Introduction

Cadmium contamination in freshwater bodies is a severe global environmental and public health problem, given its high toxicity, persistence and bio-accumulation in biota. A precedent known human carcinogen according to world health organisations, exposure to cadmium has been associated with numerous life-threatening health problems ranging from cardiovascular disease, neurological disorders, and cancer to endocrine damage such as diabetes1,2,3. Unlike essential trace elements, cadmium serves no beneficial biological function. However, it is extensively introduced into waterbodies by anthropogenic activities, e.g., electroplating, battery production, textile dyeing, and agricultural runoff, bringing serious environmental pressure4. In aqueous environments, Cd²⁺ ions are very soluble and mobile and can be assimilated through aquatic food webs, bioaccumulate in organisms, and finally transfer to human beings5. Long-term exposure may result in severe effects, including immunodeficiency, kidney and liver damage, demineralisation of the bones, and interference with enzyme actions. Moreover, cadmium also interferes with plant–microbe interactions, reduce yield, and menace to ecological diversity. The industrial emissions of Cd2+ in the ecosystem suggest an imminent necessity for the development of sustainable and effective technologies to clear cadmium-polluted water1,2,4,5,6.

As cadmium in industrial wastes is an environmental and human health threat, the development of an efficient removal method is necessary. Cadmium has been removed from wastewater using various conventional methods, including membrane filtration, ion exchange, coagulation, chemical precipitation, solvent extraction, and adsorption7,8. Among these, adsorption is widely recognized for its affordability, simplicity, and high removal efficiency. Kaolin-derived geopolymer materials show strong potential as heavy metal adsorbents due to their favorable structural and physicochemical characteristics. Their aluminosilicate framework carries a negative charge owing to the substitution of silicon with aluminum, which is balanced by exchangeable cations like sodium or potassium. These cations play a crucial role in binding cadmium ions during the adsorption process7,9. Compared to conventional materials like activated carbon, geopolymers offer a viable and attractive alternative for wastewater purification. Geopolymer combine excellent mechanical and chemical durability with economic benefits and a relatively low-energy synthesis process, while powdered forms of adsorbents often pose challenges in terms of handling and post-treatment recovery10. Geopolymers mitigate many of these limitations by enabling easier fabrication and regeneration, all at a lower production cost11,12.

To address the challenges posed by powdered adsorbents, developing structured adsorbents has become a beneficial method. Among the different manufacturing methods, additive manufacturing (AM) provides exceptional control over geometry, porosity, and surface area, enabling the creation of specialized structures designed to improve mass transfer and lower pressure drop13. DIW, an additive manufacturing technique, facilitates the rapid creation of monolithic geopolymer adsorbents. These adsorbents exhibit graded porosity and enhanced reusability, rendering them a viable and eco-conscious choice for adaptable water treatment applications3,14,15. However, the application of additive manufacturing in structured adsorbent production is still in its infancy. Research is limited regarding its use in creating complex, high-performance geopolymer based monolithic structures for wastewater treatment.

Yadav et al. (2025) reported the use of fly ash-based geopolymers for the adsorption of Cd²⁺ from aqueous solutions. Batch adsorption experiments revealed a maximum adsorption capacity of 18.75 mg/g at an optimum pH of 4.5, using 0.1 g of adsorbent with an equilibrium contact time of 90 min16. In a related study, Mladenović et al. (2019) examined metakaolin-based geopolymers for Cd²⁺ removal and achieved a removal efficiency of 84.1% at pH 6 with an adsorbent dosage of 1 g/dm³ and a contact time of 60 min. These studies highlight the effectiveness of geopolymer-based materials as promising adsorbents for heavy metal remediation17. However, most reported works have employed geopolymers in powdered form, whereas the present study emphasizes the development and application of structured geopolymer adsorbents for Cd²⁺ removal. Velayos et al. (2023) developed 3D-printed monoliths using a noncommercial filament composed of iron oxide (Fe₃O₄) and polylactic acid (PLA) at 15 wt% Fe₃O₄. These monoliths were employed for the continuous degradation of ofloxacin in a down-up flow reactor, achieving 55–82% degradation under optimized conditions18. Bauza et al. (2024) developed a cost-effective 3D-printed adsorbent for paraben removal by integrating MOF-derived porous carbon using DIW. A ZIF-8 MOF was in situ grown on a 3D-printed device and carbonized to form a zinc-free, micro-mesoporous carbon coating. These c-ZIF-8@3D structures were capable of more than 90% of removal of ethylparaben in one hour response time and showed an extraction efficiency of 83–92% after 10 cycles when tested in continuous-flow experiments. This device proved to be easily regenerated and very efficient in removing parabens from realistic water samples and was suggested as a promising technology for sustainable water treatment10. Instead, Gonçalves et al. (2023) explored a new methodology using the bauxite residue (red mud) for the manufacture of DIW porous structures based on the adsorption of methylene blue (MB). The use of a RM/MK 50:50 wt % mixture resulted in composites with compressive strength of 10.70 MPa, open porosity of 62.4% and surface area of 55 m²/g. Considering the adsorption testing was performed under ideal conditions of high concentration methylene blue solution and longer contact time, the materials exhibited a MB adsorption capacity of 19.96 mg/g, which is significantly high for a bulk form geopolymer adsorbent. These types of lattice structures also withstood 10 regeneration cycles, exhibited favourable behaviors in continuous flow systems, thus making them viable candidates for wastewater treatment12. In a similar study, Novais et al. (2016) prepared porous geopolymer monoliths from fly ash, which exhibited 6.34 mg/g of lead adsorption capacity19. However, a conventional moulding technique was used in this study for preparing geopolymer monoliths. Lukkonen et al. (2020) investigated highly porous 3D-printed geopolymer filters for water treatment, and they exhibited a total porosity of 28 vol% and a compressive strength of 16 MPa, maintaining mesoporosity, water permeability, and structural integrity. Metal impregnation with Ag and Cu influenced their oxidation state and stability, with minimal leaching (pH 3–7), ensuring safe concentrations within drinking water standards20.

