Introduction

During the decontamination and decommissioning of nuclear power plants (NPPs), radioactive 90Sr is of serious concern owing to its high mobility and long half-life of 28.8 years [1]. In addition, 90Sr behaves like Ca in the human body, and it accumulates in the bones and teeth, resulting in associated cancers [2]. 90Sr is known as a common contaminant of structural materials of NPPs [3] and also observed in NPP accidental sites such as the Chernobyl and Fukushima Daiichi sites [4]. Recently, the government of the Republic of Korea announced the termination of the operation of the Kori-1 (pressurized light water reactor) and Wolsung-1 (Canadian deuterium reactor) NPPs, and decommissioning plans are underway including challenging tasks to be demonstrated before being adopted in the field.

In this work, we focus on decontaminating 90Sr-contaminated concrete waste. Concrete is normally composed of an aggregate matrix (60–75%) and a binder (25–40%) which holds the matrix materials together. The aggregate matrix is a mixture of fine materials (such as sand) and coarse ingredients (such as gravel or crushed rocks). SiO2 is known to be the main constituent of these materials, with Al2O3 the second most abundant component. Various oxides such as Na2O, MgO, SO3, K2O, CaO, and Fe2O3 are also present in the aggregate matrix as minor constituents [5]. Portland cement is the most widely used binder, consisting of CaO (~ 63%), SiO2 (~ 22%), Al2O3 (~ 7%), and minor amounts of Fe2O3, MgO, and SO3 [6]. Adding water to the binder induces a hydration reaction, producing cement paste that glues the matrix constituents together.

Currently, decontamination techniques for concrete waste can be categorized into mechanical and chemical techniques [7]. The key idea of mechanical techniques is that the contamination is concentrated at the surface of the concrete waste. Thus, various mechanical surface-exfoliating techniques are employed for materials in this category [8]. The chemical technique employs various mineral and organic acids, surfactants, gels, and foam chemicals for the dissolution of radioactive waste with or without non-radioactive concrete itself. These wet chemical techniques can achieve higher decontamination factors than mechanical techniques, though the generation of secondary waste remains as a critical issue. A brand new approach that employs chlorine gas as a selective reaction medium for the decontamination of 90Sr-contaminated concrete waste is introduced in this study. Some metal oxides are reactive with chlorine gas without the support of a reducing agent such as carbon. Normally, metal chlorides have considerable solubility in water and higher volatility than their oxide forms. The key idea of this chlorination decontamination technique is the selective conversion of radioactive contaminants into their chloride forms so that they can be separated via water washing or a heat treatment (i.e., distillation). Among the various constituents of concrete waste, a couple of experimental results related to chlorination reactions of CaO [9] and CaCO3 [10] have been found. Yake and Ulrichson investigated the chlorination reaction behavior of CaO at 333–472 °C, and the formation of CaCl2 was confirmed [9]. The reaction between CaCO3 and Cl2 was demonstrated by Orosco et al. [10] using a thermogravimetric analysis technique, indicating that the chlorination reaction occurred at 420–900 °C. However, there are no experimental data available for SrO. In the present study, theoretical and experimental approaches were conducted in order to investigate the feasibility of the chlorination decontamination technique for application to the decontamination of 90Sr-contaminated concrete waste.

Experimental

Theoretical calculations were conducted using the “Reaction Equations” and “Equilibrium Compositions” modules of HSC chemistry 9.5.1 [11]. The Reaction Equations module was utilized to calculate the Gibbs free energy change (ΔG) and equilibrium constant (K) values for various reaction equations of the representative species in concrete waste in the temperature range of 0–1000 °C. The Equilibrium Compositions module was utilized for a quantitative analysis of the chlorination reaction behavior of SrO in the presence of CaO, CaCO3, or MgO. These calculations were performed at various reaction temperatures (200–1000 °C) and chlorine input. The initial amount of chlorine input was set to 1/10 of SrO and was then increased to 100 times the initial amount (= 10 times the initial amount of SrO). Idential calculations were also repeated for SrCO3 instead of SrO. The input data for the calculations are listed in Table 1. During the calculations, the pressure of the system was set to 1.0 bar.

