Introduction

The radioactive nuclide-contaminated water generated at Tokyo Electric Power Company (TEPCO)’s Fukushima Daiichi Nuclear Power Plant (FDNPP) is stored for long periods in on-site storage tanks after Advanced Liquid Processing System (ALPS) treatment. Although ALPS cannot remove tritium in water, it can remove nearly all other fission-derived radioactive nuclides [1]. Owing to limitations in tank capacity, the Japanese government decided to discharge the treated tritium-containing water into the ocean [2]. Ocean discharge began on August 24, 2023. The treated water was diluted to less than 1/40th of the discharge criteria for tritium-contaminated wastewater. However, as of April 2024, the volume of water yet to be discharged is approximately 1.33 million m3 or 97% of the FDNPP’s total tank capacity [3]. According to TEPCO’s implementation plan, the completion of ocean discharge is expected to take 29 years or until 2051 [4]. During this time, the tanks may age, break, or leak owing to major earthquakes, leading to increased contamination. Concerns have also been raised about the coexistence of long-living fission-derived radionuclides, such as 135Cs (half-life, 1.33 × 106 years) or 129I (half-life, 1.57 × 107 years), in the treated water because these materials are difficult to measure. In addition, the ocean discharge of treated tritium-containing water may exert a negative impact on the fishing industry.

We conducted this study to contribute to the safe storage of treated tritium-containing water. Superabsorbent polymers (SAPs) are widely used in various products [5, 6]; these materials can hold water at up to 100–1,000 times their weight, depending on the material. The water-absorption performance of SAPs decreases with the inclusion of electrolytes in the water, but remains virtually unchanged when non-electrolytes are included. Given these characteristics, SAPs have recently been utilized in agriculture to adjust the moisture content of soil [7]. Research has been conducted to investigate the safe storage of waste liquids generated from radioiodine therapy, which involves the use of radioactive I [8]. In a previous study, we showed that a solid SAP can safely store treated water in the tanks of the FDNPP with less risk of leakage than its liquid form; this material does not require energy and can evaporate naturally under a controlled tritiated water concentration [9]. We also reported [9] that an SAP can retain Cs+ in its structure; once retained, Cs+ is relatively stable in the SAP, even if water is added to it, as long as the amount of water added does not exceed its absorption capacity. These findings demonstrate the saturated adsorption capacity of SAPs for Cs+.

The capture and retention of I from water is not a simple task because inorganic I naturally takes various chemical forms, such as I−, I3−, or IO3−. The capture of I from radioactive waste, especially that generated from nuclear medicine treatment, has been actively studied [10]. Activated carbon is commonly used for this purpose; however, its high adsorption capacity for various compounds and the presence of impurities can lead to a rapid decrease in its I retention capacity. Therefore, various adsorbents have been developed as alternatives to this material [11]. For instance, we developed a hydrogel capable of retaining anions within its structure [12]. Song et al. [12] applied a hydrogel composed of N-[3-(dimethylamino)propyl]acrylamide (DMAPAA) and N,N-dimethylamino propyl acrylamide, methyl chloride quaternary (DMAPAA-Q) to simultaneously remove cationic heavy-metal ions via OH− formation and anionic ions through ion exchange; these authors used the gel to recover hexavalent Cr as an insoluble hydroxide from wastewater. Note that alkali metals, such as Cs and alkaline earth metals, have high solubility in hydroxides; therefore, they are not immobilized within the hydrogel.

We suggest that the combination of an SAP and a hydrogel can enable the safe storage of treated water at the FDNPP while retaining long-living fission-derived radionuclides such as Cs+ and I; atmospheric release without heating may also be possible. However, some areas within the FDNPP site in which the tanks are located have a high ambient radiation dose rate because of the residual presence of 662 keV gamma rays emitted from 137Cs, which is a fission-derived radionuclide. Therefore, understanding the radiation resistance of the SAP and hydrogel is necessary for their practical implementation. To the best of our knowledge, there are no studies of radiation resistance studies on the water absorption capacity of an SAP; a few studies on radiation tolerance of Cs+ and I adsorbents [13,14,15,16] have been reported; however, no studies on the effect of radiation resistance on the water absorption capacity of an SAP have yet been published. Therefore, in this study, we externally irradiated the SAP and hydrogel with 662 keV gamma rays and investigated the changes in the water absorption capacity of the SAP and the saturated Cs+ and I adsorption capacities of the hydrogel under different absorbed radiation doses.