To the best of our current understanding, the application of 3D printed geopolymer monoliths for cadmium removal has not been extensively investigated. This study seeks to explore their potential as robust, efficient, and scalable adsorbents for advanced wastewater treatment applications. The geopolymer ink formulation, necessary rheology, and printing parameters are thoroughly discussed in our prior research21. To make monoliths more porous, hydrogen peroxide (H2O2) was incorporated as a foaming agent, enabling precise control over internal porosity, maximising the effective surface area while maintaining both printability and structural integrity. A key objective was to innovate sorption bed design by integrating patterned lattice packing, which enhances flow control, minimises pressure drop, improves mass transfer, and increases the material’s resistance to attrition. Furthermore, the same ink formulation was used to produce beads for a systematic evaluation and comparison to determine if monolithic structures can replace or supplement bead-based systems. Continuous-flow column experiments validated the potential for industrial-scale applications. In eight consecutive adsorption-desorption cycles, the 3D-printed structures’ reusability was carefully assessed; in water-based systems, 3D printed geopolymer monoliths continuously demonstrated strong structural integrity and high cadmium removal efficiency.

Materials and methods

Geopolymer matrix formulation

Zillion Sawa Minerals in Chittorgarh, India, provided the Kaolin, which was thermally treated to synthesise metakaolin (MK) by heating it to 750 °C for three hours in a muffle furnace. This calcination process triggered dehydroxylation, thereby enhancing the material’s pozzolanic activity. For the synthesis of the geopolymer binder, a liquid-to-solid (L/S) ratio of 1.25 on a weight basis was adopted, as it was found optimal for achieving high compressive strength, minimising water uptake, and improved resistance to freeze-thaw cycles, parameters crucial for mechanical performance22. Analytical-grade sodium hydroxide (98.5% purity) and commercial sodium silicate, which contained 29.69 weight per cent SiO₂, 15.31 weight per cent Na₂O, and 55 weight per cent H₂O, were used to prepare the alkaline activating solution. To achieve uniformity, these ingredients were combined in a 2:1 mass ratio (NaOH to sodium silicate) and stirred for 15 min at room temperature. The solution was prepared at least a day ahead of time and kept in the refrigerator at 4°C until it was needed in order to optimise the activation potential. The metakaolin was then thoroughly mixed with this activator solution, maintaining the selected L/S ratio of 1.25, resulting in the formation of a cohesive and structured geopolymer matrix. The synthesised geopolymer was subsequently ground using a ball mill and passed through a 355 μm sieve to ensure uniform particle size distribution. The sieved powders were then transferred into airtight containers and stored in a desiccator until further use to prevent moisture uptake and preserve reactivity.

Geopolymer ink formulation

The geopolymer ink for 3D printing was prepared by first synthesising the geopolymer matrix, into which Xanthan gum was incorporated to tailor the rheological behaviour. To enhance structural integrity and compressive strength, 10 wt% kaolin was subsequently introduced as a binder. Finally, 14 wt% of a 1% H₂O₂ solution was added as a foaming agent to induce porosity, and the mixture was thoroughly kneaded to obtain a homogeneous paste suitable for extrusion. 1 wt% H₂O₂ solution used in this study was prepared by diluting commercially available 35 wt% laboratory-grade H₂O₂ with deionised water. This process generated gas within the geopolymer, leading to the formation of a macroporous network upon solidification. As demonstrated in our prior research, this composition represents the optimised formulation of the geopolymer ink21. For the DIW technique, the resulting ink was subjected to rheological testing and printability assessments.

Monolith fabrication via DIW

The geopolymer ink was loaded into a 60 mL Luer lock syringe, which was subsequently connected to a 3D printer that employed a syringe-based extrusion system. Lattice structures were produced under ambient environmental and the ink was dispensed at a rate of 10 mm/s through a blunt-tip nozzle with a diameter of 0.840 mm. The resulting cylindrical lattices were composed of parallel geopolymer filaments arranged in a 30% rectilinear infill pattern. The thickness of each layer was set to 70% of the nozzle’s diameter, and the dispensing rate was adjusted in accordance with the ink’s retention time. The printed monoliths measured 25 mm in diameter and 5 mm in height. In order to perform dynamic adsorption experiments and optimise printing parameters, multiple prints were produced. The printed monoliths were then used for adsorption tests after the samples had been cured for 24 h at 60 °C in a hot air oven.

Beads fabrication via the line injection method

Direct extrusion was used to fabricate the geopolymer beads. In order to achieve uniform deposition, the geopolymer paste was manually extruded from a syringe while maintaining a constant force. The extruded filaments were then manually shaped into beads using a calibrated scale to ensure that equal-length pellets were produced, which were subsequently shaped manually into beads. Before being used in adsorption tests, these beads were first aged for 20 days and then cured for 24 h at 60 °C in a hot air oven. Additionally, some of the beads were ground into powder for batch adsorption tests to assess the material’s performance.

Characterization

X-ray powder diffraction (XRD) was used to characterise the mineral phases found in kaolin, metakaolin (MK), and the synthesised geopolymer (GP), carried out on a Tescan Vega 3 system. The PANalytical X’Pert High-Score Plus software was used for phase analysis and identification. GP powder, bead, and monolith samples were subjected to Fourier-transform infrared spectroscopy (FTIR) in attenuated total reflectance (ATR) mode before and following cadmium ion (Cd²⁺) adsorption using a Bruker Tensor 27 spectrometer to examine chemical bonding and structural alterations. Spectral data were collected over the range of 4000 to 400 cm⁻¹ with a resolution of 4 cm⁻¹. The specific surface area (SSA) was measured via nitrogen adsorption using the multi-point Brunauer–Emmett–Teller (BET) technique on a Nova Touch LX2 instrument. Surface morphology and elemental distribution of the geopolymer samples were observed using a Zeiss Neon 40 field-emission scanning electron microscope (FESEM), which was equipped with an energy-dispersive spectroscopy (EDS) system for elemental mapping. The concentrations of Cd2+ ions before and after the adsorption experiments were measured using atomic absorption spectrophotometry (AAS), ANALYST 100 Spectrophotometer, Perkin-Elmer, Norwalk CT/USA.

