XRD patterns of control, foamed, and hydrothermally treated samples are shown in Figure 2. All the untreated samples (MS, MS-F4, MS-N8) showed the common phases of geopolymers, i.e., quartz (PDF #01-086-1629, #01-089-8936), mullite (PDF #01-074-4145, #01-079-1450), and a typical diffused hump representing aluminum substituted C-A-S-H gel (Kim et al., 2017; Park et al., 2018), in addition to the peaks of the internal standard. It is well known that the quartz and mullite are the inherited phases of fly ash which mostly stay inert under normal curing conditions, therefore, these phases are often detected in XRD of geopolymers (Ma et al., 2016).

The hydrothermal treatment of samples resulted in formation of zeolite Na-P1, confirmed through relevant XRD peaks (PDF #01–074-1787, #01-071-0962). Additionally, quartz peaks were significantly reduced in these samples showing its involvement in the reaction. This suggests that the reactivity of geopolymers can be enhanced under high temperature and pressure conditions. This enhanced reactivity under hydrothermal conditions is particularly significant as it demonstrates the potential for accelerated zeolite formation, a property that could be exploited in radioactive waste treatment where rapid immobilization of contaminants is often critical. The dissolution and recrystallization of quartz under these conditions also suggests that even relatively inert phases can be mobilized to contribute to the formation of more reactive zeolitic structures.

Lastly, it should be noted that no significant differences were observed in the XRD spectra of un-foamed, foamed, and different NaOH molarity samples. All the samples showed same crystalline phases. This could be attributed to the fact that the alkali activation of fly ash under hydrothermal conditions mainly results in the formation of zeolite Na-P1, irrespective of the minor changes in the mixture composition as has been reported in the literature (Aldahri et al., 2016; Zheng et al., 2019). Moreover, H₂O₂ foaming agents used in this study only interacted with the liquid matrix, decomposing to H₂O and O₂ (Ducman and Korat, 2016), hence only the physical changes were expected as observed in SEM images shown in the next section.

SEM images were taken to study the morphology of synthesized zeolite Na-P1 crystals as well as the formation of pores in foamed samples. SEM images of samples before and after hydrothermal treatment are shown in Figure 3. It can be clearly seen that the control samples mainly had amorphous gel phases. In addition, different-sized pores were easily seen in foamed samples due to the foaming action. Alternatively, zeolite Na-P1 crystals were seen in all the hydrothermally-treated samples. Similar morphology of zeolite Na-P1 crystals were also reported by Cardoso et al. (2015b), Cardoso et al. (2015a).

Another observation made from SEM is that the Na-P1 crystals in MS-N8-H samples were not fully developed yet. This might be attributed to the relatively high molarity of NaOH solution used in these samples. High molarity alkali solutions improve geopolymer strength by forming denser gel matrices (Mustafa Al Bakri et al., 2012; Palanisamy and Kumar, 2018), which inhibit the diffusion and reorganization required for complete zeolite crystallization. Hence, prolonged hydrothermal treatment might be required to fully develop Na-P1 crystals in these samples. This reveals an important trade-off: higher NaOH molarity provides more Na⁺ for ion exchange but simultaneously retards zeolite formation. Similar observations were made in previous studies (Lee et al., 2016; Khalid et al., 2018). Understanding this balance is crucial when optimizing materials for specific applications.

The pores characteristics (i.e., BET surface area, BJH pore volume and average pore diameter, and pore size distribution (PSD) of samples are summarized in Table 3; Figure 4. The control samples (not treated under hydrothermal conditions) showed lower surface area and pore volume compared to the hydrothermally-treated samples, e.g., the surface area and pore volume of sample MS-H were 101.1 m²/g and 0.16 cm³/g compared to only 25.5 m²/g and 0.08 cm³/g, respectively, for respective control sample MS. This confirms that the formation of zeolite crystals results in substantial increase of surface area which is normally desired for adsorbents (Abazari et al., 2019). Moreover, it can be noted that the samples with foaming agent (MS-F4) and high molarity NaOH solution (MS-N8) had less surface area and pore volume compared to the control samples MS. This can be because of foaming agent mainly give rise to the formation of macropores which are out of the detection range of nitrogen sorption method (>100 nm). The lower surface area and pore volume of high molarity samples can be related to the discussion in the preceding section i.e., high molarity of NaOH is known to form denser matrix (Mustafa Al Bakri et al., 2012; Palanisamy and Kumar, 2018) that potentially resulted in decrease of surface area and pore volume.

