Gel-Inspired Trapping Networks: Fe(III)-Activated Palygorskite Nanorod Aggregates for Enhanced Congo Red Sequestration

TL;DR

A new method uses clay nanorods and iron to create a gel-like material that efficiently traps dye pollutants from water.

Contribution

Introduces a pollutant-induced gelation strategy using Fe(III)-activated palygorskite nanorods for enhanced dye sequestration.

Findings

Fe(III)-activated palygorskite nanorods showed 95.4–277% higher adsorption capacity for Congo Red compared to unmodified clay.

The adsorption process followed pseudo-second-order kinetics and the Temkin isotherm model.

CR molecules acted as inducers to reinforce the gel structure, enabling strong physical immobilization of dye aggregates.

Abstract

Developing adsorbents that combine high capacity with structural robustness remains a critical challenge for dye wastewater treatment. In this study, we propose a “pollutant-induced gelation” strategy to address this limitation, using Fe(III)-activated palygorskite nanorod aggregates as a model system for the highly efficient sequestration of Congo red (CR). Unlike conventional modification methods that rely solely on surface functionalization, this approach leverages the adsorbed dye itself as a synergistic assembly promoter. The addition of CR significantly consolidates the Fe(III)-mediated aggregation of palygorskite nanorods, leading to the formation of an integrated three-dimensional porous network with distinct gel-like rheological behavior. This dye-induced gel network not only provides abundant confined spaces for pollutant entrapment but also enhances the structural integrity of the adsorbent, facilitating separation and potential reuse. Compared to pristine palygorskite, the Fe(III)-activated material exhibited a 95.4–277% increase in adsorption capacity across a pH range of 4–10. The adsorption process followed pseudo-second-order kinetics and the Temkin isotherm model, and was thermodynamically spontaneous and exothermic. Mechanistic studies revealed a synergistic interplay: Fe(III) served as primary cross-linking nodes to construct the network framework, while CR molecules acted as inducers to reinforce the gel architecture, enabling strong physical immobilization of dye aggregates. This work provides a new paradigm for designing intelligent, gel-based adsorbents from natural nanoclays, transforming a pollutant into a structural promoter.

Figures

Full text

1. Introduction

Environmental pollution stemming from dye-laden wastewater, particularly from textile and printing industries, poses a severe global threat due to the discharge of toxic and recalcitrant compounds [1,2]. This wastewater jeopardizes aquatic ecosystems and human health, underscoring an urgent need for efficient remediation technologies [3,4,5,6].

Adsorption is a widely adopted method for dye removal, prized for its operational simplicity and cost-effectiveness [7,8]. A broad spectrum of adsorbents, including activated carbon, clay minerals, and advanced materials like metal–organic frameworks (MOFs), has been explored [9,10,11,12,13,14,15,16,17,18,19,20]. However, challenges persist regarding cost, regenerability, or selectivity [21]. In recent years, functionalized natural minerals have attracted considerable attention as low-cost adsorbents. Among various modification strategies, layered double hydroxide (LDH) and layered triple hydroxide (LTH) coated diatomites have been developed for Congo red removal. For instance, Sriram et al. modified diatomite with Mg-Al LDH and achieved a maximum adsorption capacity of 305.8 mg g^−1^, with the removal efficiency increasing from 15% to 98% at pH 7 [22]. However, the adsorption capacity significantly declined after three regeneration cycles, indicating limited reusability. In a subsequent study, the same group further engineered a Zn-Mg-Al LTH-modified diatomite, which exhibited high selectivity and efficient desorption performance [23]. Despite these advances, the synthesis of LDH/LTH composites often involves multi-step procedures and relatively high chemical consumption. Moreover, the structural stability of the modified minerals under continuous operation remains a challenge. A very recent comprehensive review by Ullah et al. systematically assessed clay–polymer nanocomposites for emerging contaminants removal and highlighted that the rational design of clay-based adsorbents with both high capacity and structural robustness remains an ongoing pursuit. The authors emphasized that integrating functional components (e.g., polymers, metal ions) into clay matrices often improves adsorption performance, yet the dynamic role of the contaminant itself in assisting the assembly process has rarely been explored [24]. Moving beyond conventional materials that rely predominantly on surface interactions, recent research strategies emphasize designing three-dimensional porous networks capable of actively trapping pollutants. In this context, gel-based systems emerge as a compelling paradigm. Their characteristic crosslinked, hydrated networks provide not only high surface area but also a confined microenvironment that can synergistically enhance pollutant retention through mechanisms such as adsorption, size exclusion, and hindered diffusion [25].