Batch adsorption experiment

A 1000 ppm cadmium stock solution was formulated by dissolving 2.031 g of cadmium chloride in double-distilled water. Subsequent serial dilutions were carried out to obtain the desired working concentrations, with pH adjustments made using 0.01 M HCl and 0.01 M NaOH.

The geopolymer ink was finely ground and subjected to batch adsorption experiments to assess its cadmium removal performance. The study focused on four primary operational parameters: pH, contact time, initial cadmium concentration, and the amount of adsorbent used. The effect of pH on adsorption was explored over a range of 1 to 8, with all tests conducted under continuous stirring for 24 h to ensure equilibrium conditions. Initially, different dosages of the powdered adsorbent, between 0.5 mg/L and 2 mg/L, were tested to determine the optimal amount. Contact time was varied from 5 to 100 min to analyse the interaction duration between the adsorbent and cadmium ions. Cadmium adsorption kinetics were analysed using a batch equilibration system, with contact times spanning 5 to 90 min. All the experimental study was carried out in triplicate and the process kinetics were evaluated by monitoring the change in the amount adsorbed over time.

The amount of cadmium adsorbed at equilibrium, denoted as qe, (mg g−1), was calculated using the following expression

$${q}_{e}=\frac{\left({C}_{i}-{C}_{e}\right)V}{m} $$
(1)

In this equation, \({C}_{i}and{C}_{e},\) mg g−1 refer to the initial and equilibrium concentrations of cadmium ions, respectively. V(L) is the volume of the solution, and m (g) is the adsorbent mass used.

To analyse the adsorption kinetics, the experimental data were fitted using two commonly applied models: the pseudo-first-order and pseudo-second-order equation8. The pseudo-first-order kinetic model is represented by the following Eq. (2):

$$\frac{{dq}_{t}}{dt}={k}_{1}\left({q}_{e}-{q}_{t}\right)$$
(2)

In this context, \({k}_{1}\left({min}^{-1}\right),\)represents the rate constant for the pseudo-first-order model, \({q}_{t}\) \((\text{m}\text{g}/\text{g})\)is the amount of cadmium adsorbed at a given time, and ​ \({q}_{e}(\text{m}\text{g}/\text{g})\) denotes the equilibrium adsorption capacity.

Alternatively, the pseudo-second-order kinetic model, when applied using non-linear regression, is expressed by the following equation.

$$\frac{{dq}_{t}}{dt}={k}_{2}{({q}_{e}-{q}_{t})}^{2}$$
(3)

where \({k}_{2}\) (\(g{mg}^{-1}{min}^{-1}\) ), represents the rate constant for the pseudo-second-order model.

The effect of varying initial cadmium concentrations was examined while keeping the adsorbent dosage and pH constant. To characterize the equilibrium relationship between cadmium ion adsorption and its concentration in solution, both the Langmuir and Freundlich isotherm models were applied23. Similar to previous analyses, the Langmuir isotherm model was fitted using the following Eq. (4):

$${q}_{e}=\frac{{q}_{m}{K}_{L}{C}_{e}}{1+{K}_{L}{C}_{e}}$$
(4)

In the Langmuir isotherm model \({q}_{e}\left(mg{g}^{-1}\right)\) represents the equilibrium adsorption capacity, \({C}_{e}\) (\(mg{L}^{-1}\)) is the metal ion concentration at equilibrium, \({K}_{L}\)(\(L{mg}^{-1})\) denotes the Langmuir constant reflecting the affinity between adsorbent and adsorbate, and \({q}_{m}\) (\(mg{g}^{-1}\)) the theoretical maximum adsorption capacity.

The Freundlich isotherm model, on the other hand, was applied using the following expression (5):

$${q}_{e}={K}_{F}{\left({C}_{e}\right)}^{1/n}$$
(5)

where \({K}_{F}\) is the Freundlich constant (\(mg{g}^{-1}\)) indicating the adsorption capacity, and n dimensionless constant representing the adsorption intensity.

Column adsorption experiment

As illustrated in Fig. 1, a column test was conducted using beads and monoliths inside a glass column reactor to simulate an industrial adsorption process and assess the viability of continuous adsorption. The experiments utilized a glass column with an internal diameter of 26 mm. Monolithic adsorbents, each measuring 25 mm in diameter and 5 mm in height, were arranged in a stacked configuration within the column reactor. The total bed height was adjusted between 50 mm and 150 mm. The bead form of the adsorbent was tested using the same experimental conditions, including a pH of 5 and an initial cadmium concentration of 100 ppm. A centrifugal pump was used to regulate the cadmium solution’s recirculation rate from 5 mL/min to 15 mL/min, with the flow moving in an upward direction. The cadmium solution’s recirculation rate was modulated from 5 mL/min to 15 mL/min via a centrifugal pump, with the flow oriented in an upward direction. Experimental data, gathered from the column under varied operating conditions, were analysed and compared using well-established and easily applicable mathematical models, namely the Thomas and Bohart–Adams24,25.

Fig. 1
figure 1

Experimental setup for column study.