Table 1 Input data for calculations of equilibrium composition using the HSC chemistry software
Full size table

Chlorination experiments were conducted using a horizontal quartz tube reactor with a diameter of 4 cm. On the inlet side, the reactor was connected to a gas feed system which separately controls the flow rates of argon and chlorine gases using mass flow controllers (MFCs, Kofloc co., Japan, Model 3660 for Ar and 5440 for Cl2). In the middle of the reactor, a heater was positioned to control the reaction temperature. Schematic diagram of the reaction system is shown in Fig. 1. As it is known that SrO reacts with CO2 in the air to form SrCO3, SrO powder (Sigma-Aldrich, 99.9%) was heat treated at 800 °C for 4 h under a 200 mL/min O2 flow before chlorination experiments. Approximately 1.20 g of heat treated SrO was employed for each chlorination experiment. The samples were loaded in an alumina boat and then placed in the middle of the quartz reactor. After purging the reactor with an argon gas flow, the heater began to heat the reactor to the target temperature at a ramping rate of 10 °C/min. After the heater temperature was stabilized at the reaction temperature, the gas flow rate was changed from 150 mL/min Ar to 96 mL/min Ar + 4 mL/min Cl2 and maintained during the reaction. The effects of the reaction time and temperature were investigated for SrO by repeating the abovementioned process. The chlorination experiments were also conducted using CaO (Sigma-Aldrich, 99.9%, powder), CaCO3 (Sigma-Aldrich, > 99.0%, powder), and SrCO3 (Sigma-Aldrich, > 98%, powder) as a function of the reaction temperature in order to compare the corresponding chlorination reaction characteristics with those of SrO.

Fig. 1
figure 1

Schematic diagram of the experimental set-up

Full size image

Phase changes of the reacted samples were analyzed using an X-ray diffraction system (XRD, Bruker D8 Advance). The XRD signals were measured in the 2θ range of 10–80° with a step size of 0.0104° and measurement time of 0.15 s during each step.

Results and discussion

Calculating the ΔG value of a reaction is the first step when estimating the feasibility of a reaction under a certain condition. Figure 2a shows the ΔG calculation results for chlorination reactions of the various oxides that make up the major constituents of the aggregate matrix. A positive value of ΔG indicates that the reaction products have higher chemical potential than the reactants and that the reaction will not likely proceed toward the right-hand direction. On the other hand, a negative ΔG value indicates that the reaction is thermodynamically feasible to generate reaction products. Among the reactions considered in this work, it is clear that Al2O3, SiO2, and Fe2O3 are not reactive with chlorine up to 1000 °C. MgO exhibited mixed values, mostly negative except in the temperature range of 465–714 °C. Figure 2b displays the ΔG calculation results for chlorination reactions of various oxides and carbonates of Ca and Sr. Calcium compounds, CaO, CaCO3, and Ca(OH)2, showed negative ΔG values in the entire temperature range matching those of Sr compounds. It is important to note that various chemical forms of Ca were considered here, as CaO, the initial form as an ingredient of concrete, reacts with water to produce Ca(OH)2 and then is slowly converted into CaCO3 by reacting with CO2 in air. The calculations for Ca(OH)2 were conducted only up to 400 °C as it decomposes at 396–481 °C [12]. According to the ΔG values, the thermodynamic preference for the chlorination reaction was CaO > (Ca(OH)2) > CaCO3 in the entire temperature range. For Sr, SrO is known as the major chemical form in used oxide fuels [13]. However, it can react with water or CO2 to produce Sr(OH)2 or SrCO3, respectively, after being exposed to air. The calculations for Sr(OH)2 were limited to 500 °C as it decomposes in the tempearture range of 475–575 °C [12], and it was found that Sr(OH)2 exhibits higher ΔG values than those of SrO. The calculations for SrCO3 and CaCO3 were limited to 700 °C for the same reason described above [14, 15]. Among the three Sr compounds, SrO exhibited the lowest ΔG value in the entire temperature range. These calculation results suggest that, along with Sr compounds, Ca compounds and MgO may also be converted to CaCl2 and MgCl2, respectively, during the chlorination reaction.