It was further noted from Figure 4 that the PSD curves of hydrothermally-treated samples shifted a little to the left side, i.e., towards the smaller pore sizes, thus the average pore size of these samples was less than the control samples (Table 3). It can be stated that the formation of zeolitic crystals under hydrothermal conditions caused the pore refinement, resulting in smaller connected pores; consequently, giving rise to the surface area and pore volume contrary to the foaming action. This pore refinement phenomenon has significant implications: while it increases surface area that is typically considered beneficial for adsorption, it simultaneously restricts the diffusion pathways for adsorbate species, particularly in bulk materials. This creates a nuanced relationship between pore structure and adsorption performance that is explored in detail in Section 3.2. The pore refinement represents restructuring at the mesopore scale (2–50 nm), where zeolite formation creates new channels while blocking original gel pores. This explains how materials can simultaneously exhibit higher surface area but reduced accessibility.

The adsorption-desorption isotherms of samples are shown in Figure 5. The isotherms were of type IV with hysteresis loops somewhat similar to H4 type (Sing et al., 1982). This confirms the existence of mesoporous networks, characteristic of materials with both microporous zeolite crystals and mesoporous geopolymer gel. This hierarchical pore structure is particularly advantageous for applications requiring both high selectivity (from micropores) and efficient mass transport (from mesopores), such as in radioactive waste streams containing multiple contaminant species of different sizes.

Infrared spectra of samples are shown in Figure 6. The bands with peaks around 668, 876, 980–995, 1475, and 1635 cm⁻¹ were observed. The small peak at 668 cm⁻¹ is due to the symmetrical stretching of Si-O-T (T: Si or Al) bonds (Bernal et al., 2010). The peak around 877 cm⁻¹ is representing asymmetric stretching of AlO₄ ⁻ groups (Bernal et al., 2010). The broad band around 1000 cm^-1 is the characteristics band of geopolymers and zeolites due to asymmetric stretching of Si-O-T in SiO₄ tetrahedra (Lecomte et al., 2006; Visa and Popa, 2015). The band around 1475 cm^{− 1} and shoulder around 870 cm⁻¹ represent anti-symmetric stretching and out-of-plane bending vibrations of O-C-O bonds in the carbonate group (CO₃ ^2-) (Lecomte et al., 2006; Siyal et al., 2016). The band at 1635 cm⁻¹ represents the bending vibration of O-H bonds in molecular water bound in the hydrated products (Siyal et al., 2016).

The adsorption capacities of geopolymer-zeolite beads for Pb(II) and BG dye are summarized in Table 4. The maximum adsorption capacities of about 130 and 115 mg/g were noted for Pb(II) and BG, respectively. These values represent a significant advancement over most bulk-type geopolymeric adsorbents reported in the literature, demonstrating that careful control of synthesis parameters can yield materials with performance approaching that of pulverized adsorbents while maintaining the practical advantages of bulk forms.

Among the samples studied, it was found that the hydrothermally-treated samples exhibited relatively less adsorption capacities for both the contaminates, compared to their respective control samples. Both the geopolymers and zeolites have shown high adsorption potential in the pulverized form, owing to the direct interaction with adsorbate ions (He et al., 2021; Wang C. et al., 2021). In the case of bulk-type samples, low permeability hinders the transportation of adsorbate species within the bulk matrix. The conversion of some of the amorphous gel in geopolymers to crystalline zeolites results in increased surface area and pore volume, which should help in the diffusion of adsorbate species within the matrix of these bulk-type samples. Thus, the adsorption capacity was expected to increase. However, the current results were found to be the opposite. This counter-intuitive finding represents a key novelty of this study and challenges the conventional wisdom that higher surface area automatically translates to better adsorption performance in bulk adsorbents. This highlights that some other parameters also play a critical role in the adsorption efficiency of bulk-type samples.