Experimental

SAP

The SAP (BC-283FHA; TAISAP, Taipei, Taiwan) used in this study was prepared from polyacrylate polymers. The SAP had a granular form with a particle size of 106–850 µm. The pure-water absorption capacity of the SAP was 400 g/g.

Hydrogel

The hydrogel (MG-1; Gel Techno Research Co., Ltd., Higashi-Hiroshima, Japan) used in this study was a copolymer of DMAPAA (KJ Chemicals Co., Tokyo, Japan) and DMAPAA-Q (KJ Chemicals Co.).

Gamma-ray irradiation

The SAP or hydrogel (5 g) was placed in a 500 mL polyethylene container (diameter, 7.8 cm; height, 15 cm) and added with 500 mL of ultrapure water. Because the weight of the water was 100 times that of the SAP, the latter existed in gel form owing to its lower water absorption capacity than that of typical SAP. Additionally, because the water absorption capacity of the hydrogel was extremely low, it existed as granules dispersed in the water. The containers were fixed inside a 137Cs irradiation apparatus (PS-3200 T; Pony Industries Co., Ltd., Osaka, Japan) at the Tokyo Metropolitan Industrial Technology Research Institute. As shown in Fig. 1, the polyethylene containers were oriented horizontally to achieve a central absorbed dose rate of 200 Gy/h. Seven samples of both the SAP and hydrogel were prepared for irradiation for 0.1, 0.2, 0.3, 0.4, 0.5, 5.0, and 50 h while the irradiation platform was horizontally rotated; thus, the cumulative absorbed dose these samples was 20 Gy, 40 Gy, 60 Gy, 80 Gy, 100 Gy, 1 kGy, and 10 kGy, respectively. After irradiation, each sample was dried by heating at 60 °C for 6 d, followed by heating at 110 °C for 1 h. Photographs and videos of the samples after irradiation are shown in the Supplemental Information.

Fig. 1
figure 1

Positional relationship of external radiation exposure from 137Cs to the samples

Full size image

Water absorption capacity of the irradiated SAP

All irradiated SAP samples (0.2 g each, n = 3) were placed in nonwoven fabric bags (6.5 cm × 9.5 cm; maximum bottom width, 3 cm) and then inserted into separate polyethylene containers with a volume of 250 mL. Subsequently, 200 mL of ultrapure water was added to each container, which was then sealed and left undisturbed at room temperature (20℃) for 7 d. Next, all the liquid outside of the bags was removed, and the masses of the irradiated SAP were measured to determine their absorption capacity per gram of sample.

A practical approach to investigate the water absorption capacity of the irradiated SAP is to observe the change in retention after irradiation under water-absorption conditions. However, the radiation resistance of the nonwoven fabric bags that were placed in the SAP for solid–liquid separation was unknown, and the bags could be degraded at high radiation doses. Therefore, we could not employ this method in this study.

Saturated Cs+ adsorption capacity of the irradiated SAP

Based on the results of the water absorption capacity experiments, the saturated Cs+ adsorption capacity experiments were conducted using only the SAP sample with an absorbed dose of 40 Gy, which did not show significant changes in water absorption capacity within the margin of error compared with the control group (0 Gy), and that with an absorbed dose of 60 Gy, which showed a significant decrease in water absorption capacity. The SAP samples (0.2 g each; n = 3) were placed in nonwoven fabric bags of the same size described above. Nonradioactive CsI (Kojundo Chemical Laboratory Co., Ltd., Saitama, Japan) solutions with concentrations of 2.5, 5.0, 10, 25, 50, and 100 mmol/L were prepared in 250 mL polyethylene bottles. In addition, 80 µL of 137Cs radioactive standard solution (0.1 MBq/mL, CS010; Japan Radioisotope Association, Tokyo, Japan) with a CsCl concentration of 0.05 mg/g was added into the solutions as a radioactive tracer to achieve a final activity of 8 kBq per bottle. The specific radioactivities of Cs in the 2.5, 5.0, 10, 25, 50, and 100 mmol/L CsI solutions were 16, 8.0, 4.0, 1.6, 0.80, and 0.40 MBq/mol, respectively. The bottles were sealed with caps, and the solutions were left undisturbed at room temperature (20℃) for 7 d to reach equilibrium. Subsequently, the liquid outside of the bags was removed, and the moisture content of the samples was determined by measuring their mass. A 2 mL sample of the removed liquid was then subjected to 137Cs radioactivity concentration [Bq/kg] quantification using a NaI(Tl) scintillation detector (Cobra 5003; Packard, CT, USA) for 600 s. The saturated Cs+ adsorption capacity of the samples was determined by analyzing the results using the Langmuir adsorption isotherm model.