This gel-inspired paradigm directs attention toward natural materials that inherently possess or can be engineered to form such functional porous architectures [26]. Among them, fibrous clay minerals are promising inorganic precursors. Palygorskite, a natural hydrated magnesium aluminum silicate, is of particular interest due to its unique rod-like nanocrystalline structure [27]. In aqueous suspensions, these nanorods exhibit a pronounced tendency to intertwine and self-assemble into voluminous three-dimensional networks. This behavior originates from their high aspect ratio and surface interactions, which renders their colloidal state inherently gel-like [28]. This fundamental property suggests that palygorskite can be strategically engineered not merely as a particulate adsorbent but as the structural scaffold for a hybrid inorganic gel matrix.

However, the direct application of pristine palygorskite networks for the removal of anionic dyes such as Congo red faces two primary limitations. First, its negatively charged surface offers poor inherent affinity for anionic dyes [29]. Second, while a physical network exists, its stability and specific binding functionality are often inadequate for strong, irreversible sequestration. Traditional strategies to enhance the adsorption capacity of clay minerals, such as acid activation, heat treatment, and organic modification [30,31,32,33], have often overlooked a critical aspect: the active role of exchangeable and structural metal ions, especially in minerals with appreciable cation exchange capacity (CEC) [27,34]. For instance, thermal treatment can alter the palygorskite structure, affecting its CEC and potentially mobilizing metal ions from the octahedral layer, thereby influencing adsorption performance [34,35]. This highlights the complex synergy at the solid–liquid interface between the mineral framework, its ionic constituents, and pollutant molecules.

Inspired by these insights, we propose that integrating multivalent metal ions, specifically Fe(III), can simultaneously address the aforementioned limitations. Fe(III) is well-known for its strong affinity for oxygen-containing groups and its capacity to act as a cross-linker in polymeric networks [36]. We hypothesized that Fe(III) activation of palygorskite will serve a dual purpose: (i) introduce specific adsorption sites for anionic dyes via electrostatic and coordinative interactions, and (ii) more critically, modulate the inter-rod interactions (e.g., through charge screening or ionic bridging). This process was expected to reinforce the native gel-like network, guide its assembly into a more defined and stable porous aggregate, and ultimately create a synergistic trapping system where physical confinement within the network works in concert with specific chemisorption on Fe(III) sites.

Although iron-ion-modified clay minerals have been studied for pollutant removal, existing research predominantly focuses on inorganic anions (e.g., fluoride) and seldom provides a mechanistic analysis of the iron’s role in microstructural evolution or the long-term immobilization strength of the adsorbate [37,38,39,40]. The interaction of complex organic dyes like CR within such metal-ion-bridged gel networks is far more intricate and remains underexplored. A lone study on Fe-modified clinoptilolite for CR adsorption exists [40], but the profound structural differences between fibrous palygorskite and tectosilicate clinoptilolite necessitate a dedicated investigation to uncover unique structure–property relationships.

To address these challenges, in this work we propose a conceptually distinct “pollutant-induced gelation” strategy that fundamentally differs from conventional surface-functionalization approaches. Unlike previous modifications that relied solely on pre-introduced active sites, our method leverages the target pollutant (Congo red) itself as a synergistic structural promoter. Specifically, we fabricated Fe(III)-activated palygorskite aggregates as a model gel-inspired trapping system and to elucidate the mechanism behind the enhanced CR removal. We focused on understanding how the Fe(III) treatment transforms the nanorod dispersion into an effective porous network and how the resulting gel-like aggregate structure governs the adsorption capacity, kinetics, and, most distinctly, the immobilization (trapping) of the dye molecules.