Full size image

The Thomas model relies on the assumptions of plug flow dynamics and Langmuir adsorption-desorption equilibrium within the adsorption bed. The model’s linearised form is expressed as follows:

$$\text{ln}\left(\frac{Co}{Ct}-1\right)=\frac{{k}_{th}{q}_{0}m}{Q}-\frac{{k}_{th}{C}_{o}t}{Q}$$
(6)

where \({k}_{th}\) (mL min–1 mg–1) represents the Thomas rate constant; \({q}_{0}\) (mg g–1) denotes the adsorption capacity; \(m\) (g) signifies the mass of adsorbent within the column, and Q (mL min–1) indicates the feed flow rate.

This model facilitates the determination of essential parameters, such as bed height, adsorbent mass, and column dimensions, necessary for achieving desired removal efficiencies at an expanded operational scale26,27,28. When assessing breakthrough performance in fixed-bed adsorption columns, the Adams-Bohart model is a useful tool, especially when it comes to the first part of the breakthrough curve. It is mainly used to approximate bed depth, service duration, and the kinetics of initial sorption. The linearised equation of the Adam Bohart model is represented as follows.

$$\text{ln}\left(\frac{Ct}{Co}\right)={k}_{AB}{C}_{o}t-\frac{{k}_{AB}{N}_{o}Z}{U}$$
(7)

where \({k}_{AB}\left({s}^{-1}\right)\)represents the Adams rate constant; \({N}_{0}\) \(\left(mg{g}^{-1}\right)\) denotes the adsorption capacity of the bed; \(U\left(cm{s}^{-1}\right)\) signifies the superficial velocity through the bed, and \(Z\left(cm\right)\) the length of the adsorption bed.

Regeneration and reusability

Ethylene diamine tetra acetic acid (EDTA) stands out as a highly effective chelating agent due to its exceptional ability to form stable complexes with heavy metals, a property that is crucial in environmental remediation and analytical chemistry23. The selection of Na2EDTA for extracting adsorbed Cd2+ from the adsorbent is based on its strong complex-forming capabilities, facilitating the detachment of cadmium ions from the adsorbent material24. However, challenges can arise during the desorption process, particularly.

When dealing with powdered adsorbents, as the physical separation of the adsorbent from the extraction solution can lead to material loss, hindering accurate evaluation of desorption efficiency25. Thus, powdered adsorbent was not considered for the column study. The adsorbent, when shaped into beads and monoliths, was regenerated and reused for up to 3 and 10 cycles, respectively. After three cycles, cracks began to appear in the beads.

Results and discussion

Structural analysis

Figure 2 illustrates the X-ray diffraction patterns of kaolin, metakaolin, geopolymer, and geopolymer ink. X-ray diffraction analysis reveals that kaolin is mainly composed of kaolinite and quartz. The removal of water molecules from the kaolinite structure during heat treatment, which produces metakaolinite, is the reason why kaolinite peaks vanish after calcination29. The crystalline phases dissolved in the alkaline solution during the activation process, causing a geopolymerization reaction to form an aluminosilicate phase on the metakaolin surface30. A broad peak, characteristic of amorphous materials, is observed in all patterns within the 2θ range of 20–40°, suggesting the creation of amorphous aluminosilicates. Albite and Quartz were identified as the primary crystalline phases in the geopolymer and geopolymer ink31.The XRD analysis did not reveal any significant changes in the mineralogical phases of the geopolymer ink prepared with H2O2 and other additive contents, which aligns with previous findings reported by Faris et al. (2022).

Fig. 2
figure 2

XRD patterns for Kaolin, Metakaolin, Geopolymer, and Geopolymer Ink.

Full size image

Scanning electron microscopy images in Fig. 3a–e depict the morphology of a geopolymer bead and a 3D-printed monolith. The bead exhibits a spherical shape with a smooth surface, measuring 2.5 mm ± 0.2 mm. The surface of both the bead and the monolith shows slight porosity, while the internal core displays a higher degree of porosity as depicted in Fig. 3a, b and c, d, respectively. Figure 3e presents an overview of the printed geopolymer monolith, and Fig. 3f provides a detailed view of its cross-section. The structure comprises filaments with consistent diameter and spacing, indicating the suitability of the slurry for additive manufacturing. The monolith’s dimensions were measured as 25 mm ± 0.2 mm and 0.5 mm, which closely aligns with the intended design.

Fig. 3
figure 3

SEM images of the bead and monolith: a geopolymer bead, b internal porous surface of the bead, c 3D printed geopolymer monolith, d internal porous surface of the monolith, e overview of the 3d printed monolith, f cross section of the monolith.

Full size image

Table 1 presents the total and open porosity results, along with the specific surface area, for both the geopolymer and geopolymer ink, considering the H2O2 content. According to BET analysis, the geopolymer ink containing H2O2 has a higher specific surface area (SSA) of 16.73 m2/g than GP without H2O2. These outcomes are in line with earlier research on geopolymer synthesis that employed H2O2 as a foaming agent9. Although some studies have reported higher SSA values for sodium-based geopolymers, this is typically due to an increased amount of NaOH used to prolong the geopolymerization process32,33. Given that the primary focus of our study was to formulate a geopolymer ink suitable for 3D printing, the NaOH concentration was optimised accordingly. Furthermore, the inclusion of additives and binders like xanthan gum and kaolin in our sample may impede pore formation.