Fig. 2
figure 2

Gibbs free energy change (ΔG) values of a Al2O3, SiO2, Fe2O3, and MgO and b CaCO3, CaO, Ca(OH)2, SrO, Sr(OH)2, and SrCO3 for the chlorination reaction in the temperature range of 0–1000 °C calculated using the HSC chemistry code

Full size image

A quantitative analysis was conducted using the Equilibrium Compositions module to predict the chlorination reaction behaviors of SrO and SrCO3 in the presence of CaO, CaCO3, or MgO, which were identified as being reactive with chlorine in the ΔG calculations. The quantitative analysis results are shown in Fig. 3 as functions of the reaction temperature (200–1000 °C) and the input amount of Cl2. As an indicator of conversion selectivity for SrO and SrCO3, the ratio of SrCl2 to CaCl2 or MgCl2 in an equilibrium state is employed. Figure 3a shows the effects of the reaction temperature and the input amount of Cl2 on the SrO–CaO system with an initial SrO concentration of 1 ppm versus CaO. It is clear in the figure that an increase in the reaction temperature results in a decrease in the SrCl2/CaCl2 ratio due to the chlorination reaction of CaO. Figure 3b shows the ratio of SrCl2 over the initial amount of SrO, in which it is clear that an increase in the reaction temperature leads to a decrease in the SrCl2 ratio. This result indicates that the CaO chlorination reaction is preferred over the SrO chlorination reaction at temperatures greater than 400 °C. In the SrO–CaCO3 system shown in Fig. 3c, d, especially high SrCl2/CaCl2 values were observed at 200 and 400 °C when the Cl2 molar ratio was smaller than 1. In addition, the SrCl2/CaCl2 ratio increased from 0.0184 in the SrO–CaO case to 0.0853 in the SrO–CaCO3 case at 600 °C and at Cl2 molar ratio of 10. These results indicate that the selective conversion of SrO is more convenient in the SrO–CaCO3 system compared to the SrO–CaO system because the chlorination reaction of CaCO3 occurs at higher temperatures than that of CaO. The calculations for the SrO–CaCO3 were conducted up to 700 °C due to the thermal decomposition of CaCO3 into CaO and CO2 above 700 °C [15]. With regard to MgO, as shown in Fig. 3e, f, the SrCl2/MgCl2 ratios exceeded 0.111 regardless of the reaction temperature and the Cl2 molar ratio. Figure 3f clearly shows that the conversion of SrO into SrCl2 occurs even at low Cl2 molar ratio and high temperatures, indicating high selectivity for Sr in the presence of MgO. In summary, it is clear from the calculation results that reacting at a low temperature is beneficial for achieving high selectivity for Sr when it exists in the SrO form in concrete waste.

Fig. 3
figure 3

Equilibrium compositions (a, c, e) and conversion ratios of SrCl2 (b, d, f) for the SrO–CaO (a, b), SrO–CaCO3 (c, d), and SrO–MgO (e, f) systems at various temperatures and Cl2 molar ratios with an initial SrO concentration of 1 ppm versus CaO, CaCO3, or MgO

Full size image

Identical calculations were repeated for SrCO3 with CaO, CaCO3, and MgO, and these results are shown in Fig. 4. Interestingly, an increase in the reaction temperature resulted in an increase in the SrCl2/CaCl2 ratio in the SrCO3–CaO system (Fig. 4a). It is clearly shown in Fig. 4b that the SrCl2 molar ratio increases with an increase in the reaction temperature. These results indicate that thermodynamic preference for the chlorination reaction moves from CaO to SrCO3 with an increase in the reaction temperature. The opposite behavior was observed with the SrCO3-CaCO3 system shown in Fig. 4c, d. The SrCl2/CaCl2 ratio rapidly dropped with an increase in the reaction temperature, especially in the low temperature region. It is important to note that the effect of the Cl2 molar ratio was negligible in Fig. 4c, whereas the SrCl2 molar ratio increased linearly with an increase in the Cl2 molar ratio (Fig. 4d). This result means that an increase in the Cl2 molar ratio produces both SrCl2 and CaCl2 at an identical ratio, which is dependent only on the reaction temperature. High SrCl2/MgCl2 values were identified in the SrCO3–MgO system (Fig. 4e) regardless of the reaction temperature. The effect of the reaction temperature on the SrCl2 molar ratio was not distinguishable, as shown in Fig. 4f, indicating that SrCO3 may react with chlorine prior to MgO under any of the conditions considered in this work. The calculation results with SrCO3 suggest that operation at a high temperature is beneficial, except for the SrCO3–CaCO3 system, in order to ensure a high SrCl2/CaCl2 ratio. Recalling that a high SrCl2/CaCl2 ratio was observed at low temperatures with SrO, it can be summarized that the optimum reaction temperature may vary according to the chemical forms of Sr and Ca.

