Abstract
BACKGROUND:
Removal of radioactive substances, such as cesium (Cs) and strontium (Sr), has become an emerging issue after the Fukushima Daiichi nuclear power plant disaster. It has been reported that hydroxyapatite (HA) and aluminosilicate composite powders can be used to remove Cs and Sr. However, the film type of these materials for the removal of Cs and Sr has not been reported.
OBJECTIVE:
The aim of this study was to assess the possibility of using HA, aluminosilicate, and aluminosilicate/HA composites for the removal of Cs and Sr radioactive substances.
METHODS:
Aluminosilicate films and HA films were fabricated using a sputtering technique with diatomaceous earth and HA targets, respectively. The aluminosilicate film was observed by X-ray diffraction (XRD) and scanning electron microscopy (SEM). A comb-shaped HA/aluminosilicate composite film was prepared to take advantage of the adsorption properties of the HA and the aluminosilicate films. The Cs and Sr adsorption on these films were also evaluated.
RESULTS:
In the XRD patterns, the film sputtered from a diatomaceous earth target under 5.0 Pa of Ar pressure showed aluminosilicate peaks (Na1.82(Al2Si3O10) and Al2SiO5) after 8 h of vapor-phase hydrothermal treatment. The film showed higher adsorption of Cs than Sr in Cs and Sr solutions, while the HA film adsorbed far more Sr than Cs. A HA/aluminosilicate composite film was successfully fabricated, and the SEM images showed that the width of the HA region was 230–260 μm, and that of the aluminosilicate region was 170–200 μm. The HA/aluminosilicate composite film showed 84.8 ± 11.5% Cs adsorption and 28.3 ± 1.4% Sr adsorption in a mixed solution of Cs and Sr.
CONCLUSION:
This study shows the feasibility of using HA films, aluminosilicate films, and HA/aluminosilicate composite films for the removal of radioactive substances such as Cs and Sr.
Introduction
Radioactive waste is harmful to living organisms, natural resources, and the environment. The accident at the Fukushima Daiichi nuclear power plant that was caused by the Great East Japan Earthquake on March 11, 2012, reminded us of the serious hazards of radioactive substances when spread over a wide area. Radioactive cesium (137Cs) and strontium (90Sr) were emitted from the plant and diffused into the environment after the accident. There is still great concern over the effect of contaminated soil and water on the health and safety because the half-life of 137Cs and 90Sr is 30.2 years and 28.8 years, respectively [1]. From this experience, the decontamination of water, including the removal of radioactive substances, is regarded as a worldwide problem [2].
Aluminosilicate is a candidate as an adsorbent for 137Cs because the aluminosilicate adsorbs Cs+ ions by the ion exchange reaction [3]. Zeolites (such as mordenite) and montmorillonite, which are both a type of aluminosilicate, have a high affinity for 137Cs [4–7].
On the other hand, hydroxyapatite (Ca10(PO4)6(OH)2, HA) is known to have a high adsorption affinity for Sr2+ ions by ion exchange with Ca2+ ions. HA has a high adsorption ability for various cations as well as Sr, and exhibits high stability in an alkali solution [8–10]. Therefore, a mixture of aluminosilicate and HA can be expected to adsorb both Cs and Sr. Ozeki et al. reported high Cs and Sr adsorption of a mixture of aluminosilicate and HA powders [11].
These adsorbents have been used in powdered or solid form. For the effective decontamination of radioactive substances such as 137Cs and 90Sr, adsorbent films, as well as powder and solid adsorbents, are needed under various decontamination conditions. It is also expected that HA and the aluminosilicate films could adsorb Cs and Sr ions. Several film deposition techniques for adsorbents are being investigated [12–15]. One of these is sputtering deposition, which involves bombarding a target with Ar+ ions, scattering material, and depositing it onto a substrate. The deposited film has a high density and has high adhesion to the substrate because the ions have high energy (1–100 eV) compared with the energy (approximately 0.1 eV) of evaporated atoms by a vacuum deposition method [16]. The preparation of HA films by sputtering has been reported for use in medical devices such as dental implants and artificial joints [15], however, the adsorption ability of Sr in HA films has not yet been reported. For aluminosilicate, some studies have reported mordenite or zeolitic films deposited by pulsed laser deposition (PLD) [12,14], but the adsorption ability of Cs on these films has not been investigated. There are presently no reports describing the preparation of aluminosilicate films with the sputtering technique. In the present study, we fabricated an aluminosilicate film, HA film, and HA/aluminosilicate composite film by sputtering deposition, and investigated their Sr and Cs adsorption capacities.
Materials and methods
Materials
The substrates used were Ti plates (10 × 5 × 1 mm3). Powdered diatomaceous earth (Wako Corp., Japan) and powdered HA (Taihei Corp., Japan) were used as sputtering targets for the aluminosilicate film and HA film, respectively. The Si/Al ratio of the diatomaceous earth was 2.47 ± 0.12.
Preparation of the films
Aluminosilicate film
RF magnetron sputter coating was performed using a magnetron sputtering deposition system (L-210HS-D; Anelva Corp., Japan). An RF generator operating at 13.56 MHz was coupled to the target electrode through an impedance-matching network. The sputtering chamber was evacuated to a pressure below 1 × 10−5 Pa using an oil-diffusion pump equipped with a liquid nitrogen trap. Argon gas (99.999%) was then introduced into the chamber using a mass flow controller at a constant flow rate of 1.5 sccm. Before deposition occurred, the target was covered with a shield and pre-sputtered using Ar ions for 15 min. The aluminosilicate film was coated using an Ar gas pressure of 0.5–5.0 Pa and a discharge power of 100 W to obtain a 1-μm-thick film on the both sides of the Ti substrate.
The coated samples were then set in a PTFE-lined stainless steel vessel, as shown in Fig. 1, for the vapor-phase hydrothermal treatment to crystallize the film [12]. The vapor-phase hydrothermal treatment was performed at 120 °C and 170 °C with distilled water for 1–24 h to recrystallize the aluminosilicate film.