Table 1 BET analysis of geopolymer ink with H2O2 and without H2O221.
Full size table

The infrared spectroscopy used to determine the functional groups on the adsorbent surface and clarify the adsorption mechanism is shown in Fig. 4. Between 4000 and 400 cm⁻¹, spectra of geopolymer, regenerated geopolymer, and cadmium-loaded geopolymer were acquired. The –OH groups and adsorbed water on the geopolymer surface are responsible for the absorption bands found in the IR spectra at about 3465 cm⁻¹, both before and after cadmium adsorption1. Additionally, the H–O–H bending vibration absorption band at 1627 cm⁻¹ was consistently observed across all three samples. The infrared spectrum of the original geopolymer showed a peak at 1480 cm⁻¹, suggesting the presence of asymmetric stretching vibrations associated with carbonate groups. This peak’s disappearance after cadmium adsorption indicates that the carbonate groups either underwent structural modification through ion exchange or surface complexation, or they interacted with Cd²⁺ to form cadmium carbonate complexes. This observation implies an effective interaction between the geopolymer matrix and cadmium ions. This finding suggests that the cadmium ions and the geopolymer matrix interact effectively. In both the fresh and regenerated geopolymer samples, a distinctive peak was observed at 2069 cm⁻¹. This peak showed a shift to 2078 cm⁻¹ after cadmium adsorption, indicating a change in the local bonding environment due to interaction with Cd²⁺ ions34. The coordination of cadmium with functional groups, such as Si–O or Al–O sites, within the geopolymer matrix may be the cause of the blue shift, which indicates a stronger or more rigid bond formation. The asymmetric stretching vibration of Si–O–T bonds is responsible for the characteristic peak that was consistently found at 1020 cm⁻¹ in all geopolymer samples with slight shifts from 1020 cm⁻¹ (fresh geopolymer) to 1024 cm⁻¹ (regenerated) and 1026 cm⁻¹ (Cd²⁺-loaded). These shifts reflect minor modifications in the aluminosilicate framework due to Cd²⁺ interaction, while the persistence of the peak confirms that the fundamental geopolymer structure remains intact9,12.

Fig. 4
figure 4

IR spectra (a) geopolymer ), (b) regenerated geopolymer , and (c) Cd2+ loaded geopolymer.

Full size image

To determine the elemental composition of the geopolymer, Energy Dispersive X-ray Spectroscopy was utilised across three different states: after preparation, after cadmium adsorption, and after regeneration via eight adsorption-desorption cycles. Oxygen, silicon, aluminium, and sodium were the most common elements found in the initial state, which is consistent with the usual aluminosilicate structure of geopolymers. Alkali and transition metal traces were also found, most likely from the precursors used in the synthesis16. The formation of cadmium peaks coincided with a noticeable drop in silicon and aluminium content following Cd2+ adsorption. Partial leaching or the occlusion of the geopolymer surface by Cd2+ ions during the adsorption process could be the cause of this decrease in Si and Al concentrations34. Increased carbon content and calcium detection point to possible ion exchange or surface complexation processes in which Cd2+ interacts with residual carbonate species or hydroxyl groups on the geopolymer surface7.

The elemental composition partially returned to the original geopolymer after regeneration. In particular, concentrations of silicon and aluminium rose, suggesting some structural recovery in the geopolymer network. Even after several regeneration cycles, cadmium was still detectable, indicating that some Cd2+ ions were still firmly attached to active sites. The partial restoration of the original geopolymer matrix is further supported by the reemergence of sodium and a slight increase in iron7. The consistent oxygen content supports preserving hydroxyl and silicate groups, which are crucial for maintaining adsorption activity16. The findings indicate that the geopolymer maintains a high degree of structural integrity and adsorption capacity despite multiple applications, underscoring its practical potential for heavy metal removal. A strong binding affinity and the potential for selective adsorption of Cd2+ ions are further suggested by the continuous detection of cadmium after regeneration as shown in Fig. 5.

Fig. 5
figure 5

FESEM micrographs and corresponding EDS spectra a fresh adsorbent, b cadmium-loaded adsorbent, and c regenerated adsorbent.

Full size image

Batch adsorption study

Influence of initial pH on Cd²⁺ uptake

The intricate interplay of pH on the adsorption dynamics of Cd²⁺ onto geopolymer surfaces reveals a multifaceted phenomenon governed by electrostatic interactions, chemical complexation, ion exchange mechanisms, and the consequential influence of cadmium hydrolysis and precipitation processes. As shown in Fig. 6, at the lower end of the pH spectrum, specifically within the range of 1 to 3, the presence of an elevated concentration of hydronium ions (H₃O⁺) engenders a competitive environment, wherein these ions vie with Cd²⁺ ions for occupancy of the available active sites on the geopolymer matrix3,10. In addition, protonation of surface hydroxyl groups (≡ Si–OH, ≡Al–OH) imparts an overall positive charge to the geopolymer surface, which causes electrostatic repulsion against Cd²⁺ ions, thereby lowering adsorption efficiency35. As the pH ascends, typically beyond a value of 3, the geopolymer surface undergoes deprotonation, resulting in an augmented negatively charged sites on the aluminosilicate network, particularly associated with Si–O–Al and Si–O–Si linkages. These negatively charged surface sites exhibit stronger electrostatic attraction toward Cd²⁺ ions, enhancing adsorption. This attraction is most effective within a pH range of 5 to 6, where the adsorption capacity is at its highest, indicating ideal electrostatic conditions for Cd²⁺ sequestration3.

Fig. 6
figure 6

Influence of starting solution pH on Cd²⁺ adsorption performance of geopolymers, along with corresponding final pH values post-adsorption.