Saturated I adsorption capacity of the irradiated hydrogel

All irradiated and nonirradiated (control; 0 Gy) hydrogel samples (0.2 g each; n = 3) were placed in nonwoven fabric bags of the same size as described above. Nonradioactive CsI (FUJIFILM Wako Pure Chemical Co., Tokyo, Japan) solutions with concentrations of 2.5, 5.0, 10, 25, 50, and 100 mmol/L were prepared in 250 mL polyethylene bottles. Additionally, 80 µL of 125I radioactive standard solution (0.1 MBq/mL, ID010; Japan Radioisotope Association) containing 0.02 mg/g Na2S2O3, 0.03 mg/g NaOH, and 0.05 mg/g Nal was added into the solutions as a radioactive tracer to achieve a final activity of 8 kBq per bottle. The specific radioactivity of a solution containing only the 125I radioactive standard solution added to ultrapure water was identical to that of the carrier (8.1 TBq/mol).

The bottles were sealed with caps, and the solutions were left undisturbed at room temperature (20℃) for 7 d to reach equilibrium. Subsequently, the liquid outside of the bags was removed, and the moisture content of the samples was determined by measuring their mass. A 10 mL sample of the removed liquid was also subjected to 125I radioactivity concentration [Bq/kg] quantification using a high-purity Ge detector (GX4018; Mirion Technologies (Canberra) Inc.). Further details of the high-purity Ge detector system have been described in our previous studies [17,18,19]. The counting time for each sample was set to range from 600 to 3,600 s, depending on its radioactivity, to ensure that the measurement error was below 3%. The saturated I adsorption capacity of the hydrogels was finally determined by analyzing the results using the Langmuir adsorption isotherm model.

Results

Water absorption capacity of the irradiated SAP

The water absorption capacity of the nonwoven fabric bag was experimentally determined to be 3.82 g/g. Table 1 shows the variation in water absorption capacity per gram of SAP sample with various absorbed doses. No significant decrease in water absorption capacity within the margin of error was observed between the control sample (absorbed dose, 0 Gy) and those with absorbed doses of 20 and 40 Gy. The water absorption capacity of the irradiated SAP significantly decreased at absorbed doses exceeding 60 Gy.

Table 1 Water absorption capacity of the irradiated SAPs
Full size table

Saturated Cs+ adsorption capacity of the irradiated SAP

Figure 2 shows the Langmuir plots of the irradiated SAP samples with absorption doses of 40 and 60 Gy. The plot for the 40 Gy sample closely approximated a straight line, with a high correlation (r2 = 0.940). The saturated Cs+ adsorption capacity of the sample was calculated from the reciprocal of the slope of this line to be 0.511 ± 0.007 mg/g. By contrast, the plot for the 60 Gy sample showed only slightly resembled a line, with an intermediate correlation (r2 = 0.617). Moreover, owing to the influence of outliers, such as the negative C/q values at 25 mmol/L, the line was difficult to approximate as straight even when the C/q values at 25 mmol/L were zero. Consequently, an accurate saturated Cs+ adsorption capacity for the 60 Gy sample could not be determined. Attempts were made to fit the plots to the Freundlich adsorption isotherm; however, the results of both the 40 and 60 Gy samples showed low linearity.

Fig. 2
figure 2

Langmuir adsorption isotherms of irradiated SAPs with absorbed doses of a 40 and b 60 Gy for saturated Cs+ absorption capacity determination

Full size image

Saturated I adsorption capacity of the irradiated hydrogel

The water absorption capacity of the hydrogel was experimentally determined to be approximately 5 g/g, regardless of the absorbed dose. Figure 3 shows the Langmuir plots of the irradiated hydrogel samples. The plots were approximately linear within the concentration range in which the correlation coefficient was maximized. For samples with absorbed doses of 100 Gy and 1 kGy, the correlation coefficient was maximized with a linear approximation at concentrations of < 25 mmol/L. For samples with other absorbed doses, the correlation coefficient was maximized with a linear approximation across all concentration ranges tested. The results of the saturated I adsorption capacity experiments are summarized in Table 2. No significant difference was observed between the samples with an absorbed dose of up to 100 Gy and the control sample. However, when the absorbed dose of the sample exceeded 1 kGy, its saturated I adsorption capacity decreased. Attempts were also made to fit the plots to the Freundlich adsorption isotherm; however, the results of all the samples showed low linearity.

