Waste-Derived Porous Geopolymers for Pb(II) Removal: Kinetics, Thermodynamics, and Regeneration

Abstract

Lead (Pb) is a highly toxic heavy metal frequently found in industrial wastewater, posing serious risks to both human health and the environment. In this study, a porous geopolymer synthesized from fly ash, metakaolin, and red mud was evaluated for Pb(II) removal via batch adsorption experiments under varying pH, dosage, contact time, temperature, and initial concentration. The synthesized material exhibited a favorable mesoporous structure, with a BET surface area of 42.05 m² g⁻¹ and an average pore width of 6.26 nm, making it suitable for heavy metal uptake. Adsorption kinetics followed the pseudo-second-order model (R² = 0.9993), while the Langmuir isotherm (R² ≈ 0.999) best described the equilibrium data, indicating monolayer chemical adsorption as the dominant mechanism, with a maximum capacity of 74.26 mg g⁻¹ at 318 K. Thermodynamic analyses confirmed that the adsorption was spontaneous (ΔG° < 0), endothermic (ΔH° > 0), and accompanied by increased entropy (ΔS° > 0). Desorption and regeneration tests revealed EDTA to be a more effective agent than HNO₃, maintaining a reuse efficiency of 81.35% after four cycles. These results highlight the potential of waste-derived porous geopolymers as regenerable, low-cost, and efficient adsorbents for lead removal.

1. Introduction

Heavy metal pollution, particularly from industrial sources, poses a serious and persistent threat to environmental and human health. Among these contaminants, Pb(II) is especially concerning due to its high toxicity, non-biodegradability, and ability to bioaccumulate, even at trace concentrations. Exposure to Pb(II) is linked to neurological, developmental, and renal disorders, necessitating effective strategies for its removal from water systems [1,2].

Traditional lead removal methods, such as chemical precipitation, ion exchange, and membrane filtration, mainly suffer from sludge generation, high operational costs, and difficulties in their applications. In contrast, adsorption has been favored in recent years due to its cost-effectiveness, simplicity, and high efficiency, with growing interest in using low-cost, environmentally sustainable adsorbents [3,4].

Geopolymers, synthesized by the alkaline activation of natural and waste aluminosilicate feedstocks, have recently emerged as promising candidates for heavy metal adsorption [5,6]. Particularly, using industrial by-products like fly ash and red mud both valorizes waste streams and lowers the material’s carbon footprint relative to conventional adsorbents. In other words, the simultaneous role of waste utilization and heavy-metal capture makes geopolymers attractive for sustainable remediation [7,8,9].

On the other hand, many geopolymer-based adsorbents suffer from limited adsorption capacity, low surface area, or require harsh synthesis conditions. In general, waste-derived geopolymers tend to exhibit Pb(II) uptake capacities on the order of tens of mg g⁻¹ [8,10], while higher values (hundreds of mg g⁻¹) have been reported mainly in engineered composites [11,12], but such enhancements often require more complex synthesis or non-sustainable additives. Thus, most simple alkali-activated sorbents trade off ultimate capacity for easier preparation and stronger sustainability.

The present work offers a porous geopolymer adsorbent prepared by a simple one-pot activation of mixed-waste precursor (coal fly ash, metakaolin, and red mud) without the need for expensive templates or high-temperature steps. A favorable mesoporous structure was synthesized through a combination of the multi-component feedstock and H₂O₂-assisted foaming for Pb(II) removal. Pb(II) stock solutions were prepared from Pb(NO₃)₂, which serves as a representative soluble lead source to simulate industrial wastewaters such as those generated from battery manufacturing, electroplating, and pigment production processes. Pb(II) removal efficiency was thoroughly examined under various processing conditions. Adsorption kinetics, isotherms, thermodynamics, and desorption behavior of the geopolymer were also assessed. This work contributes to the development of low-cost, regenerable, and sustainable adsorbents for heavy metal remediation.

2. Materials and Methods

2.1. Materials

In the synthesis of the geopolymer, aluminosilicate sources comprised coal fly ash (CFA), metakaolin (MK), and red mud (RM) obtained from local industrial facilities. All materials were only oven-dried without any other pre-treatments prior to use to ensure a sustainable approach. Alkali activation was achieved using 3 modulus sodium silicate (Na₂SiO₃) and NaOH solutions. To induce porosity and ensure pore stability within the geopolymer matrix, H₂O₂ and sodium dodecyl sulfate (C₁₂H₂₅OSO₂ONa, SDS) solutions were employed, respectively. During the adsorption experiments, the pH of the solutions was adjusted using 0.1 M HCl and NaOH solutions. All experiments were conducted using deionized water and reagents of analytical grade.

The chemical compositions and basic physical properties of the starting materials are presented in

Table 1. The chemical composition and physical properties of the starting materials.