Full size image

However, it is crucial to note that electrostatic forces do not solely dictate the adsorption process; chemisorption plays a significant role, mediated by inner-sphere complexation reactions between Cd²⁺ ions and the geopolymer surface. In this chemisorption process, covalent bonds are formed between the Cd²⁺ ions and the geopolymer’s surface functional groups, leading to a stronger and permanent adsorption. Furthermore, the contribution of ion exchange mechanisms cannot be overlooked, where Cd²⁺ ions displace Na⁺ native cations, present on the geopolymer surface12. The adsorption capacity and selectivity of the geopolymer for Cd²⁺ ions are enhanced by the ion exchange process. The peak in Cd²⁺ adsorption observed at pH 5.0 is the result of the synergistic interaction of electrostatic attraction, surface complexation, and ion exchange mechanisms. Beyond this optimal pH, a slight decrease in adsorption capacity is seen, potentially attributable to the onset of Cd²⁺ hydrolysis and the resulting decrease in the quantity of free Cd2+ ions3. The formation of cadmium hydroxide, a solid precipitate, at elevated pH levels may further reduce Cd²⁺ concentration in the aqueous phase35. The concentration of free Cd2+ ions available for adsorption is reduced by this precipitation phenomenon. By covering the geopolymer surface and obstructing access to active sites, it may also prevent adsorption. Since it encourages significant Cd²⁺ adsorption while reducing the effect of other concurrent processes, such as cadmium hydrolysis and precipitation, which can also help lower the pollutant concentration, a pH of 5.0 was found to be the most effective for this investigation. In addition, a leaching stability test was performed, in which the 3D-printed monolithic geopolymer specimen was immersed in deionised water for 24 h. The pH of the solution was measured before and after immersion, and no significant change in pH was observed, confirming that the monolithic structure did not release appreciable alkaline species into the medium at pH 5. The monolithic geopolymer exhibits high chemical stability due to its compact structure, silica-rich passivation layer, and dense aluminosilicate network, which collectively restrict Na⁺ diffusion and alkali release36. Furthermore, the structural integrity of the geopolymer was validated by immersing the monoliths in water for 10 days, during which no deformation or shrinkage was observed, as also reported in our previously published work21. Hence, under the experimental conditions (pH = 5), the geopolymer monolith remains chemically and structurally stable, ensuring that Cd²⁺ adsorption is not influenced by alkaline leaching.

Influence of adsorption time and sorbent quantity

Over the course of five to ninety minutes, the impact of the starting Cd+ 2 concentration on the adsorption capacity was assessed within the range of 50 to 400 mg/L. The process proceeded swiftly, as seen in Fig. 6a; for a starting cadmium concentration of 100 mg/L, the removal percentage reached a maximum of 45% after 20 min. After 60 min, the adsorption capacity stabilized, and all cadmium concentrations showed comparable patterns. As a result, 100 mg/L was selected as the starting concentration as the ideal setting for further research into the effects of different experimental parameters.

Fig. 7
figure 7

a Influence of initial cadmium concentrations on adsorbate removal (m adsorbent = 0.5 gm/L, pH 5), and b influence of adsorbent dosage on adsorbate (conc.= 100 mg/L, pH 5).

Full size image

The effect of geopolymer amount on Cd2+ adsorption was studied between 0.5 and 2.0 g/L. High amount of geopolymer improved adsorption, but performance decreased at 2.0 g/L as depicted in Fig. 7b. This is likely because more geopolymer provides more surface area and binding sites for cadmium ions. A dose of 1.5 g/L achieved 79% removal and an adsorption capacity of 52.5 mg/g, as depicted in Fig. 8. Additional geopolymer beyond this amount did not significantly increase adsorption.

Fig. 8
figure 8

Influence of adsorbent dosage on adsorption capacity (Conc.= 100 mg/L, pH 5).

Full size image

Adsorption isotherms

To determine the maximum adsorption capacity and clarify the adsorption mechanism of the powdered adsorbent, adsorption isotherms were developed. As shown in Fig. 9, these isotherms show a graphical relationship between the equilibrium concentration of Cd2+ ions in the supernatant following the adsorption experiment and the equilibrium adsorption capacities of the solid phase. With a strong correlation coefficient of R2 = 0.989, the isotherm data closely matched the Langmuir equation, indicating that the sorption data fit the Langmuir isotherm model better. Table 2 provides specifics on the isotherm constants and fitting results.

Fig. 9
figure 9

Nonlinear fit of a Langmuir, and b Freundlich adsorption isotherm for removal of Cd2+.

Full size image
Table 2 Langmuir and Freundlich isotherm model constants for the adsorption of Cd2+ ions
Full size table

To assess the effectiveness of the adsorption process, the dimensionless separation factor was determined using the formula RL = 1/ (1 + bC0), where b represents the Langmuir constant and C0 signifies the initial concentration of metal ions. An RL value between 0 and 1 indicates favourable adsorption, whereas a value greater than 1 suggests unfavourable adsorption. The computed RL value of 0.21 for Cd2+ ions falls within the desirable range of 0–1, implying a propensity for the adsorption of metal ions by the GP adsorbent37,38.

Adsorption kinetics

As seen in Fig. 10, the correlation coefficient values are close to unity, indicating that the adsorption kinetics of cadmium ions onto the synthesised metakaolin-based geopolymers are accurately described by the pseudo-second-order kinetic model, which is typical of a mechanism dominated by chemisorption39. This finding is consistent with the field’s consensus, which states that metakaolin-based geopolymers are commonly reported to have a sizable capacity for adsorption of divalent metal ions through a pseudo-second-order adsorption process40. The kinetic behaviour demonstrates the geopolymer surface’s heterogeneity by showcasing a range of active sites and energy levels that make it easier to extract metal ions from aqueous solutions.

Fig. 10
figure 10

Nonlinear fit of a pseudo-second-order kinetic model, and b pseudo-second-order kinetic model.

Full size image

Column study

The effectiveness of a fixed-bed adsorption column in removing cadmium ions from polluted aqueous solutions was assessed through breakthrough curve analysis under various operational conditions. The key parameters such as the temporal changes in effluent concentration, time to breakthrough, and time to bed exhaustion are essential for designing and operating adsorption columns41,42,43,44,45. To find out how operating parameters impact the adsorption of cadmium ions onto both geopolymer monolithic and bead beds, continuous fixed-bed column experiments were conducted. Under carefully monitored experimental conditions, several breakthrough curves were measured. To conduct a systematic investigation, the feed flow rate was varied at a fixed bed height while the feed concentration remained constant throughout all experiments. Further tests were carried out with constant flow rates but at different bed heights. Table 3 provides specifics about the experimental setup. Plotting the normalised concentration against time showed a typical sigmoidal shape for all curves, illustrating the various adsorption phases as the heavy metal solution passes through the column. The points at which the effluent concentration achieved 10% and 90% of the influent concentration, respectively, were designated as the breakthrough point and saturation points in this investigation.