Fig. 3
figure 3

Langmuir adsorption isotherms of irradiated hydrogels with absorbed doses of a 0 Gy, b 20 Gy, c 40 Gy, d 60 Gy, e 80 Gy, f 100 Gy, g 1 kGy, and h 10 kGy for saturated I absorption capacity determination

Full size image
Table 2 Saturation I adsorption capacities of the irradiated hydrogels
Full size table

Discussion

In this study, we investigated the external radiation resistance of an SAP and a hydrogel to assess their applicability to the safe storage of treated water at the FDNPP. No significant difference within the margin of error was observed between the water absorption capacities of irradiated SAP samples with absorbed doses of up to 40 Gy and that of the nonirradiated SAP. The average dose rate of natural radiation in Japan is 30.1–59.4 nGy/h [20, 21]. The measured dose rate at the location of the storage tank at the FDNPP was reported to be 2.4 µGy/h as of April 2021 [22]. The internal dose rate of SAP in the storage tanks was difficult to estimate. However, the dose rate outside the tanks can be estimated to be higher than that inside the tanks for several reasons. First, at the FDNPP, the dose rate outside the tanks is higher than the average dose rate of natural radiation in Japan because of the presence of 137Cs, which have a long half-life of approximately 30 years and emit strong β-rays. In addition, 137Cs emits γ-rays of 662 keV. On the other hand, the tanks are made of stainless steel, which provides shielding against β- and γ-rays. The treated water also exerts a slight shielding effect. A potential source of radiation inside the tank is tritium, which only emits β-rays with a maximum energy of 18.6 keV. Given the very low energy of the β-rays of tritium, the main factor influencing the dose rate inside the storage tank is radiation from external 137Cs. Even if the internal dose rate in the storage tanks is 100 µGy/h, approximately 400,000 h or 45.6 years could pass before an absorbed dose of 40 Gy is reached. Therefore, the SAP is practical for use during the ocean discharge of treated tritium-contaminated water.

Based on the reciprocal of the slope of the Langmuir plots, the saturated Cs+ adsorption capacity of the 40 Gy SAP sample was 0.511 mg/g; however, the corresponding value for the 60 Gy sample could not be determined. A saturated Cs+ adsorption capacity of 2.40 mg/g was previously reported for the nonirradiated sample (0 Gy) [9]. Therefore, the decrease in the saturated Cs+ adsorption capacity of the SAP is greater than that of its water absorption capacity.

We previously discussed two reasons [9] behind the selective retention of Cs+ by an SAP. First, the SAP exists in the form of a Na carboxylate salt when dried; after hydration, Na+ dissociates from the material. Both Na+ and Cs+ exist as hydrated ions in aqueous solutions, and the hydrated ion radius of Cs+ (0.329 nm) is smaller than that of Na+ (0.358 nm). Consequently, Cs+ can more easily penetrate the polymer mesh structure than Na and is thus more readily retained by the SAP. Second, the structure of the SAP, which has carboxyl groups at its end, is similar to that of weakly acidic cation-exchange resins. In cation-exchange resins, when ions of the same charge are considered, the smaller the hydrated ion radius, the higher its adsorption rate. Therefore, the exchange of Na+ with Cs+ occurs at the carboxyl–Na group, leading to its retention in the SAP.

The decrease in water absorption capacity of the SAP with increasing absorbed dose is due to radiation chemical reactions. Because γ-rays have a low linear energy transfer, they exhibit a higher proportion of indirect effects mediated by water than direct effects. When radiation strikes water, the water molecules undergo ionization and excitation reactions, generating reactive species, such as OH− radicals, hydrogen radicals, and hydrated electrons. These products can recombine to produce not only water but also highly oxidizing H2O2. These substances are believed to reduce the number of water molecules that can be retained by SAP by cleaving the structural bonds, such as C–C bonds, of the material. New groups and additional crosslinks may also form [13].

Orechovská and Rajec [14] developed composite sorbents for Cs+ based on potassium nickel ferrocyanide embedded in a silica gel matrix. The saturated Cs+ adsorption capacity of this material was 0.50 mmol/g (6.7 mg/g) at an absorbed dose of 10 kGy, and its radiation resistance was maintained at absorbed doses of up to 10 kGy. Wang et al. [15] developed a three-dimensional uranyl organic framework that could selectively remove Cs+ from aqueous solutions; even after irradiation with 200 kGy 60Co γ-rays at a dose rate of 1.2 kGy/h, no degradation of the material’s structure or crystallinity occurred. Markova et al. [13] studied sorbents for Cs+ removal from high-level alkaline waste; these materials demonstrated a 24%–50% reduction in adsorption capacity at absorbed doses of up to 2 MGy. However, in contrast to our study, the irradiation conditions adopted in these previous studies neither indicated the presence of water nor demonstrated the impact of the radiation-induced chemical reactions of water.