Fig. 4
figure 4

Equilibrium compositions (a, c, e) and conversion ratios of SrCl2 (b, d, f) of the SrCO3–CaO (a, b), SrCO3–CaCO3 (c, d), and SrCO3–MgO (e, f) systems at various temperatures and Cl2 molar ratios with an initial SrCO3 concentration of 1 ppm versus CaO, CaCO3, or MgO

Full size image

The effects of the reaction temperature and duration on the chlorination reaction of SrO were investigated and the results are shown in Fig. 5. The conversion ratio on the y-axis indicates the ratio of SrO converted to SrCl2 (and Sr4OCl6) relative to the initial amount of SrO, which was calculated using the equation below,

$${\text{Conversion}}\;{\text{ratio}} = \left( {w_{{\text{f}}} {-}w_{0} } \right)/\left( {w_{0} \times \left( {{\text{MW}}_{{{\text{SrO}}}} /{\text{MW}}_{{{\text{SrCl2}}}} } \right)} \right)$$
(1)

where wf is the weight measured after reaction, w0 is the initial weight, MWSrO is the molecular weight of SrO (= 103.62 g/mol), and MWSrCl2 is the molecular weight of SrCl2 (= 158.53 g/mol). It is clear in the figure that the conversion ratio increases with an increase in the reaction time. It is interesting to observe that the maximum conversion ratio values were less than 0.4 in the reaction temperature range of 300–600 °C. This result may stem from the limited diffusion of chlorine within the solid phase at this temperature range. The conversion ratios at this temperature range require additional investigations. After 4 h of reaction, the conversion ratio increased from 0.155 to 0.320 when the temperature increased from 300 to 400 °C. When the temperature was increased further to 500 °C, the conversion ratio dropped profoundly to 0.0966 and then returned to 0.350 at 600 °C. This trend was confirmed by repeated experiments to ensure that it was not a result of any experimental errors. The reason for the poor conversion ratio at 500 °C is not clear at this point; however, it is obvious that the chlorination reaction should be conducted at or above 700 °C to achieve a high conversion ratio with SrO. These results eliminate the feasibility of a low-temperature operation for the high SrCl2/CaCl2 ratio proposed in the thermodynamic calculations with SrO (Fig. 3). Based on the outcomes of this work and previous documents [9, 10], it is obvious that the conversion of CaO and CaCO3 into CaCl2 during the chlorination reaction of SrO is not avoidable. Here it should be noted that the highest reaction temperature was limited to 800 °C, below melting point of SrCl2, 874 °C, in order to avoid volatilization of the reaction product. As the conversion ratio is derived from weight change, volatilization of SrCl2 results in reduced conversion ratio. After repeated experiments, the quartz reactor was clean and high conversion ratio of 0.914 was observed after reaction at 800 °C for 3 h.

Fig. 5
figure 5

Conversion ratio of SrO as a function of both the reaction temperature and the time