A schematic of the vapor-phase hydrothermal method and the vessel used.
After the treatment, the films were identified by X-ray diffractometry (XRD) with a Cu K𝛼 radiation source operating at 40 kV with a 30 mA excitation current. Surface observation of the films was carried out using scanning electron microscopy (SEM; VE-9800; KEYENCE) with an accelerating voltage of 20 kV. The X-ray photoelectron intensities of the Si and Al elements of the films were measured by X-ray photoelectron spectroscopy (XPS; JPS-9010TR; JEOL) with Mg K𝛼 radiation. Each final result was obtained from an average of three samples.
The HA film was prepared using a magnetron sputtering deposition system following the procedure reported by Ozeki et al. [17]. The film was coated using an Ar gas pressure of 0.5 Pa and a discharge power of 100 W to obtain a 1-μm-thick film on both sides of the Ti substrate. The hydrothermal treatment was then performed at 120 °C and 0.2 MPa in a NaOH solution (pH = 9.5) in a stainless-steel vessel for 24 h to recrystallize the HA film [18]. After the treatment, the films were washed in distilled water to remove any residual NaOH solution.
Preparation of the HA/aluminosilicate composite film
A 1-μm-thick HA film was coated on the titanium substrate as described in Section 2.2.2. The HA-coated plates were then covered with 1.5 mm-thick stainless mask (composed of aligned rectangular holes, 150 μm wide) placed in the substrate chamber in the sputtering equipment for the aluminosilicate coating. The aluminosilicate film was coated using an Ar pressure of 5.0 Pa and a discharge power of 100 W. The HA/aluminosilicate composite films were coated as shown in Fig. 2, and the films were coated on both sides of the Ti substrate. After the coating, the films were treated with vapor-phase hydrothermal treatment at 170 °C above distilled water for 24 h to recrystallize the aluminosilicate film. The prepared film pattern was observed by SEM with an accelerating voltage of 20 kV.

A schematic of the HA/aluminosilicate film.
Six plates of the HA, aluminosilicate, and HA/aluminosilicate composite films were immersed in 2.0 mL of Cs, Sr, or a mixture of Cs and Sr solutions (1.0 × 10−5 M) for 30 min at room temperature (25 °C). The preparation conditions of all the aluminosilicate films used were 100 W and 5.0 Pa of sputter coating and 170 °C for 24 h of the vapor-phase hydrothermal treatment. Samples of each film were separated using the filtration method after immersion, and the Cs or Sr concentrations before and after the removal operations were analyzed using inductively coupled plasma mass spectroscopy (ICP-Mass; 7500CX, Agilent Technologies). The ratio of the adsorbed Cs or Sr was calculated according to the following equation:
Statistical analysis
For all test groups, the results from each experiment are expressed as mean ± standard deviation (SD). The statistically significant difference (p > 0.05) was determined using the Student’s t test.

