3.1. Interzeolite Transformation to Sodalite
The results of the interzeolite transformation of Na-zeolites into the different structures are summarized in
Table 1 (run 1–18).
Figure 2a shows the powder XRD patterns for the representative interconverted products obtained after heating at 180 °C for 1 day (run 2, 7, 9, 12, 13, and 17), and the XRD patterns of all the others are depicted in
Figure S2. For comparison, the XRD patterns of parent Na-A and Na-X are also displayed in
Figure 2a. Since the LTA zeolite with Si/Al = 1 is known to be easily transformed into the SOD structure [
19], the interconversion of Na-A was performed under various hydrothermal synthesis conditions (run 1–10). The X-ray peaks for the LTA phase were still observed in the synthesis with ca. 0.7 M of NaOH (NaOH/Al = 1) aqueous solution at 180 °C for 1 day (
Figure S2). However, all the LTA phases disappeared and the generation of SOD was confirmed (run 1) after extended crystallization, up to two days. When the NaOH concentration had more than doubled (NaOH/Al = 2–4), a pure SOD phase was crystallized within 1 day (run 2–4). Because they were crystallized under OH
- media, these sodalities were HSOD, and their XRD patterns were in good agreement with the literature results [
12,
19,
27]. The I/Al ratio of ISOD with Si/Al = 1 was ca. 0.3, because two iodide anions fit per SOD unit cell [
26]. Thus, the interzeolite transformation was performed under the conditions of NaI/Al = 0.3 in the synthesis mixture (run 5–7). Moreover, pure SOD phases were obtained only when the NaOH/Al ratio exceeded two in the synthesis mixture.
Since Ag-zeolites are known to have better CH
3I adsorption capacity than Na-forms, we tried to perform the interzeolite transformation of Na-A in the presence of AgI instead of NaI (run 8–10). For all the runs, although no zeolitic phases other than SOD were confirmed, the X-ray peaks for bulk AgI clusters and bulk Ag metal phases at 2θ = 22.4°, 23.6°, 39.2°, and 46.2° and 2θ = 38.1° and 44.3°, respectively, were also observed together with the SOD phase. This indicates that unlike NaI, most of the AgI clusters were not dissolved under the hydrothermal condition at 180 °C and existed as a physical mixture together with the generated SOD. However, since the presence of metallic Ag was identified, it can be speculated that some portions of Ag-I were dissociated under the hydrothermal conditions, and the dissociated I
− contributed to stabilizing the formation of
sod-cage of SOD together with Na
+. On the other hand, the amount of CH
3I adsorbed using the Ag-zeolite is around 100 mg/g-zeolite [
10], which corresponds to 0.1 AgI/Al. So, only 0.1 times as much AgI as Al was added to the synthesis mixture, but the X-ray peaks for AgI and Ag were still observed (run 10). This suggests that the AgI may be difficult to immobilize on the SOD structure by interzeolite transformation, and Na-zeolites may be a better adsorbent from the immobilization point of view, even though their adsorption of CH
3I is somewhat lower than the Ag-zeolites. Since sodium is cheaper than silver, the former is also more advantageous from an economic point of view.
As described above, the large-pore zeolites like FAU have better CH
3I adsorption capacity than the small-pore materials like LTA zeolite [
6,
8]. Accordingly, the interzeolite transformation was performed for Na-X with Si/Al = 1.2, which is similar to that (1.0) of Na-A (run 11–15). The ANA phase was observed as an impurity level together with SOD regardless of the presence or absence of NaI (run 11 and 12). This suggests that the relatively excess amount of silica contributed to the stabilization of ANA phase compared to the interconversion of Na-A [
20]. Since FAU and ANA have 6-rings in common, it can be easily converted to ANA by the interconversion of FAU, especially under certain cationic systems like Na
+ and Cs
+ [
20,
21,
22]. The ANA structure is not suitable for immobilizing radioactive elements because it has 8-ring open pores (ca. 4.2 Å) [
26]. However, when comparing the two strong X-ray peaks of the ANA and SOD phases in the run 12 sample, i.e., 2θ = 26.0 and 24.4 corresponding to (400) and (211) Miller indices for ANA and SOD, respectively, the ANA portion was estimated to be only about 9% of the sample. In contrast, when AgI was added instead of NaI under the identical hydrothermal conditions (run 13–15), although the ANA phase was not observed, the X-ray peaks for bulk AgI and Ag metal phases were also observed, like the results of run 8–10.