2. Results and Discussion

2.1. Effect of pH

The adsorption capacity as a function of the initial solution pH is shown in Figure 1a. The adsorption capacity of both adsorbents decreases as the pH value of the dye solution increases. Specifically, the decrease is rapid in acidic systems and slows down when pH > 7. Under the same pH conditions, the adsorption capacity of the Fe(III)-activated sample is significantly greater than that of palygorskite, increasing by 95.4277% (Figure 1b). Its adsorption capacity is 1.953.77 times greater than that of palygorskite. This pH dependence of adsorption capacity can be attributed to the known effect of pH on the surface charge of clay minerals [41].

The zeta potential results (Figure 2) show that the isoelectric points of palygorskite and Fe(III)-palygorskite are 2.9 and 3.1, respectively. At pH > 4, both adsorbents carry net negative charges, which is electrostatically unfavorable for anionic CR adsorption [42]. Nevertheless, Fe(III)-palygorskite exhibits significantly higher adsorption capacity across all pH values, confirming that electrostatic attraction is not the dominant mechanism. Besides electrostatic interactions, surface complexation between Fe(III) sites and the sulfonate groups (-SO_3_^−^) of CR, hydrogen bonding with surface hydroxyl groups (-OH), and π–π stacking interactions between the aromatic rings of CR and the siloxane surface likely contribute to the adsorption process. Additionally, the pH values of the supernatant after adsorption of dyes on both palygorskite and the modified sample increased with the increase in the initial pH of the dyes, with the latter being generally lower than the former and weakly acidic (Figure S1). This is mainly attributed to the strong acidity of the ferric chloride solution, which contains a high concentration of hydrogen ions. While the iron ions act on palygorskite, the latter also undergoes a certain degree of acid activation, that is, hydrogen ions are exchanged for palygorskite crystals and are partially released into the solution during the adsorption of dyes [43]. CR is an anionic dye containing two sulfonate groups (-SO_3_^−^) and two amino groups (-NH_2_). Its pK_a_ is approximately 4.1. At pH < 4.1, the amino groups are protonated (-NH_3_^+^), reducing the overall negative charge of the dye. At pH > 4.1, CR exists predominantly in its deprotonated anionic form. The sharp decrease in adsorption capacity at pH > 7 is therefore attributed to: (i) increased electrostatic repulsion between the negatively charged adsorbent surface and the fully deprotonated CR anions; and (ii) competition between hydroxide ions (OH^-^) and CR sulfonate groups for Fe(III) binding sites. Although the maximum adsorption capacity was observed at pH 4–5, pH 6 was selected for further experiments to balance adsorption efficiency with practical applicability, as near-neutral conditions reduce chemical consumption and equipment corrosion risks in real wastewater treatment scenarios.

2.2. Adsorption Kinetics

Figure 3 shows the variation in adsorption capacity with respect to both contact time and temperature. As the contact time increases, the adsorption capacity of Fe(III)-activated palygorskite initially increases rapidly, then more slowly, reaching equilibrium after 240 min. The adsorption capacity decreases with increasing temperature, indicating that the adsorption of CR by the activated sample is an exothermic process.

The pseudo-first-order and pseudo-second-order adsorption kinetic models are two commonly used mathematical models to describe the adsorption rate of solutes on adsorbent surfaces, which are suitable for water treatment, environmental remediation and other fields [44]. The former is an empirical model describing the relationship between the adsorption rate and the adsorption capacity during the adsorption process. It assumes that the adsorption rate is proportional to the number of unoccupied adsorption sites and is applicable to processes dominated by physical adsorption or surface diffusion. The latter assumes that the adsorption rate is controlled by the chemical reaction between the active sites on the adsorbent surface and the adsorbate. This model is applicable to chemical adsorption processes, such as covalent bonding, ion exchange and complexation. The integral forms of the two models are:

$$[eqn]$$ $$[eqn]$$

In these equations, qt and qe (mg g^−1^) represent the adsorption capacity at time t and at equilibrium, respectively, while k1 (h^−1^) and k2 (g mg^−1^ h^−1^) are the rate constants for the pseudo-first-order and pseudo-second-order models. As summarized in Table 1, the pseudo-second-order model provided a markedly superior fit (higher R^2^) to the kinetic data compared to the pseudo-first-order model (Figure S2). This indicates that the adsorption process is best described by pseudo-second-order kinetics, suggesting a rate-limiting step governed by diffusion or chemical interaction [45]. Furthermore, the increase in the k2 value with rising temperature denotes an acceleration of the adsorption rate, facilitating faster attainment of equilibrium.