The XRD patterns of the aluminosilicate film at 0.5 Pa of Ar pressure after vapor-phase hydrothermal treatment at (a) 120 °C and (b) 170 °C.
XRD patterns of aluminosilicate and HA films
Figure 3 shows the XRD patterns of the aluminosilicate films coated using an Ar gas pressure of 0.5 Pa with the vapor-phase hydrothermal treatment time. In the film treated at 120 °C, only peaks from the titanium substrate were observed (2𝜃 = 35.1°, 38.4°, and 40.1°) (Fig. 3(a)), whereas in the film treated at 170 °C, the Al2SiO5 peak at 2𝜃 = 26.0° appeared after 24 h of vapor-phase hydrothermal treatment (Fig. 3(b)). Figure 4 shows the XRD patterns of the aluminosilicate films coated using an Ar gas pressure of 5.0 Pa with the vapor-phase hydrothermal treatment time. In Fig. 4(a), peaks from Na1.82(Al2Si3O10) were observed at 2𝜃 = 14.7° and 29.6°, in addition to a peak from Al2SiO5 (2𝜃 = 26.0°) after 8 h of vapor-phase hydrothermal treatment. In the film treated at 170 °C, peaks from Na1.82(Al2Si3O10) appeared after 8 h of vapor-phase hydrothermal treatment (Fig. 4(b)). Peaks from both Na1.82(Al2Si3O10) and Al2SiO5 were observed at 120 °C in the vapor-phase hydrothermal treatment, whereas only peaks from Na1.82(Al2Si3O10) were observed at 170 °C. These results indicate that Na1.82(Al2Si3O10) is more stable than Al2SiO5 at high-temperatures. At an Ar gas pressure of 0.5 Pa, no peak was observed except for the 24 h vapor-phase, which indicated that the film sputtered at 5.0 Pa favored crystal growth more than the film sputtered at 0.5 Pa.

The XRD patterns of the aluminosilicate film at 5.0 Pa of Ar pressure after vapor-phase hydrothermal treatment at (a) 120 °C and (b) 170 °C.
Figure 5 shows the XRD patterns of an as-sputtered HA film (0 h) and a sputtered film subjected to hydrothermal treatment (24 h). In the as-sputtered film, no peak was observed except for peaks from titanium substrate. After the 24 h-hydrothermal treatment, additional HA peaks appeared, indicating the growth of HA crystals. These results are consistent with the findings of previous studies [18].

The XRD patterns of the HA film after the hydrothermal treatment.
Figure 6 shows the Si/Al ratio of the aluminosilicate films. In the films with 0.5 Pa of Ar pressure, the Si/Al ratio of the as-sputtered film at 0.5 Pa was 3.26, which is higher than that of the target materials (2.47) (Fig. 6(a)). The Si/Al ratio of the as-sputtered film at 5.0 Pa was 2.76, which was lower than that (3.26) at 0.5 Pa (Fig. 6(b)).

The Si/Al ratio of the aluminosilicate film coated at (a) 0.5 Pa and (b) 5.0 Pa of Ar pressure after vapor-phase hydrothermal treatment. *P < 0.05.
After vapor-phase hydrothermal treatment at 120 °C, the ratio decreased with the treatment time. The decrease in the ratio of the film treated at 170 °C was larger than that at 120 °C.
Figure 7 shows SEM images of the aluminosilicate films prepared at 0.5 Pa after the vapor-phase hydrothermal treatment. No change was observed in the surface topology of the film under any of the treatment conditions. This is in agreement to the result of the XRD patterns in Fig. 3, where only the peak corresponding to the film treated at 170 °C for 24 h was observed. Figure 8 shows the SEM images of the aluminosilicate films prepared at 5.0 Pa after the vapor-phase hydrothermal treatment. Particles were observed on the surface after the treatment. At 120 °C, the particles appeared after 16 h of vapor-phase hydrothermal treatment, and at 170 °C, the particles were observed after 8 h of treatment. The width of the particles was approximately 50–150 nm. These particles may correspond to crystal phases of Na1.82(Al2Si3O10) and Al2SiO5 because the peaks from Na1.82(Al2Si3O10) and Al2SiO5 were observed in the XRD patterns of vapor-phase hydrothermally treated films prepared at 5.0 Pa (Fig. 4(a) and (b)).

Scanning electron micrograph of the aluminosilicate films at 0.5 Pa of Ar pressure after vapor-phase hydrothermal treatment at 120 °C and 170 °C.