Among the zeolite adsorbents, zeolite Y, which has a higher Si/Al ratio (2.6) than zeolite X, is known to exhibit excellent CH
3I adsorption performance [
6,
11]. So, additional interzeolite transformation was attempted for Na-Y under the same conditions described above. Only a pure ANA phase was formed, instead of SOD, and bulk AgI and Ag metal phases were also observed in the presence of NaI (run 16) and AgI (run 17), respectively. This can be explained by the much higher silica content, which preferably stabilized the ANA phase rather than SOD. When the interzeolite transformation of Na-Y was attempted with a lower Si/Al ratio of 1.0, by adding more Al precursor, the SOD phase was generated, although not purely (run 18). From the above interzeolite transformation results, it can be concluded that zeolites A and X can be used as CH
3I adsorbents to form an SOD structure by interzeolite transformation, but zeolite Y is not suitable from the immobilization point of view, even though it has superior CH
3I adsorption capacity.
Figure 3 (left) shows the SEM images of the representative interconverted products discussed above. The run 7 sample obtained from the interconversion of Na-A in the presence of NaI showed only agglomerated SOD particles ca. 5 μm in size, and there was no impurity phase like NaI. The EDS result also showed that the iodine content on the particle surface was very low, compared to the bulk analysis by ICP (I/Al = 0.01 vs. 0.16 in
Table 2). This implies that most of the iodide anions are immobilized in the
sod-cage in SOD. The SEM–EDS result of the run 12 sample, which was obtained by the interconversion of Na-X with NaI, also showed almost no iodine on the particle surface (
Table 2). In addition, as observed in the XRD, a polyhedron impurity crystal about 20 μm in size of something other than SOD particles was observed. This is consistent with the crystal morphology of a typical ANA zeolite in the literature [
21]. Meanwhile, the SEM-EDS analysis of the run 9 and 13 samples synthesized in the presence of AgI clearly showed the presence of AgI and/or Ag metal together with SOD particles, which corresponds to the XRD results (
Figure 2a).
Figure 4 shows the
127I MAS NMR spectra of the representative synthetic products produced by the interzeolite transformation. It is interesting to note that for the run 7 product, no peak was observed at 0 ppm corresponding to NaI, and only a strong
127I NMR band was observed at −260 ppm, which is a typical chemical shift for iodine immobilized in an
sod-cage [
12,
27]. This strongly indicates that almost all the iodide anions were immobilized in the
sod-cage of the SOD structure in the run 7 sample, where a pure SOD phase was formed by the interconversion of Na-A in the presence of NaI. For this product, the number of I
−/
sod-cage can be estimated to be about 0.5 based on the results of the elemental analysis (
Table 2). This means that approximately one iodide anion is immobilized in just one of the two
sod-cage per SOD unit cell. Furthermore, a strong
127I chemical shift was also observed around −260 ppm for the run 12 product, which was synthesized by the interconversion of Na-X in the presence of NaI. For this product, it can be estimated that almost one iodine is immobilized per
sod-cage in the SOD structure (
Table 2). Thus, although the ANA phase co-existed as an impurity level in the run 12 sample, most of the iodine used in the synthesis was immobilized in the
sod-cage of the mainly generated SOD structure. In contrast, for the run 9 product generated by the interconversion of Na-A with AgI, no peak was observed at −260 ppm, and only a strong band was observed near −70 ppm, which corresponds to bulk AgI clusters (
Figure S3). This is also in good agreement with the results from the powder XRD and SEM-EDS analyses. These results indicate that no iodine exists in the
sod-cages of the SOD structure generated in the run 9 product.