Chemical Constituents (%)	CFA	MK	RM
SiO2	53.4 ± 2.67	50.7 ± 2.54	15.2 ± 0.76
Al2O3	27.2 ± 1.36	44.5 ± 2.23	22.1 ± 1.10
Fe2O3	7.8 ± 0.39	2.1 ± 0.11	33.3 ± 1.67
CaO	1.4 ± 0.07	0.4 ± 0.02	3.6 ± 0.18
MgO	2.2 ± 0.11	0.2 ± 0.01	0.4 ± 0.02
Na2O	0.4 ± 0.02	0.1 ± 0.01	9.4 ± 0.47
K2O	4.6 ± 0.23	0.2 ± 0.01	0.7 ± 0.03
TiO2	1.2 ± 0.06	1.4 ± 0.07	5.1 ± 0.25
Loss on ignition (LOI, %)	1.8	0.3	10.3
Physical Properties
Average particle size (µm)	34.50	6.51	0.97
Specific gravity	2.19	2.49	2.94
BET surface area (m2 g−1)	1.48	1.08	47.06

Note: The total oxide content was normalized to 100%. Minor variations are attributed to trace elements. Table 1 has two subsets which are chemical constituents and physical properties of the precursors.

. The oxide compositions are presented as mean ± standard deviation (SD) values (SD ≤ 5%) based on triplicate XRF analyses to account for the inherent variability in the chemical composition of these industrial by-products. According to this table, 53.4% SiO₂ and 27.2% Al₂O₃ contents highlight the main aluminosilicate framework of CFA. Notably high Al₂O₃ (44.5%) and similar SiO₂ (50.7%) contents were determined for MK, confirming its suitability as a reactive aluminosilicate source material. In contrast, RM sample, a by-product of aluminum production, displayed a distinctly different composition, characterized by a low SiO₂ content (15.2%), high Fe₂O₃ content (33.3%), and a relatively high loss on ignition (10.3%). Similar chemical compositions were reported in previous studies [8,13,14].

As also shown in Table 1, it is evident that CFA exhibits a larger particle size (34.50 µm) and a lower specific gravity (2.19) compared to MK and especially RM. The significantly higher surface area of RM (47.06 m² g⁻¹) relative to the other materials is likely attributed to its finer particle size (0.97 µm) and high loss on ignition (10.3%), which may indicate the presence of residual carbon. Similar observations were reported in previous studies [8,13,15].

2.2. Synthesis of the Porous Geopolymers

A 10 M NaOH solution was initially prepared and subsequently mixed with Na₂SiO₃ at a 1:3 mass ratio to obtain the alkali activator solution. Based on the preliminary trials, the precursor (60 wt.% CFA, 20 wt.% MK, and 20 wt.% RM) was added into the activator solution at a solid-to-activator mass ratio of 0.83, and the mixture was stirred for 10 min to achieve homogeneity. A 3.5% H₂O₂ solution was then introduced as a foaming agent at a dosage of 0.5–2.5 wt.% relative to the total solid precursor, while a 300 mg L⁻¹ SDS solution was used at a fixed dosage of 1.5 wt.% to stabilize the unstable foam, maintain pore structure, and prevent coalescence during formation. The selected H₂O₂ dosage range (0.5–2.5 wt.%) was determined through preliminary optimization trials conducted in triplicate to evaluate porosity and bulk density of the resulting geopolymers, as well as lead removal efficiency based on statistical analysis of variance (p < 0.05). Following the secondary mixing, the homogeneous geopolymeric slurry was poured into molds and initially cured in an oven at 343 K for 24 h. After demolding, the samples were sealed and allowed to cure at room temperature for 7 days. The hardened products were ground and sieved through 75 µm to ensure reproducible adsorption performance, and to minimize size effect and mass transfer limitations during batch experiments. They were then neutralized with diluted HCl and deionized water to remove residual alkalis. The neutralized samples were finally dried in an oven at 343 K for 24 h.

2.3. Characterization of the Materials

An IQ X-ray fluorescence (XRF) instrument (SPECTRO, Kleve, Germany) was utilized to determine the major chemical constituents of the raw materials. Loss on ignition (LOI) was assessed by calculating the mass loss following heating in an air oven at 750 °C for 3 h. X-ray diffraction (XRD) analysis was performed using an Empyrean diffractometer (Malvern Panalytical, Almelo, The Netherlands). The diffraction patterns were recorded over a 2θ range of 10–80°, operating at 45 kV and 40 mA, with Cu-Kα radiation (λ = 1.54051 Å). Microstructural properties and point elemental contents were examined using a Sigma 300 scanning electron microscope (SEM) equipped with an energy-dispersive X-ray spectroscopy (EDX) system (Carl Zeiss, Oberkocken, Germany). Surface and pore properties were determined using a 3Flex instrument (Micromeritics, Norcross, GA, USA) based on nitrogen adsorption at −197 °C. The Brunauer–Emmett–Teller (BET) method was applied over a relative pressure (P/P₀) range of 0.0–1.0 to calculate the specific surface area, while the pore volume was determined at P/P₀ = 0.95. The Barrett–Joyner–Halenda (BJH) model was employed to derive pore size distributions from the adsorption isotherms.

2.4. Batch Adsorption Experiments

The applicability of porous geopolymers for lead adsorption was evaluated through batch laboratory experiments using synthetic wastewater samples. Pb²⁺ solutions were derived from a 1000 mg L⁻¹ Pb(NO₃)₂ stock solution. Pb(NO₃)₂ is a stable and soluble lead salt commonly used to simulate industrial effluents. Since this approach represents the ionic Pb²⁺ species typically found in wastewater from battery, pigment, and metal plating industries, the results can be extended to real lead-containing systems where Pb²⁺ dominates.

