sorbent. The content of zirconium hydroxide (after water deduction) was 46% for the T-3K sorbent and 61% for the T-3 sorbent; the data agree with certificate data for these sorbents. All other oxides were as trace impurities. For example, sodium content in the T-3K sorbent was 1.78% and aluminum content was 1.42%.

Table 1 Chemical composition (as element oxides) of the T-3K sorbent, %wt
Full size table

The transmittance IR spectrum of the T-3 and T-3K sorbent is given at Fig. 1.

Fig. 1
figure 1

Transmittance IR spectrum of the T-3K sorbent

Full size image

A wide adsorption band at 2800–3700 cm−1, which is typical for stretching vibrations of OH— groups [31], was observed in the IR spectrum. The adsorption peak at 1635 cm−1 was conditioned by bending vibrations of absorbed water molecules [32]. Presence of absorbed carbon dioxide molecules was confirmed by spectral adsorption bands at 2364 cm−1 (asymmetric stretching vibration) and 668 cm−1 (symmetrical bending vibration) [33]. The sharp peaks at 2859 cm−1 and 2924 cm−1 indicated vibrations of the Zr–O bond [34]. In contrast to the T-3 sorbent, the IR spectrum of the T-3K contains the absorption peaks at 1384 cm−1, 1076 cm−1 and 840 cm−1 indicating vibrations of the CO3 group [35]. Adsorption peaks at 1080, 1040 and 845–850 cm−1 correspond to various types of surface carbonate groups [36]. The adsorption peak at 1390 cm−1 indicates the presence of the CO32− group [36], the peaks series at 1080–1030 cm−1 and 1337 cm−1 can be associated with bidentate carbonates on the surface of zirconium dioxide [36,37,38].

Table 2 shows the parameters of the pore structure of the T-3 and T-3K sorbents being determined by the low temperature nitrogen adsorption method.

Table 2 The parameters of the pore structure of the sorbents
Full size table

Figure 2 presents the isotherm of nitrogen adsorption and size distribution of micropores for the T-3K sorbent.

Fig. 2
figure 2

Isotherm of nitrogen adsorption (a) and differential distribution of micropores size (b) for the T-3K sorbent

Full size image

The isotherm of nitrogen adsorption was Type I of Langmuir’s isotherm. The isotherm was convex relative to the P/P0 axis and the amount of adsorbate approached the limit as P/P0 increased to 1. This type of isotherm is basically typical for microporous materials. The analysis of micropores distribution of the T-3 и T-3K sorbent was performed using the V-t method. The results have shown that the T-3 sorbent was a micropore sorbent with the surface are of 144 ± 7 m2/g. Chemical modification of the T-3 sorbent by carbonate resulted in a decrease of its porosity as well as the area and volume of micropores. The contribution of external surface increased after modification.

Figure 2 showed a very narrow micropores distribution T-3K sorbent. The average diameter of micropores was 1.5–2 nm, the volume of micropores was (9.6 ± 1.0) × 10–3 cm3/g and the area of micropores was 14 ± 1 m2/g. Thus, the sorbent was mainly microporous with a small contribution of mesopores with the average diameter of 2–5 nm. The volume of mesopores was (9.3 ± 0.9) × 10–4 cm3/g. The effective diameter of pores was 3.1 ± 0.3 nm. The surface area of the sorbent was determined using the BET method. This value was 95 ± 5 m2/g and it was mainly conditioned by the contribution of external surface area (81 ± 4 m2/g). Therefore, we suggest that reactions of the sorbent with strontium should occur mainly on the surface of the sorbent.

The study of pH effect on strontium sorption by the T-3K sorbent

Figure 3 shows the pH dependence of the degree of strontium sorption by the T-3K sorbent.

Fig. 3
figure 3

The effect of pH in strontium sorption by the T-3K sorbent: a the dependence of sorption degree on pH; b the dependence of pH after sorption on the pH value in the solution before sorption

Full size image

The sorbent possessed buffer capacity that resulted in pH shifting to pH 7–8 during sorption of strontium at the initial pH range of 4–10. Because of this significant buffer capacity, we have failed to obtain experimental points at the pH range of 4–6 even after triple acidification of the solution during 3 days of sorption. The dependence showed a decrease of strontium sorption by the T-3K sorbent with the decrease of pH value that is typical for weak acid cation exchangers. The maximal sorption of strontium achieved at pH 7–10; distribution coefficient of strontium was Kd = (2.08 ± 0.04) × 103 mL/g. Thus, the working pH range for the T-3K was optimal for strontium separation from seawater and fresh waters.