Influence of flow rate and bed height

The adsorption of cadmium in geopolymer-packed columns was studied by changing the flow rate and bed height, while keeping the initial concentration at 100 mg/L. Higher flow rates led to shorter breakthrough times and lower adsorption capacity and removal efficiency for both bead and monolithic beds, as illustrated in Fig. 11a–f. This can be attributed to a reduced interaction duration between the Cd²⁺ ions and the adsorbent surface, thereby impeding adsorption equilibrium. Furthermore, the mass transfer zone narrowed, suggesting a quicker saturation of the active layer because of accelerated solute transport. Conversely, augmenting the bed height from 0.5 to 1.5 cm prolonged breakthrough and saturation times but expanded the MTZ and diminished adsorption capacity. This decline in specific uptake likely stems from incomplete utilisation of deeper bed sections and axial dispersion effects, which reduce the adsorption efficiency per unit mass. Comparative analysis indicates that monoliths displayed a consistently shorter MTZ, implying enhanced mass transfer characteristics and uniform flow distribution. Monoliths had a slightly lower adsorption capacity compared to bead-packed beds. However, the removal efficiency was about the same, with only a small variation of ± 2%. Table 3 depicts the column parameters for both the bead and monolithic beds.

Fig. 11
figure 11

Breakthrough curves at different flow rates a bed height 0.5 cm beads, b bed height 0.5 cm monoliths, c bed height 1.0 cm beads, d bed height 1.0 cm monoliths, e bed height 1.5 cm beads, and f bed height 1.5 cm monoliths

Full size image
Table 3 Column parameters and removal percentage for both beads and monoliths.
Full size table

The breakthrough curves illustrated in Fig. 12a, b demonstrate the impact of bed height on the dynamic adsorption of cadmium, employing both geopolymer beads and monolithic columns under consistent parameters. In both configurations, increasing bed height delayed breakthrough time, consistent with increased interaction time between the adsorbate and adsorbent. The increased availability of active sites and extended residence time in the longer bed enhance overall adsorption performance.

Fig. 12
figure 12

Breakthrough curves at different bed heights and constant flow rate of 5 mL/min a bead, and b monolith

Full size image

Monolithic columns exhibit steeper breakthrough curves compared to bead-packed beds, indicating a sharper mass transfer zone and more efficient, uniform mass transfer with minimal axial dispersion. The broader curves in bead beds suggest greater dispersion, potential channelling, and delayed utilisation of deeper bed sections. Despite a slightly lower breakthrough time in monoliths at each bed height, the sharper profile indicates superior control over flow dynamics, minimising dead zones and ensuring better predictability. Thus, the structured architecture of monoliths proves enhanced flow distribution and accelerates the adsorption front progression, advantageous for scalable and reproducible applications.

Adsorption modelling

The Thomas model parameters, derived from linearised Eq. (6), elucidate column performance. The adsorption capacity, q0, predicted by the Thomas model, diminishes with increasing flow rate in both bead and monolith configurations. Beads exhibit marginally superior q0 values compared to monoliths across all tested flow rates. The high regression coefficient R2 values, approaching unity, affirm the Thomas model’s robust fit. Table 4 summarises the calculated Thomas parameters for beads and monoliths. Elevated flow rates curtail the residence time of cadmium ions, thereby impeding adsorbate-adsorbent contact. This constrained contact time attenuates mass transfer, resulting in diminished adsorption site utilisation. Consequently, increased flow rates correlate with reduced adsorption capacity, a trend substantiated by both experimental findings and model predictions. Beads manifest slightly enhanced adsorption capacity relative to monoliths, potentially attributable to their augmented surface area and porous architecture, which facilitates cadmium assimilation. Monoliths, however, demonstrate comparable efficacy coupled with diminished flow distribution variability and a shortened mass transfer zone, indicative of more uniform kinetics as shown in Fig. 13a–d. The parameter kTh escalates with flow rate, consistent with the anticipation that accelerated flow rates expedite breakthrough, thereby augmenting the apparent adsorption rate. This observation aligns with prior investigations indicating that the Thomas model accurately portrays adsorption phenomena in scenarios where the rate-limiting steps are not external and internal diffusion46,47.

Fig. 13
figure 13

Linear plots based on the Thomas model for beads and monoliths: a, b at varying flow rates, and c, d at different bed heights. Experimental data points are shown as dots, with lines representing the model predictions.

Full size image
Table 4 Fitted Thomas model parameters for Cd²⁺ ion adsorption.
Full size table

The Adams-Bohart model parameters, derived from linearised Eq. (7), elucidate column performance. Table 5 summarizes the Adams-Bohart model parameters (kAB and N0) for beads and monolithic beds at varying flow rates and bed heights under constant Cd²⁺ concentration. With increasing flow rate, both systems exhibit higher N0 and kAB, reflecting improved mass transfer. At low flow (5 ml/min), monoliths slightly outperform beads due to reduced pressure drop and better flow distribution. However, at higher flow rates, both systems show comparable kinetics, with beads demonstrating consistently higher N0, indicating superior adsorption capacity as shown in Fig. 14a–d. An increase in bed height reduces N0, suggesting enhanced contact time and delayed breakthrough. Nearly constant kAB across bed heights supports the model’s applicability to early-stage adsorption. Beads offer greater adsorption efficiency due to higher surface area, while monoliths provide operational advantages like uniform flow and thus the lower energy demand. Overall, beads are preferable for maximum removal efficiency, whereas monoliths are better suited for scalable, continuous applications. However, the model’s accuracy is limited to a specific range, and R2 values suggest it may not always be applicable23.

Fig. 14
figure 14

Linear plots based on the Adams–Bohart model for bead and monolith: a, b at varying flow rate, and c, d at different bed heights. Experimental data points are shown as dots, with lines representing the model predictions.