Figure 5 shows the I 3d XPS spectra of representative products synthesized by interzeolite transformation. There are two types of peaks in the ranges of 617–620 eV and 629–632 eV which correspond to I 3d
5/2 and I 3d
3/2, respectively [
12]. Here, the discussion is based on I 3d
5/2, because I 3d
3/2 shows the almost same trend for all samples. For the NaI cluster with the FCC structure, six Na
+ ions are coordinated around one I
− anion, and the XPS peak for I 3d
5/2 is observed at 618.4 eV [
12]. For the run 7 product, the peak for I 3d
5/2 was observed at 618.7 eV, which is the typical binding energy for I
− coordinated with four Na
+ in the
sod-cage [
12]. We should note here that although the immobilization of iodine in the
sod-cage of SOD was confirmed in the run 12 product, the XPS peak of I 3d
5/2 for this sample was observed at 619.5 eV, which is ca. 0.8 eV higher than that of the run 7 product. The binding energy of I
− anions can change depending on the type and number of surrounding cations [
12,
28]. As shown in
Table 2, since the run 12 product has the lowest Na/I ratio of 1.8 among the samples analyzed, the strongest I 3d binding energies may be observed. On the other hand, for the run 9 and 13 products, the XPS peak for I 3d
5/2 was observed at 619.3 eV, which is the typical binding energy corresponding to a bulk AgI cluster [
28].
3.2. Adsorption of CH3I and Immobilization of Iodine in Sod-Cages
Based on the characterization of products obtained from the interzeolite transformation of Na-A, Na-X, and Na-Y in the presence of NaI or AgI, the four different zeolites (i.e., Na-A, Ag-A, Na-X, and Ag-X) were prepared to compare their CH
3I adsorption properties and the immobilization of iodine by successive interzeolite transformation. As shown in
Figure S4, the XRD patterns of the Ag
+-exchanged pristine Ag-A and Ag-X zeolites displayed no other diffractions like Ag metal or Ag
2O besides the LTA and FAU phases, respectively [
29]. This supports the conclusion that the Ag
+ ions were mostly dispersed in the zeolite pores and did not diffuse out onto the external surface of the zeolite during the preparation, even after the final high temperature calcination. As shown in
Table 3, the Ag contents of the two Ag-zeolites prepared using the identical ion-exchange method are quite similar, and ca. 48% and 43% of the anion sites generated by the framework Al were exchanged with Ag
+, respectively. From the elemental analysis of commercial Na-A and Na-X, it was confirmed that Na-A (Na/Al = 0.89) contained twice as much Na as Na-X (Na/Al = 0.43).
Figure 6 and
Table 3 show the results of CH
3I adsorption at 100 °C for the four zeolites. While Na-A hardly adsorbed CH
3I, the other three zeolites were almost saturated after 2 h of adsorption, and the total amount of adsorbed CH
3I was 140–280 mg/g-zeolite, which is similar to the values reported in the literature [
6,
10]. Unlike Na-A, the Ag-A showed a CH
3I adsorption capacity of about 260 mg/g-zeolite. Although the adsorption by Na-X was less than Ag-A, it showed a moderate CH
3I adsorption capacity of 140 mg/g-zeolite. Ag-X showed about double the CH
3I adsorption capacity of Na-X. As can be easily predicted from these adsorption results, it thus appears that Ag
+ is better than Na
+, and zeolite X is better than zeolite A in terms of CH
3I adsorption capacity.
As shown in
Figure 2b, after CH
3I adsorption the Na-A-ad and Na-X-ad samples showed no X-ray peaks other than the LTA and FAU phases, respectively. Although no X-ray peaks for bulk AgI clusters were observed in the Ag-A-ad, the characteristic peaks of AgI were observed in the Ag-X-ad. As described above, one of the CH
3I adsorption mechanisms on the metal sites is the formation of NaI or AgI species, where the successive dissociation of CH
3-I and the combination of metal iodides occur [
4,
5,
9]. However, XRD detectable AgI may not be formed inside such small zeolite micropores, and therefore AgI clusters or larger particles could form on the external surface of the Ag-X-ad crystallite (
Figure S5). It has been reported that a rapid growth of AgI species occurred inside the super cage of the Ag-Y zeolite, and at the same time they diffused out toward the external surface as the spent adsorbent was exposed to ambient conditions [
5]. However, since the pore size of LTA is much smaller than that of FAU, it is relatively difficult for AgI formed inside the pores to get out of the pores even after exposure to air, and thus the characteristic AgI X-ray peaks may not be detectable for Ag-A-ad.