In the experiments, solution volume of 100 mL (0.1 L) and shaking speed of 150 rpm were kept constant, while the effects of H₂O₂ dosage (0.5–2.5 wt.% relative to the total solid), adsorbent dosage (0.5–2.5 g L⁻¹), initial lead concentration (25–200 mg L⁻¹), pH (3–11), and temperature (298–318 K), were systematically investigated as a function of time. At predetermined time intervals, 5 mL samples were withdrawn and centrifuged at 5500 rpm for 5 min. To ensure complete removal of solid particles from the solution, the supernatant was filtered through a syringe filter and made ready for analysis. Lead concentrations were determined spectrophotometrically using Merck lead test kits (Product No: 1097170001; measurement range: 0.01–5.00 mg L⁻¹).

Triplicate measurements and one-way ANOVA (p > 0.05) were used to evaluate results through batch adsorption experiments. Lead removal efficiency (RE, %) and the adsorption capacity of the geopolymer at equilibrium (q_e, mg g⁻¹) were used as performance indicators. These values were calculated using the following equations:

$$RE (\%)={\displaystyle \frac{\left(C_0-C_t\right)}{C_0}}\times 100$$

(1)

$$q_e \left(\mathrm{mg} {\mathrm{g}}^{-1}\right)={\displaystyle \frac{{(C}_0-C_e)V}{W}}$$

(2)

Here, C₀, C_t, and C_e (mg L⁻¹) represent the initial, time-dependent and equilibrium concentrations of lead in the solution, respectively. Additionally, V (L) denotes the volume of the solution, and W (g) refers to the mass of the adsorbent.

2.5. Reusability Studies

The ability to reuse an adsorbent is an important factor in evaluating the cost-effectiveness of an adsorption system, especially when it is not naturally abundant [16]. In this context, an initial adsorption was carried out using a 2 g L⁻¹ adsorbent dose, 100 mg L⁻¹ Pb²⁺ concentration, pH 7, at 298 K for 4 h. After filtration and overnight drying at 343 K, the lead-loaded geopolymers were subjected to desorption using 100 mL (0.1 L) of either 0.01 M HNO₃ or 0.05 M EDTA solutions at 298 K for 16 h. Following each extraction, the material was rinsed with deionized water and dried. The sorbent was then reused in the subsequent adsorption cycle under the same conditions as the initial process. This desorption–adsorption cycle was repeated four times. Desorption (D, %) and reuse efficiencies (R, %) were calculated using Equations (3) and (4), respectively [17]:

$$D (\%)=\left({\displaystyle \frac{C_d}{C_0-C_t}}\right)\times 100$$

(3)

$$R \left(\%\right)=\left({\displaystyle \frac{C_{R0}-C_{Rt}}{C_0-C_t}}\right)\times 100$$

(4)

Here, C_d (mg L⁻¹) indicates the concentration of Pb²⁺ remaining in the solution after desorption. C_R0 and C_Rt (mg L⁻¹) represent the Pb²⁺ concentrations in the solution before and after adsorption using the reused material, respectively. Similarly, C₀ and C_t (mg L⁻¹) denote Pb²⁺ concentrations in the solution before and after the initial adsorption, respectively.

3. Results and Discussion

3.1. The Effect of H₂O₂ Dosage

The effect of H₂O₂, used as a foaming agent, on pore formation was investigated within the range of 0.5–2.5 wt.% relative to the total solid content. The quality of the resulting products was assessed through porosity and bulk density analyses, as well as lead adsorption experiments. As shown in Figure 1a, an increase in the H₂O₂ content during geopolymer synthesis leads to higher porosity in the final products, accompanied by a corresponding decrease in bulk density. A porosity of 38.5% was obtained with a 0.5 wt.% H₂O₂ addition, whereas this value increased to 63.4% and 67.7% for 1.5 wt.% and 2.5 wt.% H₂O₂ additions, respectively. The corresponding bulk density values were determined as 1.44, 0.85, and 0.75 g cm⁻³ for 0.5 wt.%, 1.5 wt.%, and 2.5 wt.% H₂O₂, respectively. These findings indicate that a 1.5 wt.% H₂O₂ addition is sufficient to achieve desirable porosity while maintaining acceptable structural properties [18].

As shown in Figure 1b, the removal efficiency increases with both the H₂O₂ content and contact time. After a 3 h adsorption period, the removal efficiency for the sample without H₂O₂ was recorded as 60.48%, while the efficiencies for samples with 0.5 wt.%, 1.0 wt.%, and 1.5 wt.% H₂O₂ additions increased to 82.68%, 86.72%, and 96.96%, respectively. Slightly higher removal efficiencies of 97.64% and 98.56% were achieved with H₂O₂ dosages of 2.0 wt.% and 2.5 wt.%, respectively. However, considering economic and environmental sustainability, 1.5 wt.% H₂O₂ was identified as the optimal dosage. The resultant geopolymer was designated as PGEO and selected for subsequent analyses.