According to the Arrhenius equation, the activation energy of the adsorption process can be calculated.

$$[eqn]$$

where k denotes the rate constant; R is the molar gas constant (8.314 J mol^−1^ K^−1^); T is the absolute temperature (K); Ea is the activation energy of the adsorption process (kJ mol^−1^); and A is the preexponential factor. The pseudo-second-order kinetic rate constant, which best describes the adsorption behavior, is selected and plotted as lnk versus 1/T. The activation energy Ea of the adsorption process can be obtained from the slope, as shown in Figure 4. In general, the activation energy required for physical adsorption is small, typically no greater than 4.2 kJ mol^−1^, but if pore diffusion resistance is present, the apparent activation energy may be slightly higher (520 kJ mol^−1^); the activation energy for chemical adsorption, however, is 80400 kJ mol^−1^ [46]. In this study, the adsorption activation energy of CR on Fe(III)-activated palygorskite is 11.30 kJ mol^−1^, which falls within the hydrogen-bonding range (530 kJ mol^−1^) [47]. This suggests that hydrogen bonding plays a significant role in the adsorption process, while the absence of a higher activation energy (80400 kJ mol^−1^) rules out pure chemisorption. It also indicates the presence of pore diffusion resistance during the adsorption process, likely related to the nanopores and micropores developed in the palygorskite particles.

2.3. Adsorption Isotherms

The most common method for evaluating the adsorption capacity of adsorbents is to use adsorption isotherms to describe the entire adsorption process. Adsorption isotherms were generated by performing experiments at varying temperatures and initial CR concentrations (Figure 5). The data reveal a steady rise in adsorption capacity with increasing equilibrium dye concentration. Notably, no adsorption plateau is observed throughout the isotherm and its shape differs significantly from that of a typical Langmuir isotherm, suggesting that monolayer adsorption does not occur [48]. The observed decline in adsorption capacity with rising temperature confirms that the adsorption of CR onto the Fe(III)-activated palygorskite is an exothermic process. This is consistent with previous kinetic results. The equilibrium data were modeled against four classic adsorption isotherms: Langmuir, Freundlich, Dubinin–Radushkevich (D–R), and Temkin (Figure S3). The corresponding fitted parameters are compiled in Table 2. The linear formulas corresponding to the four models are as follows:

Langmuir equation:

$$[eqn]$$

Herein, qmax represents the theoretical maximum adsorption capacity, and b (L mg^−1^) is the Langmuir equilibrium constant. A higher value of b denotes a stronger affinity between the adsorbent and the adsorbate [49].

Freundlich equation:

$$[eqn]$$

where KF is the Freundlich constant (related to adsorption capacity), and n is a dimensionless exponent representing both adsorption intensity and favorability. It is generally stated that values of n > 1 reflect favorable adsorption conditions [50].

D-R equation:

$$[eqn]$$

β, a constant related to the mean free energy of adsorption (mol^2^ kJ^−2^); qm, the theoretical saturation capacity; and ε is the Polanyi potential, expressed as RTln(1 + (1/Ce)), with R representing the gas constant (J mol^−1^ K^−1^) and T the absolute temperature (K).

Temkin equation:

$$[eqn]$$

where B = RT/z, and z is the Temkin constant related to heat of sorption (kJ mol^−1^). A is the Temkin isotherm constant (L mg^−1^).