The study of selectivity of strontium sorption by the T-3K sorbent in presence of alkaline and alkaline earth elements

Selectivity of strontium sorption by the T-3K sorbent over the major components of natural waters was studied. Figure 4 shows the dependences of distribution coefficients of Sr and Ca as well as Sr/Ca separation factors on calcium concentration in the solution. The maximal distribution coefficients of Sr and Ca achieved in solutions at calcium concentration up to 1 mg/L; the respective values reached 1.5 × 105 and 1.5 × 104 mL/g respectively. The maximal Sr/Ca separation factor reached 13 giving a possibility of selective separation of strontium from calcium.

Fig. 4
figure 4

The effect of calcium concentration on distribution coefficients of Sr and Ca (a) and Sr/Ca separation factor (b). Concentration of stable strontium was 1 mg/L

Full size image

In case of calcium concentration in the solution more than 50 mg/L, distribution coefficients of Sr and Ca remained nearly invariable: lg Kd (Sr) = 4.2 ± 0.2 mL/g and lg Kd (Ca) = 3.9 ± 0.3 mL/g The average Sr/Ca separation factor was 2.0.

Figure 5 presents the dependence of strontium distribution coefficient on sodium concentration in the solution. Increase of sodium concentration resulted in a decrease of strontium distribution coefficient at sodium concentrations up to 0.0001 mol/L (2.3 mg/L); however, the Kd value was invariable (lg Kd (Sr) = 3.8 ± 0.1 mL/g) at higher concentrations of sodium.

Fig. 5
figure 5

The effect of sodium concentration on distribution coefficient of Sr. Concentration of stable strontium was 1 mg/L

Full size image

A high distribution coefficient of strontium in presence of significant amounts of sodium and calcium allows using the T-3K sorbent for both strontium preconcentration from natural waters in radioanalytical schemes and decontamination of radioactively polluted natural waters and liquid radioactive waste from strontium radionuclides.

Sorption of strontium from low mineralized and seawater was studied under batch conditions. Table 3 shows concentrations of major alkaline and alkaline earth ions and parameters of their sorption by the T-3K sorbent.

Table 3 Concentrations of major alkaline and alkaline earth ions and parameters of their sorption by the T-3K sorbent
Full size table

It is obvious that selectivity of sorption of various cations by the T-3K was conditioned by solubility of respective carbonates. Sodium and potassium carbonates possess high solubility resulting in low sorption of these ions. According to literature data [39], solubility products of alkaline earth carbonates follow in the series: MgCO3 = 7.9 × 10–6; CaCO3 = 4.4 × 10–9; SrCO3 = 5.3 × 10–10; BaCO3 = 4.9 × 10–9. Affinity of the T-3K sorbent for alkaline earth ions corresponds to the reverse trend of carbonates solubility: Sr > Ba > Ca ≫ Mg. Some deviation from this trend was observed in case of barium that can be probably explained by its low concentration in natural waters. Thus, the sorbent shows the highest selectivity for strontium among alkaline earths due to the lowest solubility product of strontium carbonate. Therefore, the sorbent provides selective separation in the pairs Sr/Mg, Sr/Ca and Sr/Ba allowing good strontium sorption from waters with elevated calcium content.

Separation factors of Sr/Ca obtained in both low mineralized water and seawater with natural concentrations of alkaline and alkaline earth elements were like those obtained in pure CaCl2 solutions (see Fig. 4). Iron, potassium, magnesium, and barium were not separated from seawater.

Selective preconcentration of strontium is preferable for development of a method for determination of 90Sr in natural waters. Among the main radionuclides interfering 90Sr measurement by beta radiometry is 137Cs, which is a long-lived beta emitting radioactive contaminant presenting natural waters together with 90Sr. In case of 137Cs sorption by the T-3K sorbent, this radionuclide will contaminate the measuring source resulting in a systematic error of measurement of 90Sr activity. However, cesium carbonate is a well soluble compound; therefore, considering mechanism of ions sorption by the T-3K, sorption of cesium is expected to be low.

Sorption of 137Cs by the T-3K sorbent was studied in comparison with the T-3 sorbent under the same experimental conditions. The T-3 sorbent has OH– groups as the sorption sites and thus it shows very weak selectivity for cesium. The distribution coefficient of cesium for the T-3 sorbent was 30 mL/g, and the static exchange capacity was 0.5 mg/g. The experimental test of 137Cs sorption by the T-3K sorbent from low mineralized water under batch conditions has shown that the degree of cesium sorption was very near to error of measurement, S = 0.13 ± 0.10. Chemical modification of the T-3 sorption with sodium bicarbonate with the aim of increasing its selectivity for strontium resulted in replacing the OH– groups by –CO3H and –CO3Na groups (see Eqs. 1 and 2) and thus in depression of the ability to adsorb cesium. As it was shown above, chemical modification resulted in a decrease of the surface area of the sorbent and in an increase of the contribution of the external area. Therefore, the T-3K sorbent provided selective separation of strontium from 137Cs.