Full size image

Structural changes upon the chlorination reactions were investigated. Figure 6a shows the XRD measurement results of the samples reacted at 300 °C. The formation of the SrCl2 phase (JCPDS #01-072-1537) was confirmed from the XRD peaks without any other phases. It was identical in the samples reacted at 400 °C, as shown in Fig. 6b. It is clear in the figures that the relative intensity of the SrCl2 peaks increases with an increase in the conversion ratio. When the reaction temperature increased to 500 °C, along with SrCl2, a new phase of Sr4OCl6 (JCPDS #01-086-1832) was identified regardless of the reaction time, as shown in Fig. 6c [16, 17]. These results suggest that the sudden decrease in the conversion ratio at 500 °C is related to the formation of Sr4OCl6, though additional experiments are necessary to explain the details of the relationship. A thermodynamic analysis of the formation of Sr4OCl6 was not available due to lack of data in the HSC chemistry code. The trend was similar in the 600 and 700 °C cases, which exhibited peaks of SrO (JCPDS #01-075-6979), SrCl2, and Sr4OCl6, as shown in Fig. 6d, e. It was interesting to observe the outcome at 800 °C (Fig. 6f). After 0.5 and 1 h of the reaction, tiny peaks of Sr4OCl6 were observed with high intensities of SrCl2 and SrO peaks. When the reaction time was increased to 2 h, the peaks of SrCl2 are dominant while those of SrO are relatively small. At the same time, the peaks of Sr4OCl6 disappeared, suggesting that this phase was converted into SrCl2 at 800 °C by the chlorination reaction. At this temperature, thermal decomposition of Sr4OCl6 into SrO and 3SrCl2 might be negligible due to its high thermal stability [17]. After 3 h of the reaction, only SrCl2 peaks were observed, suggesting that the amount of SrO is below the detection limit of the XRD measurement technique. These XRD measurement results show that the reaction product generated via the chlorination reaction changes with the reaction temperature and time only at 800 °C. It should be noted here that the presence of Sr4OCl6 does not affect the conversion ratio results shown in Fig. 5, because its weight is identical to SrO–3SrCl2. Therefore, the conversion ratio still represents the amount of SrO reacted with Cl2 when the reaction products are SrCl2 and Sr4OCl6. However, it is possible that other chlorides or oyxchlorides have been produced during the experiments, which are not detectable by the XRD measurements.

Fig. 6
figure 6

XRD measurement results as a function of both the reaction temperature, at a 300, b 400, c 500, d 600, e 700, and f 800 °C, and the reaction time at each temperature

Full size image

The thermodynamic calculation results suggested that the Ca compounds react with chlorine gas to produce CaCl2 during the chlorination of SrO. A set of experiments was conducted in order to identify the chlorination reaction behavior of CaO and CaCO3. Figure 7 shows the effect of the reaction temperature on the conversion ratio after 4 h of the reaction under a 96 mL/min Ar + 4 mL/min Cl2 flow. The chlorination reaction behavior of SrCO3 was also investigated under identical conditions. Here, Ca(OH)2 was not included due to its decomposition temperature of 512 °C and experimental confirmation of partial decomposition at 300 and 400 °C. It is clearly shown in the figure that CaO has higher conversion ratios than those of SrO, except at 800 °C. The conversion ratio of CaO peaked at 600 and 700 °C and then decreased slightly at 800 °C. This result may have come from evaporation of CaCl2 (melting point 775 °C) or the temperature dependence of the chlorination reaction. In contrast, the chlorination reaction of CaCO3 began at 700 °C with a conversion value of 0.361, which is only 38.7% of the conversion ratio of CaO at 700 °C. The chlorination reaction began at 300 °C in SrCO3, and the conversion ratio dramatically increased to 0.424 and 0.806 at 400 and 500 °C, respectivley. However, the conversion ratio decreased with a further increased in the reaction temperature to 600 and 700 °C. It is interesting to note that SrO exhibited an especially low conversion ratio value at 500 °C, where the highest conversion value was observed with SrCO3. These results suggest that the high selectivity for SrO estimated in the thermodynamic calculations is not readily achievable in the presence of CaO. In addition, it is recommended to operate the chlorination decontamination process at or above 700 °C.

Fig. 7
figure 7

Conversion ratio values of SrO, SrCO3 CaO, and CaCO3 as a function of the reaction temperature after 4 h of the reaction with a 96 mL/min Ar + 4 mL/min Cl2 flow

Full size image

Conclusions

A chlorination technique was introduced for use during the decontamination of 90Sr-contaminated concrete waste. Thermodynamic investigations suggested that the chlorination technique is a promising approach to convert SrO into SrCl2 with relatively low conversion ratios in Ca compounds and MgO. However, the experimental results revealed that the reaction temperature should be at least 700 °C to achieve a significant amount of SrCl2 and that CaO exhibited higher or similar conversion ratios compared to those of SrO. In conclusion, the proposed chlorination decontamination technique has the merits of a simple process stream for Sr removal, while the simultaneous removal of non-radioactive Ca may lessen its decontamination factor. In addition, further investigations including interactions between chlorides and volatilization of chlorides might be necessary before applying the chlorination technique to the decontamination of concrete waste.