Scanning electron micrograph of the aluminosilicate films at 5.0 Pa of Ar pressure after vapor-phase hydrothermal treatment at 120 °C and 170 °C.
Figure 9 shows an SEM image of the HA/aluminosilicate composite film. The HA and aluminosilicate regions were coated separately. The width of the HA region was 230–260 μm, and that of the aluminosilicate region was 170–200 μm. This difference may be influenced by the dimensional accuracy of the mask pattern. The area ratio of the HA/aluminosilicate was approximately 1.2–1.5. Needle-like HA crystals were observed in the HA region, and aluminosilicate pillars and particles were observed in the aluminosilicate region. The width of the pillars in the aluminosilicate region was thicker than that of the needle-like HA crystals found in the HA region because the aluminosilicate pillars might grow from the surface of the needle like HA crystal. In the preparation of the HA/aluminosilicate composite, HA was coated on the titanium, and then the aluminosilicate was coated on the HA surface.

Scanning electron micrograph of the HA/aluminosilicate composite films.
Figure 10 shows Cs and Sr adsorption on the HA, aluminosilicate, and HA/aluminosilicate composite films. In Fig. 10(a), the HA film adsorbed 0.7 ± 0.8% of Cs in the Cs solution, whereas the film adsorbed 17.3 ± 1.8% of Sr in the Sr solution. The adsorption of Cs and Sr on the HA films in the mixture of Cs and Sr solution were 1.2 ± 1.5% and 18.3 ± 1.8%, respectively. The amount of Sr adsorption on the HA films was much higher than that of Cs, and there was no significant difference in the Cs and Sr adsorption behavior of the separated Cs and Sr solutions and the mixed solution of Cs and Sr.