The study of strontium sorption from natural water with various salt content

Isotherms of strontium sorption were also obtained for both T-3 and T-3K sorbent. The T-3 sorbent did not separate strontium from weakly mineralized water. Among nine solutions with various strontium concentrations in the experiment, the degrees of strontium sorption in two points were 0.07 ± 0.02 and 0.09 ± 0.04 at initial concentrations of strontium of 10–6 and 10–1 mg/L respectively. At the same time, sorption of strontium was nearly zero in other seven solutions. In contrast to the T-3 sorbent, the T-3K showed well selectivity for strontium.

Isotherms of strontium sorption by the T-3K sorbent from low mineralized water (pH 7.5 ± 0.1) and seawater (pH 8.0 ± 0.1) are presented at Fig. 6.

Fig. 6
figure 6

Isotherms of strontium sorption by the T-3K sorbent from low mineralized water (1) and seawater (2). CL is concentration of strontium in the equilibrium liquid phase and CS is concentration of strontium in the equilibrium solid phase

Full size image

The isotherm of strontium sorption by the T-3K sorbent from weakly mineralized water was treated using the mathematical models by Henry, Langmuir, Freundlich and Dubinin–Radushkevich for quantitative description of the results.

The Langmuir isotherm equation is [40]:

$$C_{{\mathrm{s}}} = \frac{{K_{{\mathrm{L}}} \cdot C_{{\mathrm{L}}} \cdot \Gamma_{\infty } }}{{1 + K_{L} \cdot C_{{\mathrm{L}}} }}$$
(6)

where Cs is equilibrium strontium concentration in the sorbent, mg g−1; CL is equilibrium strontium concentration in the solution, g L−1; Г∞ is static exchange capacity (SEC) of the sorbent (sorption capacity after total saturation), mg g−1; KL is Langmuir’s constant, L g−1.

The Freundlich isotherm equation is [41]:

$$C_{{\mathrm{s}}} = K_{{\mathrm{F}}} \cdot C_{{\mathrm{L}}}^{1/n}$$
(7)

where KF is Freundlich’s constant, L kg−1; n is a parameter.

The Dubinin–Radushkevich isotherm equation is [42]

$$C_{{\mathrm{s}}} = \Gamma_{\infty } \exp ( - k \cdot \varepsilon^{2} ),$$
(8)

where k is a constant depending on the energy of adsorption, mol2/kJ2; ε is chemical potential, kJ/mol; R the molar gas constant, 8.31 J/(mol K); T is the temperature, K.

$$\varepsilon = RT\ln (1 + 1/C_{{\mathrm{L}}} )$$
(9)

The isotherms of sorption were plotted in linear coordinates and then treated using the least squares method; quantitative characteristics of the isotherms were determined. The equations were transformed into linear appearances, and the values of the respective constants were calculated (Table 4).

Table 4 Parameters of adsorption isotherms
Full size table

According to the correlation coefficients obtained, the Langmuir’s model was the most suitable model for treating the isotherm of strontium sorption by the T-3K sorbent from weakly mineralized water over the entire concentration range studied. This indicates the ion exchange mechanism of sorption. The Henry’s model was also adequate at low concentrations of strontium. Henry’s distribution coefficient of strontium was 1.3 × 103 mL/g, the static exchange capacity of the sorbent was 30.4 ± 0.5 mg/g.

Considering the dependences of strontium sorption on pH and concentrations of other cations, ion exchange is the most probable mechanism of strontium sorption by the T-3K sorbent. Since the T-3 sorbent did not adsorb strontium, we suggested that the T-3K sorbent adsorbed strontium due to its carbonate phase. Hydrogen of the hydrocarbonate group exchanges by strontium resulting in formation of sparingly soluble strontium carbonate on the surface of zirconium dioxide. The reaction of interaction of the sorbent with strontium can be presented as following (10):

$$2 {\equiv} {\text{Zr-CO}}_{3} {\text{H}} + {\text{Sr}}^{2 + } + 2_{2} \to {}_{{ \equiv {\text{Zr-CO}}_{3}^{/} }}^{{ \equiv {\text{Zr-CO}}_{3\backslash } }} {\text{Sr}} + 2{\text{H}}_{3} {\text{O}}^{ + }$$
(10)

Strontium exchange by sodium also occurs (11):

$$2 {\equiv} {\text{Zr-CO}}_{3} {\text{Na}} + {\text{Sr}}^{2 + } \to {}_{{ \equiv {\text{Zr-CO}}_{3}^{/} }}^{{ \equiv {\text{Zr-CO}}_{3\backslash } }} {\text{Sr}} + 2{\text{Na}}^{ + }$$
(11)

Exchange of sodium by strontium was confirmed by mass spectrometry; a significant increase of sodium concentration was observed in the solution after sorption.