3.2. Characterization Results

3.2.1. Mineralogic and Microstructural Analysis

Figure 2 presents the detailed XRD analysis of the samples. The Inorganic Crystal Structure Database (ICSD) was utilized to evaluate the diffraction patterns. The XRD analysis revealed that RM sample primarily consists of hematite (00-024-0072), gibbsite (00-007-0324), and labradorite (01-076-0949), with semiquantitative inclusions of 46.1%, 23.4%, and 19.2%, respectively. Minor phases such as chantalite (01-083-1450, 6.1%) and rutile (01-076-0319, 5.2%) were also detected. The dominant crystalline phases for MK were identified as sillimanite (01-074-0274) and quartz (01-083-2471), with approximate semiquantitative abundances of 54% and 43%, respectively, while it also includes 3% hematite (01-087-1165). CFA exhibited a semiquantitative composition of about 57% mullite (01-079-1454), 35% quartz (01-078-1252), and 8% hematite (01-072-0469) [15,19].

A significant reduction in crystalline peak intensity, along with the appearance of a broad hump in the 2θ range of 15–35° for PGEO, indicates a predominantly amorphous gel structure [8]. A semi-quantitative estimation of the amorphous content was conducted based on integrated intensity ratios of the crystalline and amorphous regions in the XRD patterns. The analysis revealed that the amorphous phase in PGEO accounts for approximately 62–65% of the total composition. Consistent with the precursor compositions, the dominant residual crystalline inclusions were semiquantitatively determined as mullite (01-074-2419), quartz (01-085-1054), and hematite (01-087-1166), with 64.6%, 22.2%, and 9.1%, respectively. In addition, a newly formed sodalite phase (01-076-1639) was identified with 4.1%, confirming successful geopolymerization and structural reorganization [5,11,20,21].

Figure 3 displays SEM images and EDX spectra of PGEO before (a and b) and after (c and d) the adsorption of Pb(II). Elemental compositions are presented as mean ± SD (≤5%) values based on triplicate EDX measurements at randomly chosen regions. As seen in Figure 3a, the surface morphology of PGEO appears heterogeneous, with a clear presence of pores, and mostly irregular and rough particles, including plate-like microdomains and agglomerates. As expected from a geopolymerization product, the main structure is dominated by amorphous aluminosilicate gel phases [15,22]. The elemental distribution presented in Figure 3b confirms the aluminosilicate framework, with Fe mostly originating from the red mud component. The very low Na content (0.23 ± 0.01%) suggests minimal residual alkali or effective post-treatment washing.

As shown in Figure 3c, the surface becomes more compact and granular after the adsorption, indicating the filling of pores or deposition of Pb species. These surface topography changes suggest that the adsorption mechanism is chiefly based on chemisorption rather than simple physical interaction. The emergence of the Pb signal appeared around 2.3 keV in Figure 3d, confirming successful adsorption [8]. The average Pb content after adsorption was determined as 4.25 ± 0.21%, indicating uniform Pb distribution across the geopolymer surface through inner-sphere complexation or ion exchange.

3.2.2. Surface and Pore Properties

The nitrogen adsorption–desorption isotherms and the BJH pore size distribution curves of the raw materials and PGEO are presented in Figure 4. According to IUPAC classification, both CFA and MK exhibit Type IV isotherms with H3-type hysteresis, typical for slit-shaped mesopores, and they show minimal nitrogen uptake (Figure 4a), especially at P/P₀ < 0.9. In contrast, as shown in Figure 4b, PGEO exhibits significantly higher N₂ uptake and displays a Type IV isotherm with H4-type hysteresis, often associated with narrow slit-like pores and possible microporosity. Despite its high nitrogen uptake, RM displays a Type III–IV hybrid isotherm, accompanied by a weak H3/H4 hysteresis loop, suggesting less ordered porosity [23,24]. Overall, Figure 4a,b signifies enhanced pore structure of PGEO, mostly with highly ordered mesopores, likely attributed to H₂O₂ foaming and SDS stabilization during synthesis.

A sharp and prominent peak for PGEO observed at 3–4 nm in Figure 4c represents the modal pore diameter derived from the BJH distribution, whereas the average pore width (6.26 nm) listed in the inset table corresponds to the mean value of the adsorption and desorption branches using the BET method [25]. This peak is also indicative of the uniform and hierarchical mesoporosity of PGEO, which facilitates accessibility for Pb(II) ions during adsorption [11]. In contrast, a broader distribution of RM signifies its heterogeneous pore network [26]. On the other hand, CFA and MK exhibit a limited to negligible intrinsic porosity. The overall BJH results confirm that PGEO has the most uniform and accessible mesoporous structure among all, which enhances both kinetic accessibility and adsorption capacity.

3.3. The Effects of Operational Parameters

3.3.1. Adsorbent Dosage and Initial Lead Concentration

The effect of adsorbent (PGEO) dosage (0.5–2.5 g L⁻¹) on lead removal efficiency was investigated over a contact time range of 30–180 min, while keeping the initial lead concentration constant at 50 mg L⁻¹.