Full size image
Table 5 Adams-Bohart model parameters for adsorption of Cd2+ ions.
Full size table

Reusability of the bead and monolith

In addition to a substantial adsorption capacity, effective adsorbents intended for the elimination of Cd2+ should demonstrate exceptional reusability, a crucial factor in diminishing water treatment expenses. Within this investigation, both the beads and monoliths underwent regeneration and were subsequently reused across 3 and 8 cycles, respectively, maintaining a Cd2+ ion concentration of 100 mg/L, a flow rate of 5 mL/min, and a consistent bed height of 0.5 cm. Figure 15 illustrates the reusability performance of the geopolymer beads and monoliths for Cd2+ removal, as assessed through column experiments. Following the initial run, the beads achieved a maximum adsorption capacity of 37.5 mg/g, while the monoliths exhibited a capacity of 35.9 mg/g. After the 3rd cycle, cracks were observed in the beads, potentially due to uneven densification. However, the structural integrity of the monoliths remained unchanged even after the eighth cycle, possibly due to the engineered porosity, which maintained their structural stability as a single unit. The removal efficiency remained consistent for the first 6 cycles. A slight decrease to 65% from 71.9% of cadmium removal was observed after six cycles, probably because Cd2+ was not thoroughly washed. This suggests that a fraction of Cd²⁺ ions remained strongly bound to active sites, even after multiple regeneration cycles. This finding aligns with the EDS spectra of the regenerated monolith after eight regenerative cycles, as shown in Fig. 5. Based on the regeneration results, monoliths are proven to be the most efficient form of the adsorbenst.

Fig. 15
figure 15

Adsorption-regeneration cycles of Cd2+ on bead and monolithic geopolymer bed, Conditions: initial concentration = 100 mg L− 1; bed height = 0.5 cm; flow rate = 5 mL min.−1)

Full size image

Comparative study

To the best of the authors’ knowledge, this study offers the first performance assessment of 3D-printed geopolymer monolithic and bead-type adsorption beds designed especially for the removal of Cd2+ ions. Together with the geopolymer-based adsorbents used in this study, Table 6 provides a thorough comparison of the adsorption capacities and manufacturing processes of several previously published adsorbent systems.

Among the adsorbents listed, monoliths prepared from polyethene oxide-based silica and chitosan/perlite beads exhibit significantly high adsorption capacities of 153 mg/g and 178.6 mg/g, respectively. However, these materials often rely on costly precursors or require sophisticated phase separation or inversion techniques that may not be scalable or environmentally benign. Similarly, the sodium alginate/thiol-modified lotus seedpod biochar beads show promising adsorption efficiency (138.73 mg/g) but lack information on reusability, which is critical for practical deployment.

In contrast, the 3D-printed geopolymer monoliths fabricated through DIW technique in this study demonstrate a moderate but practical maximum adsorption capacity of 35.9 mg/g. More importantly, these monoliths outperform most of the listed systems in terms of reusability, with a demonstrated stability over eight adsorption-desorption cycles, which is the highest among all compared materials. This resilience under repeated use highlights the potential of 3D-printed geopolymer monoliths as a durable and sustainable adsorbent system. Notably, many of the previously reported adsorbents, including clay-based honeycomb monoliths (e.g., illite-smectite and stevensite), exhibit low adsorption capacities (1.21–4.58 mg/g) and limited reusability data, thereby reducing their appeal for industrial applications. These systems’ scalability is restricted by their limited regeneration capacity and poor mechanical strength. The 3D-printed geopolymer monoliths developed in this work thus fill the gap between process scalability, performance, and reusability.

Table 6 Comparative overview of cd²⁺ adsorption by bulk-type adsorbents from previous studies.
Full size table

Conclusions

This study demonstrated the successful synthesis of a MK-based geopolymer and its subsequent fabrication into monolithic structures using the DIW technique. For the first time, a comparative evaluation was carried out between bead-shaped and monolithic adsorbents fabricated from the same material composition. The investigation aimed to assess their performance for Cd²⁺ adsorption from aqueous solutions and to evaluate their potential for scale-up in industrial applications. A comprehensive parametric study was conducted to optimise operational conditions for Cd²⁺ removal. Experiments showed that the geopolymer adsorbed materials in a way matched to the pseudo-second-order model for speed and the Langmuir isotherm for equilibrium. These results suggest that cadmium adsorption primarily occurs via monolayer chemisorption. The maximum adsorption capacity achieved was 87.1 mg/g at an initial Cd²⁺ concentration of 100 mg/L, pH 5, and adsorbent dosage of 1.5 mg/L, comparable to values reported for powdered geopolymer adsorbents in the literature1,8,49.

Under consistent operational parameters, column studies were performed using both bead and monolithic beds. The resulting breakthrough curves from the fixed-bed adsorption process were accurately modelled by the Thomas model. This suggests that cadmium sorption follows a pseudo-second-order mechanism, is reversible, and occurs with minimal axial dispersion. The maximum adsorption capacities were determined to be 37.5 mg/g for beads and 35.9 mg/g for monoliths. While beads exhibited slightly higher adsorption capacity, the monoliths demonstrated superior mass transfer characteristics, as evidenced by a shorter mass transfer zone, indicating more efficient intraparticle diffusion and adsorption kinetics. The monolithic bed also showed excellent regeneration potential, retaining its structural integrity and adsorption efficiency over eight consecutive adsorption–desorption cycles. Furthermore, the monoliths were easier to handle, regenerate, and maintain due to their robust and self-supporting structure.

Overall, this study highlights the feasibility of using additive manufacturing techniques, such as DIW, for the one-step fabrication of geometrically complex, porosity-controlled, structured adsorbents. The promising performance, ease of regeneration, and mechanical stability of geopolymer-based monoliths indicate their strong potential for large-scale industrial wastewater treatment applications targeting heavy metal removal.