Bezhin et al. [29] suggested another mechanism of strontium sorption by the T-3K sorbent; however, this suggestion seems to be doubtful because the authors stated that the T-3K sorbent is a hydrated zirconium dioxide rather than carbonate-containing zirconium dioxide as we have shown above. We have tested the T-3 sorbent (hydrated zirconium dioxide) and found it did not uptake strontium.

The isotherm of strontium sorption from seawater had a linear appearance at strontium concentrations up to 100 mg/L pointing to the fulfillment of Henry's law. The T-3K sorbent separated strontium at these concentrations due to the ion exchange mechanism. At the same time, the isotherm had different appearance at the concentrations above 100 mg/L. This can be explained by a change of the sorption mechanism from ion exchange to precipitation of strontium carbonate in the pore space of the sorbent. This suggestion is confirmed by low solubility product of SrCO3 (5.3 × 10–10). The value of distribution coefficient of strontium from seawater (Mediterranean Sea) was 85 ± 25 mL/g that agrees well with the respective value of 1.1 × 102 mL/g being obtained in the experiments with simulated seawater [21]. According to the mathematical treatment of the sorption isotherm, the value of static exchange capacity for strontium in seawater was 7 mg/g. However, the sorption isotherm showed that the capacity of the T-3K sorbent for strontium may reach up to 980 mg/g (the maximal value of Cs) at strontium concentration in a solution of more than 100 mg/L probably due to a mechanism of strontium carbonate precipitation in the pore space of the sorbent. This feature of the sorbent allows using it for treating contaminated seawater containing strontium radionuclides or liquid radioactive waste based on seawater.

The study of strontium desorption

Desorption of strontium from the T-3K sorbent by various acidic solutions was studied under batch conditions to define the most suitable type and concentration of stripping solution. Figure 7 shows the dependences of the degree of strontium desorption from the T-3K sorbent by hydrochloric and nitric acid solution on time of the phases contact.

Fig. 7
figure 7

The dependences of strontium desorption from the T-3K sorbent on time of the phases contact

Full size image

Nitric acid appeared unsuitable for strontium desorption since a turbidity of the solution because of destruction of the sorbent was observed even after 1 h of the phases contact. Hydrochloric acid solutions with the concentrations of 2, 3 and 5.5 mol/L were unsuitable too for the same reason. For example, destruction of the sorbent during desorption of strontium by 2 M HCl resulted in an increase of zirconium concentration in the solution that corresponded to leaching of 23% of zirconium from the T-3K sorbent. Thus, 1 M HCl was chosen as the most suitable stripping solution for strontium. The degree of strontium desorption after 3 h was 77%. The longer time of the phases contact is not recommended because a turbidity of the solution was observed after 24 h.

Conclusions

Physicochemical and sorption properties of the T-3K carbonate-containing zirconium hydroxide sorbent were studied. It was shown that the sorbent is a microporous carbonate-containing zirconium hydroxide consisting of ZrO2 (46.39%), Na2O (1.78%), Al2O3 (1.42%) as the major components excluding water.

The T-3K sorbent did not adsorb 137Cs. Affinity of the T-3K sorbent for alkaline earth ions corresponded to the reverse trend of carbonates solubility: Sr > Ba > Ca ≫ Mg. The maximal sorption of strontium achieved at pH 7–10 allowing for successful strontium separation from various aqueous solutions such as seawater and fresh waters, radioactively contaminated industrial waters and liquid radioactive waste. Distribution coefficient of strontium from low mineralized water was Kd = 1.3 × 103 mL/g and the static exchange capacity of the sorbent was 30 mg/g. Sr2+/Ca2+ separation factor depended on calcium concentration in the solution; this value was 9.3 in case of tap water with calcium content of 14.3 mg/L. Separation factors of strontium and other alkaline earth ions were determined for sorption from low mineralized water and seawater. Distribution coefficient of strontium was invariable (~ 6.3 × 103 mL/g) at sodium concentration in the solution ≥ 0.0001 mol/L.