According to Figure 5a, all curves show a sharp initial increase within the first 15 min, indicating fast Pb²⁺ uptake and strong surface affinity [27]. At 2.5 g L⁻¹, Pb²⁺ removal exceeds 95% within 60 min, while at 0.5 g L⁻¹, it plateaus around 63%, even after 120 min. This confirms that increased dosage improves both the adsorption rate and final removal efficiency, due to a greater number of available binding sites [28]. Equilibrium time was defined when successive adsorption values differed by <5%. At dosages ≥ 1.0 g L⁻¹, one-way ANOVA confirmed no significant variation (p > 0.05) beyond 120 min. Therefore, equilibrium time was accepted as 120 min. The optimal dosage range seems to be around 1.5–2.0 g L⁻¹, beyond which the gain in removal efficiency becomes marginal [10]. Figure 5a clearly demonstrates that the adsorption kinetics support the practical viability of PGEO for real-world wastewater treatment systems where rapid treatment cycles are advantageous. Based on these findings, the optimum PGEO dosage was determined to be 2.0 g L⁻¹.

The influence of initial Pb²⁺ concentration (25–200 mg L⁻¹) on removal efficiency and adsorption capacity was examined over a contact time range of 30–240 min using the optimum PGEO dosage of 2.0 g L⁻¹. According to Figure 5b, as the initial Pb²⁺ concentration increases, removal efficiency decreases due to saturation of available adsorption sites on PGEO [5]. All curves exhibit rapid initial Pb²⁺ uptake within the first 60 min, followed by gradual flattening as equilibrium is approached. At 120 min, removal efficiencies exceeding 90% were achieved for initial concentrations up to 75 mg L⁻¹, while the efficiency decreased to 82.88% for 100 mg L⁻¹ and dropped dramatically to 42.94% for 200 mg L⁻¹. At low Pb²⁺ concentrations, more active sites are available per ion, resulting in higher removal, while competition among Pb²⁺ ions leads to reduced efficiency at high concentrations [29].

As seen from the inset bar chart in Figure 5b, adsorption capacity (q) increases with increasing initial Pb²⁺ concentration. At 240 min, q rises from 12.46 mg g⁻¹ (25 mg L⁻¹) to 51.22 mg g⁻¹ (200 mg L⁻¹). At 75 mg L⁻¹, the adsorption capacity at 120 min is 34.18 mg g⁻¹, and at 100 mg L⁻¹ it reaches 41.44 mg g⁻¹. Even though removal efficiency drops at high concentrations, adsorption capacity increases due to the high availability of Pb²⁺ [30]. Based on balancing high enough efficiency (>90%) and considerable capacity (~34 mg g⁻¹), the optimum initial Pb²⁺ concentration is identified as 75 mg L⁻¹. Contact time of 120 min is sufficient to achieve near-equilibrium performance for most concentrations [30].

3.3.2. Initial pH and Temperature

Figure 6a illustrates the effect of initial pH on Pb²⁺ removal efficiency over time, along with a bar chart inset showing the final pH values after 120 min of adsorption.

At pH 3, removal efficiency never exceeds 30% and declines over time, indicating severe competition with H⁺ ions and electrostatic repulsion due to the protonation of functional groups. At pH 5, moderate improvement was observed in removal efficiency, reaching 82.93%, 91.49% and 97.06% at 60, 120, and 240 min, respectively. At pH 7, high efficiencies of 90.19%, 95.36% and 98.03% were recorded for the respective contact times of 30, 60, and 120 min. From pH 9 to 11, excellent performance (>99% removal) was achieved within 30 min due to the high deprotonation of surface groups. In addition, the inset final pH values confirm that there is no Pb(OH)₂ precipitation up to pH 11, indicating true adsorption. Overall, Figure 6a indicates that PGEO demonstrates broad pH applicability and excellent stability [31]. The optimal Pb²⁺ removal occurs at pH 7, where an efficiency of more than 95% is achieved in just one hour.

Under optimized conditions (2.0 g L⁻¹ PGEO dosage, pH 7, and 2 h contact time), the influence of temperature (298, 308, and 318 K) on Pb(II) removal efficiency and adsorption capacity was examined at five initial concentrations of 75, 100, 125, 150, and 200 mg L⁻¹. The results, illustrated in Figure 6b, reveal that both removal efficiency and adsorption capacity increased with temperature. This effect became more pronounced at higher concentrations (≥125 mg L⁻¹), indicating that the adsorption process is endothermic in nature [28,32].

At any given temperature, removal efficiency tended to decrease with increasing initial Pb(II) concentration, due to saturation of active sites; however, adsorption capacity consistently increased, as more Pb²⁺ ions were available for binding. For instance, at an initial concentration of 125 mg L⁻¹, the removal efficiencies at 298, 308, and 318 K were 87.58%, 94.37%, and 96.32%, respectively, with corresponding adsorption capacities of 54.74, 58.98, and 60.20 mg g⁻¹. At 150 mg L⁻¹, removal efficiencies declined to 74.77%, 87.68%, and 89.76%, while adsorption capacities increased to 56.08, 65.76, and 67.32 mg g⁻¹ over the same temperature range.

When considering both removal efficiency and adsorption capacity, the most balanced performance was observed at 298 K for 125 mg L⁻¹ and at 308 K for 150 mg L⁻¹. The highest adsorption capacity of 74.26 mg g⁻¹ was achieved at 318 K for 200 mg L⁻¹, highlighting the strong Pb(II) uptake potential of PGEO under thermally enhanced conditions.