The control (0 Gy) and 100 Gy hydrogel samples showed no significant difference in saturated I adsorption capacity. However, decreases of approximately 20% and 30% were observed in samples with absorbed doses of 1 and 10 kGy, respectively. This decrease in saturated adsorption capacity could be attributed to the decomposition of the gel structure owing to radiolytic chemical reactions. SAP has a high water absorption capacity, and its structure can accommodate more water molecules than that of the hydrogel. Therefore, structural degradation begins at absorbed doses exceeding 40–60 Gy, resulting in a decrease in water retention and Cs+ absorption capacity. However, the hydrogel exhibits high resistance to radiolytic chemical reactions because of its low structural water content.

The mechanism of I adsorption by the hydrogel involves ion exchange with OH− or Cl− ions on –NH(CH3)2+ or –N(CH)3+ groups [12]. The amount of I that can be retained by the gel decreases as the breakage of bonds, such as C–C and C–N bonds, in its skeletal structure increases owing to radiolytic chemical reactions. At present, we cannot determine where the breakage in the hydrogel occurs. The use of techniques such as nuclear magnetic resonance spectroscopy to search for liberated functional groups may be employed in future studies to investigate where such breakages occur.

Cyclodextrin has been utilized for the selective capture and stable retention of I. In recent years, its application has been extended to the inhibition of 131I uptake by the thyroid in the living body [23]. Ihara et al. [24] demonstrated that the saturated I adsorption capacity of β-cyclodextrin polymer beads was 1,410 mg/g, surpassing that of the hydrogel used in our study. Among the cyclodextrins that can retain I through encapsulation, 2-hydroxypropyl cyclodextrin with absorption doses of 100 Gy, 1 kGy, 5 kGy, 10 kGy, and 30 kGy in aqueous solution was examined to assess its resistance to external 137Cs (662 keV) radiation [16]. The results showed that the tolerance of 2-hydroxypropyl cyclodextrin was maintained within the range of ± 10.9% at all irradiation doses tested, suggesting that cyclodextrin has higher radiation resistance than the hydrogel used in our study. However, the authors only conducted experiments on low-I-concentration solutions, and the saturated adsorption capacity of the cyclodextrin was not measured. Therefore, whether cyclodextrin has higher radiation resistance than our hydrogel cannot be determined. In addition, in Ihara et al.’s study [24], the dose rate of the external radiation was not provided.

Evidence of a dose-rate effect on biological cells has been reported. Specifically, even at the same cumulative absorbed dose, irradiation at a lower dose rate is less likely to cause biological effects, such as cell death, than irradiation at a higher dose rate. This finding can be attributed to the natural healing ability of biological cells. However, whether a dose-rate effect exists in the case of the SAP or hydrogel is unclear and remains a topic for further investigation.

This study did not consider the inhibition of I adsorption by the hydrogel owing to the presence of other anions in the ALPS-treated water, such as Cl− and NO3−. The applications of the hydrogel in the field can be better demonstrated by examining the concentrations of these ions in ALPS-treated water and elucidating their impact on the inhibition of I adsorption. This study revealed the feasibility of combining an SAP to retain water and cations and a hydrogel to retain anions for practical application at the FDNPP. However, the optimal conditions that can best balance the mass and timing of this combination must be further investigated for future implementation.

Conclusion

We investigated the resistance of an SAP and anion-supporting hydrogel to external radiation in the presence of water and examined the changes in their saturated adsorption capacities for water and Cs+. The saturated I adsorption capacity of the hydrogel used in this study was also investigated and found to be 575 ± 28 mg/g. This saturated adsorption capacity was maintained by samples with absorbed doses of up to 100 Gy. Although the SAP was more prone to a decrease in adsorption capacity owing to external radiation, it demonstrated good water retention and Cs+ adsorption capability at cumulative absorbed doses of up to 40 Gy. Given that the spatial dose rate at the storage location is 100 µGy/h, these materials can be practically used for approximately 45.6 years until their cumulative absorbed dose reaches 40 Gy. However, the Cs retention capacity of the SAP may decrease with increasing cumulative absorbed dose. Although the effects of the presence of ions other than Cs+ and I must be investigated, we believe that the combination of an SAP and a hydrogel can safely store treated water containing tritium for at least 30 years at the FDNPP.