The ability to immobilize iodine in the
sod-cage of the SOD structure by interzeolite transformation was evaluated for the four CH
3I adsorbed samples discussed above. These synthesis results are also summarized in
Table 1 (run 19–22). Although a pure SOD phase (Na-A-it in
Figure 2b and
Figure 3 (right)) was interconverted from Na-A-ad after only 1 day of crystallization with excessive NaOH concentration (NaOH/Al = 4), no characteristic peaks were observed in the
127I MAS NMR (
Figure 4) and I 3d XPS (
Figure 5) spectra because the amount of CH
3I adsorption was insignificant, as shown in
Figure 6 and
Table 3. The Ag-A-it sample also formed an SOD phase under the same interconversion conditions as Na-A-it, but the Ag metal phase was also observed. Interestingly, in this case, unlike the results of interconversion of Na-A with AgI (run 8–10 in
Table 1), the bulk AgI phase was not observed. As mentioned above, it can be speculated that AgI molecules or nanometric species formed inside the LTA pores after CH
3I adsorption may dissociate under the interzeolite hydrothermal conditions, unlike the bulk AgI clusters, and accordingly, the dissociated Ag
+ may organize a metal cluster. Moreover, the dissociated I
− binds with Na
+ in the synthesis mixture, which can contribute to stabilizing the
sod-cage of the generated SOD. However, as shown in
Figure 4 and
Figure 5, in the case of Ag-A-it, the proportion of iodide immobilized in the
sod-cage was not so high.
Since the interzeolite transformation of FAU may generate the ANA phase with excessive concentrations of sodium as well as silica (run 11, 12, 16, and 17 in
Table 1), in the interconversion of Na-X-ad, the concentration of NaOH was lowered to 1/2 compared to those with Na-A-ad and Ag-A-ad (run 21 in
Table 1). However, the ANA phase was still formed at a proportion of about ca. 15% compared to SOD (Na-X-it in
Figure 2b), which is similar to the result for run 12. We should note here that both the
127I MAS NMR (
Figure 4) and I 3d XPS (
Figure 5) results for Na-X-it were also similar to those of run 12. It thus appears that the iodide anions were immobilized in the
sod-cage of SOD. The elemental analysis result for Na-X-it was also similar to the run 12 product, and the number of I
−/
sod-cage in the SOD structure was estimated to be 0.8, which is slightly less than that (1.0) of run 12 (
Table 2).
On the other hand, ANA and Ag metal phases were identified along with the SOD phase from the interconversion of Ag-X-ad with the identical hydrothermal conditions for the Na-X-ad (run 22 in
Table 1). Like Ag-A-it, in this run, the bulk AgI phase was not identified (Ag-X-it in
Figure 2 and
Figure 3 (right)). However, as shown in
Figure 4, a weak band for AgI was confirmed around −70 ppm with a strong band at −260 ppm in the
127I MAS NMR spectrum of Ag-X-it. This indicates that some of the AgI clusters in Ag-X-ad, identified in the XRD result (
Figure 2b), remained even after the interzeolite transformation. Even though Ag-X-ad adsorbed the highest amount of iodine among the four adsorbents (
Figure 6 and
Table 3), the iodine in Ag-X-it was only half that of Na-X-it (I/Al = 0.13 vs. 0.26 in
Table 2), so the number of I
−/
sod-cage in SOD was also calculated to be half (0.4 vs. 0.8 in
Table 2).
The comprehensive results for CH
3I adsorption and hydrothermal interzeolite transformation revealed that the Na-form zeolite has a lower CH
3I adsorption capacity than Ag-form, but a higher iodine immobilization capacity. In addition, unlike Na-A, the interconversion of Na-X to pure SOD is relatively difficult, but the CH
3I adsorption capacity of Na-X is much higher. Therefore, among the four adsorbents, the Na-X zeolite was determined to be the best adsorbent. In Na-X, the amount of adsorbed iodine measured after CH
3I adsorption and the continuous desorption of physisorbed CH
3I was about 120 mg/g-zeolite (i.e., ca. 120,000 ppm) (
Table 3). Meanwhile, the concentration of iodine in the Na-X-it solid sample, and the synthetic solution remaining after hydrothermal interzeolite transformation, were determined by analyses to be 117,000 and 2220 ppm, respectively. These results confirmed that the mass balance was well matched within the error range (i.e., 120,000~117,000 + 2220 ppm). Consequently, with the Na-X adsorbent, iodine was immobilized in the
sod-cage of the SOD structure at a very high proportion of ca. 98%, through a successive interzeolite immobilization process after CH
3I adsorption.