A comparison of the regression coefficients (R^2^) of various models shows that the Langmuir model has a poor fit, whereas the Freundlich, D-R and Temkin models all fit the isothermal adsorption process well. Among these models, the Temkin isotherms provided the best fit to the experimental data at all temperatures, with R^2^ values all above 0.97. The values of parameter n in the Freundlich model are all greater than 1, implying favorable adsorption of dye molecules by Fe(III)-activated palygorskite. The β value of the D-R model is much smaller than 1, indicating that the adsorbent has small micropores [51]. The average free energy of adsorption was calculated using the D-R model (Es= –1/(2β)^0.5^). The Es value reflects the energy required for ion exchange per mole of ions on the adsorbent surface and can be used to analyse adsorption types. Based on the magnitude of Es, the adsorption mechanism can be categorized as follows: physisorption for Es < 8 kJ mol^−1^; ion-exchange for 8 kJ mol^−1^ < Es< 16 kJ mol^−1^; and chemisorption for 20 kJ mol^−1^ < Es < 40 kJ mol^−1^ [52]. Table 2 shows Es values of 8.3~10.7 kJ mol^−1^ at three temperatures, indicating that the adsorption of CR by Fe^3+^ ions-treated palygorskite involves ion exchange. Since the adsorbent contains almost no exchangeable anions, this exchange suggests that the cations (e.g., K^+^, Mg^2+^, Ca^2+^, Fe^3+^) present in the adsorbent underwent a certain degree of cation exchange with Na^+^ in the CR molecules in the solution. As the temperature increases, the average free energy of adsorption decreases, indicating that this cation exchange is unfavourable at higher temperatures. Clearly, this mechanism will not contribute to the adsorption of anionic dyes such as CR. Given that the Es is not significantly higher than the upper limit of physical adsorption (8 kJ mol^−1^), we hypothesise that the adsorption mechanism of the adsorbent for CR is primarily driven by physical adsorption, including electrostatic interactions, hydrogen bonding, and van der Waals forces.

The Temkin constant A value decreased with increasing temperature. This suggests that low temperatures are more conducive to interaction between the adsorbate and the adsorbent. B is related to the heat of adsorption, and the B values at different temperatures are very small, indicating that the interaction between the adsorbent surface and the adsorbate is weak [53]. This is consistent with the conclusion of physical adsorption in the D-R model. Moreover, the Temkin isotherm model accurately describes adsorption behaviour, confirming significant interactions between CR dye molecules adsorbed on the modified palygorskite [54].

2.4. Adsorption Thermodynamics

Since the Temkin isotherm model provided the best fit of CR on modified palygorskite, the corresponding model parameter was selected to calculate the thermodynamic parameters. The constant A was converted to a dimensionless form to calculate Kc. This was achieved by scaling it by the molar mass of CR (696.66 g mol^−1^) and subsequently by 55.5 × 1000. The factor 55.5 accounts for the molar concentration of water (1000/18 mol L^−1^), establishing a rational basis for the thermodynamic cycle [55].

$$[eqn]$$ $$[eqn]$$ $$[eqn]$$

The enthalpy change (ΔH^0^) and entropy change (ΔS^0^) were determined from the slope and intercept of the linear plot of ln Kc against 1/T, respectively (Figure S4). Subsequently, the Gibbs free energy change (ΔG^0^) was calculated employing Equation 10. The results are presented in Table 3.

The negative ΔH^0^ value confirms the exothermic nature of the adsorption process, indicating that lower temperatures favor the uptake of CR, which aligns with the kinetic observations. Concurrently, the negative ΔS^0^ value indicates an increase in system order after adsorption. The calculated ΔG^0^ values were negative at all temperatures studied but became less negative with increasing temperature. This confirms the spontaneity of the process, although the driving force diminishes at higher temperatures. The magnitude of ΔG^0^ (between −20 and 0 kJ mol^−1^) typically corresponds to physisorption, whereas chemisorption is characterized by more negative values (typically −80 to −400 kJ mol^−1^) [56]. The ΔG^0^ values obtained in this study are higher than the typical range for physisorption yet are significantly less negative than those for chemisorption. This intermediate range implies that the adsorption is dominated by strong physical interactions such as hydrogen bonding and π–π stacking, rather than by weak van der Waals forces alone.