In our previous study (Khalid et al., 2018), the average pore diameter was found to play a key role in the adsorption capacity of these bulk-type adsorbents, regardless of the presence of zeolites and/or higher surface area. Therefore, adsorption capacities were plotted against pore diameter, pore volume, and surface area in Figure 7. The comparison gave conclusive results in the case of Pb(II) with relatively high R-square values. The adsorption capacity was directly related to the average pore diameter, while it decreased with the pore volume and BET surface area. The same trend was followed by BG, despite a lower R-square value. This significant finding establishes a clear hierarchy in the importance of pore characteristics for bulk-type adsorbents: pore diameter > pore volume > surface area. This hierarchy can be explained by mass transport limitations, i.e., larger pores facilitate easier diffusion of adsorbate species deep into the material, allowing more of the internal surface area to be accessed during the experimental timeframe. The dominance of pore diameter can be further explained by tortuosity effects: smaller pores increase diffusion resistance, preventing adsorbate species from reaching interior sites within the 120 h experimental timeframe. This is particularly significant for larger molecules like BG dye, which experience steric hindrance in narrow pores. This insight has profound implications for rational material design, suggesting that for bulk applications, strategies to maintain or increase pore diameter should be prioritized over those that simply maximize surface area.

It was observed that the 5 mm bead samples either showed comparable (MS5-F4-H) or better (MS5-H) adsorption efficiency compared to their 7 mm counterpart samples. The smaller beads tend to offer more exposed areas, potentially resulting in enhanced interaction with the adsorbate solution. This can potentially result in better adsorption efficiency. The improved performance of smaller beads also reflects shorter diffusion pathways, reducing the time required for adsorbate species to reach interior sites. This effect becomes particularly important when designing column reactors or packed beds for continuous treatment systems.

Moreover, the use of foaming agent generally resulted in marginal increase of adsorption capacity compared to their unformed counterparts (e.g., MS and MS-F4) in the case of Pb(II) while opposite results were observed for BG. The differential response of Pb²⁺ and BG to foaming likely reflects the different nature of these adsorbates, i.e., Pb²⁺ as a small cation benefits from any increase inaccessibility, while BG as a larger organic molecule may require specific pore sizes for optimal adsorption. This suggests that pore engineering strategies need to be tailored to the target contaminant’s molecular characteristics.

Lastly, the MS-N8 sample showed the highest adsorption capacity. It can be inferred that the presence of more Na⁺ species in this sample potentially resulted in more cation exchange, thus a higher adsorption capacity was seen. This finding is particularly relevant for radioactive waste applications, as many radionuclides (such as cesium (Cs⁺) and strontium (Sr²⁺)) are cationic species that would be captured through similar ion-exchange mechanisms. The ability to enhance cation exchange capacity by simply adjusting NaOH molarity provides a straightforward approach to tailoring materials for specific radionuclide capture.

It should be noted that adsorption is a complex process. In geopolymeric or geopolymer-zeolite composites, adsorption can occur through physical capture, ion-exchange with charge balancing cations (mostly sodium), and/or chemical bonding with non-bridging Si– O − or Al– O − (El-Eswed et al., 2017; Luukkonen et al., 2019). FT-IR spectra of Pb(II) and BG-adsorbed samples were taken to understand the adsorption mechanism and are plotted along with the spectra of control samples in Figure 8. A change in transmittance was observed in the Si-O-T band around 1000 cm^-1 after adsorption of Pb(II) and BG. This confirms the involvement of SiO₄ tetrahedra in the adsorption process by ion-exchange and/or chemical bonding. Another major change was observed in AlO₄ ⁻ group band around 876 cm⁻¹. The Si– O − or Al– O − in geopolymers/zeolites are known to interact with the adsorbate ions (El-Eswed et al., 2017; Luukkonen et al., 2019). Similar behavior was confirmed through the FT-IR spectra of Pb(II) and BG-adsorbed samples. This was also confirmed by the higher adsorption capacity of MS-N8 compared to MS, owing to higher content of exchangeable Na + cations in this sample. The spectroscopic evidence of chemical interaction between the adsorbates and the aluminosilicate framework suggests that these materials offer not just physical capture but also strong chemical binding, which is essential for long-term immobilization. This dual mechanism, combining ion exchange with chemical bonding, provides multiple barriers against contaminant release, enhancing the safety and reliability of waste forms.

A comparison of the adsorption capacities observed in this work is made with the values reported in the literature in Table 5. Only the studies reporting bulk-type geopolymeric adsorbents (i.e., solid blocks, not pulverized/grinded) are covered. It is evident that the adsorption capacity of studied adsorbents was much higher than the values reported in the literature. The superior performance achieved in this study validates the effectiveness of the synthesis approach and parameter optimization undertaken. This advancement is attributed to the combination of optimized NaOH molarity, controlled bead size, and strategic balance between zeolite formation and pore structure preservation. The potential reason for such behavior could be their mesoporous characteristic and high molarity of NaOH solution used during the synthesis.