3.4. Isotherm Studies

To gain deeper insight into the adsorption behavior of Pb(II) onto PGEO, the Langmuir, Freundlich, and Temkin isotherm models were applied to the experimental data. Along with the corresponding equations,

Table 2. Parameters of the adsorption isotherms used. Experimental conditions: 2.0 g L⁻¹ PGEO dosage, pH 7, 2 h-contact time at 298, 308, and 318 K.

Isotherm	Equation	Parameter	298 K	308 K	318 K
Langmuir	${\displaystyle \frac{C_e}{q_e}}={\displaystyle \frac{1}{K_L{q}_m}}+\left({\displaystyle \frac{1}{q_m}}\right)C_e$	qe,exp (mg g−1)	58.53 a	71.95 a	74.26 a
		qm (mg g−1)	59.17 b	73.53 b	75.19 b
		KL (L mg−1)	0.7647	0.7273	0.9852
		R2	0.9998	0.9995	0.9995
	$R_L={\displaystyle \frac{1}{1+K_L{C}_0}}$	C0 (mg L−1)	RL
		75	0.0171	0.0180	0.0134
		100	0.0129	0.0136	0.0100
		125	0.0104	0.0109	0.0081
		150	0.0086	0.0091	0.0067
		200	0.0065	0.0068	0.0050
Freundlich	$log{q}_e=log{K}_F+{\displaystyle \frac{1}{n}}log{C}_e$	KF	37.59	41.23	44.37
		n	8.97	6.56	6.84
		R2	0.9202	0.8941	0.9056
Temkin	$q_e=K_{T}ln{a}_T+K_{T}ln{C}_e$	KT = RT/b	5.27	8.20	8.02
		aT	1168.26	147.52	269.05
		R2	0.9430	0.9451	0.9552

^a Experimental values at equilibrium obtained for the initial Pb²⁺ concentration of 200 mg L⁻¹. ^b Theoretical maximum adsorption capacities calculated from the Langmuir isotherm.

summarizes the resulting parameter values and associated correlation coefficients (R²) calculated using linear regression analysis.

In the Langmuir model, q_m (mg g⁻¹) denotes the theoretical maximum adsorption capacity, while K_L (L mg⁻¹) represents the Langmuir constant, which is associated with the affinity between the adsorbent and adsorbate, and the free energy of adsorption [20,33]. As presented in Table 2, the value of q_m increased with temperature from 59.17 mg g⁻¹ (298 K) to 75.19 mg g⁻¹ (318 K), indicating that the adsorption process is endothermic. The highest K_L value (0.9852), recorded at 318 K, further specifies enhanced interaction strength between Pb(II) ions and the geopolymer surface at elevated temperatures. The Langmuir model exhibited an excellent fit to the experimental data across all temperatures, as evidenced by consistently high correlation coefficients (R² ≈ 0.999). Additionally, the separation factor (R_L) values calculated for all initial concentrations remained between 0 and 1, confirming that the Pb(II) adsorption process is favorable and proceeds via a monolayer adsorption mechanism [29].

The Freundlich isotherm model is widely applied to describe adsorption on heterogeneous surfaces and to evaluate non-uniform adsorption energies. In this model, the Freundlich constant (K_F) [(mg g⁻¹) (L mg)^−1/n] reflects the adsorption capacity, while the heterogeneity factor (n) provides insight into the adsorption intensity and surface heterogeneity [22,34]. As seen in Table 2, the value of K_F increases from 37.59 to 44.37 with rising temperature, further supporting the endothermic nature of the adsorption process. Additionally, the values of n at all tested temperatures is higher than 1, suggesting the favorability of physical interaction on a heterogeneous surface [19]. However, the relatively lower R² values calculated (≈0.91) specifies the weaker fit of the Freundlich model, reinforcing that Pb(II) adsorption onto PGEO is better described by a monolayer coverage on a homogeneous surface, as represented by the Langmuir isotherm.

The Temkin isotherm considers indirect adsorbate–adsorbent interactions and assumes that the heat of adsorption decreases linearly as the surface becomes increasingly occupied [35]. In this model, the Temkin adsorption energy parameter (a_T) exhibited a considerable variation with temperature, which was 1168.26 at 298 K, followed by a dramatic decline to 147.52 at 308 K, and a subsequent increase to 269.05 at 318 K. This non-linear trend suggests different dominate adsorption mechanisms at varying thermal conditions [23].

3.5. Kinetic Studies

The pseudo-first-order, pseudo-second-order, and intraparticle diffusion kinetic models were applied to the experimental data obtained over a contact time range of 10–240 min, under the conditions of 2 g L⁻¹ adsorbent dosage, 100 mg L⁻¹ initial lead concentration, 298 K temperature, and pH 7. Together with the corresponding equations used, the model parameters and associated correlation coefficients (R²) are presented in

Table 3. Parameters of the kinetic models used. Experimental conditions: 2.0 g L⁻¹ PGEO dosage, 100 mg L⁻¹ initial Pb²⁺ concentration, 298 K, and pH 7.