The adsorption ratio of Cs ions and Sr ions on (a) the HA films, (b) the aluminosilicate films, and (c) the HA/aluminosilicate composite films in the Cs, Sr, and mixed solution of Cs and Sr. *P < 0.05.
As shown in Fig. 10(b), the aluminosilicate film adsorbed 61.8 ± 2.9% of Cs in the Cs solution, whereas the film adsorbed 8.7 ± 2.1% of Sr in the Sr solution. The adsorption of Cs and Sr on the aluminosilicate films in the mixed solution of Cs and Sr was 59.3 ± 4.9% and 8.2 ± 2.1%, respectively. The aluminosilicate film has a much higher Cs adsorption than Sr adsorption.
As shown in Fig. 10(c), the HA/aluminosilicate composite film adsorbed 76.3 ± 10.5% of Cs in the Cs solution, and the film adsorbed 26.6 ± 1.5% of Sr in the Sr solution. On the other hand, Cs adsorption on the aluminosilicate film and HA film were 61.8% and 0.7% in the Cs solution, respectively (Fig. 10(a) and (b)). Even though the aluminosilicate region of the HA/aluminosilicate composite film comprises only half the total area of the pure aluminosilicate film, the composite film showed a 76.3% Cs adsorption, which is higher than that of the aluminosilicate film alone (61.8%).
The composite film showed 26.6% Sr adsorption, which is higher than that of the pure HA film (14.3%), even though the HA comprises only half the area of the composite film. The adsorption of Cs and Sr on the composite film from the mixed solution of Cs and Sr was similar to that of the separate Cs and Sr solutions. The composite film showed 84.8 ± 11.5% Cs adsorption and 28.3 ± 1.4% Sr adsorption in the mixed solution of Cs and Sr.
The Si/Al ratio of the as-sputtered aluminosilicate films is higher than that of the target materials (2.47) (Fig. 6). This can be explained by the difference in the sputter yields of Al2O3 and SiO2, which comprise the diatomaceous earth used as the target material. The deposition rate of SiO2 is 1.5–2.5 times higher than that of Al2O3 because of the difference in the binding energy and thermal diffusivity between SiO2 and Al2O3 [19–21]. A higher deposition rate of SiO2 than that of Al2O3 leads to a higher Si/Al ratio of the coated film than that of the target material.
After vapor-phase hydrothermal treatment at 120 °C, the ratio slightly decreased with the treatment time. The decrease in the ratio of the film treated at 170 °C was larger than that at 120 °C. This indicates that higher temperatures promote the dissolution of Si in the film because the solubility of SiO2 is higher than that of Al2O3. The solubility of SiO2 (quartz) is approximately 50 ppm at 100 °C, and increases with temperature. On the other hand, the solubility of Al2O3 was quite low, and its solubility was 1.8 ppm even at 500 °C [22,23].
The Si/Al ratio of the as-sputtered film at 5.0 Pa was 2.76, which was lower than that (3.26) at 0.5 Pa. A decrease in the Ar pressure leads to an increase in the energy of Ar+ ions because the mean free pass of Ar+ ions increases during the sputtering process [16]. An increase in the energy of Ar+ ions leads to an increase in sputtered SiO2 rather than Al2O3 because SiO2 is more readily sputtered than Al2O3. This is because the deposition rate of SiO2 is higher than that of Al2O3, as described above. Therefore, a lower Ar pressure such as 0.5 Pa results in a higher Si/Al ratio.
The Si/Al ratio of the film at 0.5 Pa ranged from 2.89 to 3.26, which is higher than that of the film at 5.0 Pa (1.17–2.76) because of the difference in the energy of the Ar+ ions. In the XRD patterns, the crystal growth favored the film with 5.0 Pa, compared with the film with 0.5 Pa (Figs 3 and 4). The film at 5.0 Pa showed peaks corresponding to Na1.82(Al2Si3O10) and Al2SiO5. The Si/Al ratios of Na1.82(Al2Si3O10) and Al2SiO5 are 1.5 and 0.5, respectively. These crystals have a low Si/Al ratio. This indicates that the films with a low Si/Al ratio promote the growth of Na1.82(Al2Si3O10) and Al2SiO5 crystals.
The amount of Sr adsorption on the HA films was more than 10 times that of Cs. In the HA crystal, cations larger than Ca should be accommodated at the Ca site, and Sr2+ can replace Ca2+in the structure of the HA over the entire range of composition. Sr2+ is substituted at the Ca site because Sr2+ (ionic radius 0.113 nm) is slightly larger than Ca2+ (ionic radius 0.099 nm) [24]. On the other hand, the replacement of Ca2+ by Cs+ is unfavorable because of the large difference in ionic radius and valence between Cs+ and Ca2+. The ionic radius of Cs+ (0.170 nm) was much larger than that of Ca2+ (0.099 nm) [25].
The aluminosilicate film has a much higher Cs adsorption than Sr adsorption. The aluminosilicate film prepared under the sputtering conditions of 100 W and 5.0 Pa and a vapor-phase hydrothermal temperature of 170 °C was identified as Na1.82(Al2Si3O10) from the XRD pattern (Fig. 4(b)). Cs+ can be exchanged with Na+ in the Na1.82(Al2Si3O10). However, the ratio of Sr adsorption on the film was much lower than the ratio of Cs adsorption because Na1.82(Al2Si3O10) does not include a divalent cation, such as Ca2+ and Mg2+. Aluminosilicates containing Ca2+ and Mg2+ are more likely to adsorb Sr2+ [26].
The adsorption ratio of Cs and Sr of the aluminosilicate/HA composite film was higher than that of the pure aluminosilicate and the pure HA films (Fig. 10). The high adsorption ability of the composite film might be related to the high surface roughness of the aluminosilicate part of the composite film. In the composite film, the aluminosilicate film was coated on the HA surface, which has a complex rough surface with needle-like HA crystals (Fig. 9). The aluminosilicate film has a rough surface, giving it a high surface area which contributes to an increase in Cs and Sr adsorption.
Conclusions
The aluminosilicate, HA and HA/aluminosilicate composite films were fabricated by a sputtering method, and we investigated their adsorption capacity for Sr and Cs. The following conclusions were drawn: The aluminosilicate film was successfully fabricated using the sputtering technique. In the XRD patterns, the film sputtered from a diatomaceous earth target under 5.0 Pa of Ar pressure showed aluminosilicate peaks (Na1.82(Al2Si3O10) and Al2SiO5) after 8 h of vapor-phase hydrothermal treatment. In the SEM images, the film sputtered from the diatomaceous earth target under 5.0 Pa of Ar pressure showed particles after the 8 h vapor-phase hydrothermal treatment. The width of the particles was approximately 50–150 nm. The HA/aluminosilicate composite film showed 84.8 ± 11.5% Cs adsorption and 28.3 ± 1.4% Sr adsorption in the mixed solution of Cs and Sr. The amount of Cs and Sr adsorbed on the composite films was higher than that on the aluminosilicate and HA films.
Footnotes
Acknowledgements
This research was partially supported by KAKENHI (no. 15K00578).
Conflict of interest
The authors declare no conflict of interest.