Kinetic Model	Equation	Parameter	Value
Pseudo-first-order model	$ln\left(q_e-q_t\right)=ln{q}_e-k_{1}t$	qe,cal (mg g−1)	20.68
		k1 (g mg−1 min−1)	0.0264
		R2	0.8902
Pseudo-second-order model	${\displaystyle \frac{t}{q_t}}={\displaystyle \frac{1}{k_2{q}_e^2}}+{\displaystyle \frac{t}{q_e}}$ $h=k_2{q}_e^2$	qe,exp (mg g−1)	48.49
		qe,cal (mg g−1)	49.26
		k2 (g mg−1 min−1)	0.0043
		h (mg g−1 min−1)	10.36
		R2	0.9993
Intraparticle diffusion model	$q_t=k_{id}{t}^{1/2}+C$	kid (mg g−1 min−0.5)	1.0734
		C	34.912
		R2	0.8037

.

As presented in Table 3, the pseudo-first-order model exhibited a relatively low fit (R² = 0.8902), and a significant difference exists between the experimental and the calculated adsorption capacities of 48.49 and 20.68 mg g⁻¹, respectively. This notable inconsistency suggests that the adsorption kinetics cannot be represented by this model. In contrast, the pseudo-second-order model provided an almost perfect fit (R² = 0.9993) and a strong agreement between the experimental and calculated (49.26 mg g⁻¹) uptake values, confirming that the adsorption process follows second-order kinetics. The rate constant (k₂) and initial sorption rate (h) were determined as 0.0043 g mg⁻¹ min⁻¹ and 10.36 mg g⁻¹ min⁻¹, respectively. These findings indicate predominance of chemical interactions through electron sharing or exchange between Pb(II) ions and the surface functional groups of PGEO [29,36,37]. On the other hand, a markedly lower fit (R² = 0.8037) and the intercept (C) greater than zero (34.912) were noted for the intraparticle diffusion model, signifying that the overall adsorption is significantly affected by surface adsorption and other mechanisms.

3.6. Thermodynamic Investigations

Thermodynamic parameters, the changes in Gibbs free energy (ΔG°), enthalpy (ΔH°), and entropy (ΔS°) were determined using the equilibrium data obtained at 125 mg L⁻¹ initial Pb²⁺ concentration, which provided the highest correlation coefficient (R² = 0.9734) in the van’t Hoff plot. This approach was adopted to ensure that the derived parameters accurately represent the intrinsic adsorption mechanism under well-fitted equilibrium conditions.

As seen in

Table 4. Thermodynamic parameters of the adsorption process. Experimental conditions: 2.0 g L⁻¹ PGEO dosage, 125 mg L⁻¹ initial Pb²⁺ concentration, pH 7, 2 h contact time at 298, 308, and 318 K.

Equations Used	T (K)	ΔG° (kJ mol−1)	R2	ΔH° (kJ mol−1)	ΔS° (kJ mol−1 K−1)
$∆G^o=-RTln{K}_c$ $K_c={\displaystyle \frac{C_{sorbed, e}}{C_e}}$ $ln{K}_c={\displaystyle \frac{{∆S}^o}{R}}-{\displaystyle \frac{{∆H}^o}{RT}}$	298	−4.84
	308	−7.22	0.9734	51.81	0.191
	318	−8.63

, the values of ΔG° are all negative and become more negative with increasing temperature. ΔG° values of −4.84, −7.22, and −8.63 kJ mol⁻¹ were sequentially calculated at 298, 308, and 318 K, suggesting that the adsorption process is spontaneous and thermodynamically more favorable at higher temperatures [11]. The positive ΔH° (51.81 kJ mol⁻¹) indicates endothermic nature of the adsorption, supporting the increasing thermodynamic favorability with rising temperature [10,17]. Since the ΔH° value obtained is higher than 40 kJ mol⁻¹, it can be said that the adsorption process is primarily governed by chemisorption [38,39]. The positive ΔS° (0.191 kJ mol⁻¹ K⁻¹) reflects an increase in randomness at the solid–solution interface, which can be attributed to the displacement of water molecules and the more disordered arrangement of adsorbed Pb(II) ions [22].

In summary, thermodynamic investigations indicate spontaneous (ΔG° < 0) and endothermic (ΔH° > 0) nature of the adsorption, which is mainly governed by chemical interactions like ion exchange or surface complexation. The positive entropy value indicates an increase in system disorder during adsorption, and highlight the significant role of adsorbate–surface interactions in the overall process.

3.7. Desorption and Reuse Efficiencies

Figure 7 presents the results of consecutive desorption-reuse tests conducted using EDTA and HNO₃ solutions. As shown in Figure 7a, in the first cycle, HNO₃ exhibited a higher desorption efficiency (86.49%) compared to EDTA (54.92%). However, HNO₃′s performance declined sharply with each subsequent cycle, dropping to 26.80% by the fourth cycle, indicating potential deterioration of the geopolymer’s surface functionality. In contrast, EDTA interestingly showed an increasing trend in the second cycle (74.84%), before gradually declining in the following cycles. This temporary enhancement can be attributed to a surface activation phenomenon, where the initial desorption step removes loosely bound Pb(II) ions and surface impurities, thereby exposing previously inaccessible active sites. The greater desorption potential of EDTA compared to HNO₃ were also reported in the literature [8,17].

As shown in Figure 7b, after four cycles, the EDTA-treated PGEO maintained a removal efficiency above 81%, whereas only a 48.15% was achieved with the HNO₃-treatment. These results suggest that HNO₃ acts may gradually degrade the structural integrity or diminish the surface functionality of PGEO upon repeated use though its initially strong desorption capability. On the other hand, EDTA offers greater long-term regeneration performance, making it a more viable for sustainable Pb(II) removal applications [8,17].

3.8. Comparative Evaluation of PGEO with Other Geopolymer-Based Adsorbents

To further evaluate the adsorption performance of PGEO, its Pb(II) uptake capacity was compared with that of various geopolymer-based adsorbents reported in the literature (

Table 5. Comparative evaluation of PGEO with other geopolymer-based adsorbents. Reported adsorption capacities from literature were normalized to a 2 h contact time (q_t-2h) using the pseudo-second-order kinetic model, assuming equilibrium at the time reported by each source.

Geopolymer Properties		Operating Conditions					Eq. Uptake, qe (mg g−1)	k2 (g mg−1 min−1)	qt-2h (mg g−1)	Reference
Precursors/ Shape	Surface Area (m2 g−1)	Dosage (g L−1)	C0 (mg L−1)	pH	T (K)	Eq. Time (h)
CFA, MK, and RM/ Ground	42.05	2.0	200	7	318	2	74.26	0.00427	72.36	This study
MK/Ground	NI a	NI a	100	6	RT b	1	68.60	NI a	ND c (≈69)	[40]
(MK, and SF) G–S composite/Crushed	154.27	1.0	300	6.3	298	6	386.3	0.00022	352.03	[11]
MK, and RM/ Parallelepiped	28.8	NI a	600	5	RT b	6	30.70	NI a	ND c (<30.70)	[8]
MK/Cube	7.25	4.0	50	7	298	24	11.99 d	NI a	ND c (<11.99)	[31]
CGFA, SS, and MK/ Sphere	9.73	1.5	90	5–6	308	24	59.31	0.00037	42.82	[10]
(MK) G–A–C composite/Sphere	230	NI a	300	5	298	NI a	142.67	0.0215	142.28	[12]
MK, and BFA/ Cylindrical disk	NI a	NI a	50	5	RT b	24	6.34	NI a	ND c (<6.34)	[42]
MK/Sphere	53.95	1.5	100	5	NI a	60	45.60	0.00004	7.85	[41]

^a Not included. ^b Room temperature. ^c Cannot be determined due to lack of rate constant data. ^d Determined from simultaneous removal in the ternary (Pb-Ni-Cd) system. Abbreviations: A: Alginate, BFA: Biomass fly ash, C: Chitosan, CFA: Coal fly ash, CGFA: Coal gasification fly ash, G: Geopolymer, MK: Metakaolin, RM: Red mud, S: Sulfhydryl, SF: Silica fume, SS: Steel slag.

). For consistency, all reference data were normalized to a contact time of 2 h, matching the operational conditions used in this study. Normalized adsorption capacities (q_t-2h) were calculated using pseudo-second-order rate constant (k₂) and equilibrium adsorption capacity (q_e) values. After normalization, with an adsorption capacity of 72.36 mg g⁻¹, PGEO outperforms many unmodified geopolymers synthesized from metakaolin [31,40,41], biomass fly ash [42], coal gasification fly ash [10], and red mud [8].

While certain modified composites such as geopolymer–sulfhydryl [11] and geopolymer–alginate–chitosan [12] show higher normalized capacities of 352.03 mg g⁻¹ and 142.28 mg g⁻¹, respectively, high-cost precursors and complex preparation procedures limit their scalability. In contrast, PGEO offers a balanced performance, combining low-cost, simple preparation, moderate surface area (42.05 m² g⁻¹), rapid equilibrium (2 h), neutral pH operation, and moderately high regeneration ability. These features highlight its promising applicability in sustainable water treatment applications.

4. Conclusions

In this study, a porous geopolymer (PGEO) was successfully developed and evaluated for Pb(II) removal from aqueous solutions. The material exhibited a favorable porous structure and moderate surface area (42.05 m² g⁻¹), enabling efficient adsorption. Batch experiments demonstrated that Pb(II) adsorption was strongly influenced by operational parameters. The adsorption kinetics were best described by the pseudo-second-order model (R² = 0.9993), indicating chemisorption as the dominant mechanism. The Langmuir isotherm provided the best fit (R² ≈ 0.999), supporting a monolayer adsorption process with a maximum adsorption capacity of 74.26 mg g⁻¹ at 318 K. Thermodynamic analyses confirmed that the process was spontaneous (ΔG° < 0), endothermic (ΔH° > 0), and associated with increased randomness at the solid–liquid interface (ΔS° > 0). Regeneration tests showed that EDTA was a more effective desorbing agent than HNO₃, maintaining over 81% removal efficiency after four cycles. In conclusion, the overall findings suggest that PGEO can serve as an effective adsorbent due to its strong balance of performance, sustainability, and reusability. However, its practical applicability should be proven with future studies on multi-solute systems, real wastewaters, and long-term stability under flow conditions.