2.5. SEM, TEM and XRD Analysis

Figure 6 shows the SEM and TEM images of palygorskite, modified palygorskite, and a dye-loaded modified sample. Palygorskite clearly exhibits a typical elongated, rod-shaped crystal morphology with a smooth surface. Numerous rod-shaped crystals are stacked randomly, clearly distinguishable from each other (Figure 6a,d,g). After modification with iron ions, the nanorod morphology is maintained, and initial, loose nanorod associations are observed (Figure 6b,e,h), indicating the incipient formation of a three-dimensional network mediated by Fe(III) cross-linking. The most striking transformation occurred upon interaction with CR. TEM images (Figure 6c,f,i) reveal that the dye-loaded sample exhibits significantly larger and more coherent aggregates, reaching hundreds of nanometers in size. This marked increase in aggregate size and structural integrity provides direct visual evidence that CR molecules are not passive guests but active participants in the assembly process. The dye molecules appear to bridge and consolidate the Fe(III)-primed nanorod network, leading to the formation of a denser, more extensive gel-like scaffold perfectly suited for physical entrapment.

EDS surface scan results (Figure S5 and Table 4) suggest that the iron content of the palygorskite increased significantly following treatment with iron ions, while the magnesium and silicon content decreased, suggesting possible Fe incorporation into the structure, though it cannot distinguish between lattice substitution and surface deposition. Based on ionic radii considerations, Fe^3+^ may partially replace Mg^2+^ in octahedral sites, while surface-associated Fe species are also likely present. The aluminium content increased slightly. This may be related to its location at the center of the octahedral layer of palygorskite, making it less susceptible to attack [43]. At the same time, some iron ions may have entered the mineral’s lattice defects.

After CR adsorption, the EDS analysis (Table 4) revealed the appearance of carbon (24.46 ± 1.88%) and nitrogen (19.68 ± 2.66%), which are characteristic elements of Congo red molecules. These elements were not detected in either pristine or Fe(III)-activated palygorskite before adsorption. This provided direct elemental evidence for the successful uptake of CR onto the adsorbent. Concurrently, the atomic percentages of Mg, Al, Si, and Fe decrease significantly after adsorption, which is attributed to the surface coating effect of the adsorbed dye layer, as corroborated by TEM observations (Figure 6c,f,i). The Fe content decreased from 2.42% to 1.02%, suggesting that some surface-bound Fe(III) species may be partially shielded or exchanged during the adsorption process, although the majority remains fixed within the network (Table S1).

Although EDS is not usually a reliable method for quantitative analysis, we can still draw some conclusions from this result. Compared with palygorskite, the iron content of the sample activated by iron ions increased by 0.90%. The solid-to-liquid ratio in this adsorption experiment is 1:500. Assuming that all introduced iron ions entered the liquid phase during the dye removal process, the iron ion concentration in the supernatant should increase by approximately 18.0 mg L^−1^ compared to the untreated sample. During the liquid-phase adsorption process, clay minerals inevitably leach some metal ions, which can affect their ability to adsorb dyes [35]. Since, after FeCl_3_ treatment, the various metal ions on palygorskite are more or less replaced by iron ions, the leaching behavior of iron ions was therefore a critical factor governing the improved adsorption performance of the modified sample. The results of the Fe^3+^ ion leaching experiment showed that the iron ion concentration in the supernatant of the modified sample with adsorbed dye at 303.2 K was only 0.69 mg L^−1^ (Table S1), indicating that the majority of iron ions fixed in palygorskite were not leached during the adsorption process.

In order to elucidate the role of leached Fe(III) ions in the adsorption performance, different concentrations of iron ions were introduced into the palygorskite dye adsorption system. As the iron ion concentration increased, the adsorption capacity increased only slightly and remained far lower than that of iron ion-activated palygorskite (Figure 7). This indicates that the substantial enhancement of the modified sample’s dye adsorption capacity is not primarily caused by the aggregation and sedimentation of iron ions on the dye, but rather by the trapping effect of Fe(III)-activated palygorskite rod-like aggregates on the dye. This is evident in palygorskite nanorods that encapsulate a large number of CR aggregates during aggregation and sedimentation, as demonstrated by TEM results.

X-ray diffraction is an important method for analyzing the crystal structure of clay minerals and their modified samples. X-ray diffraction analysis was conducted on palygorskite clay and its modified and dye-loaded sample. The XRD pattern revealed that the primary component of the original sample was palygorskite, accompanied by trace amounts of montmorillonite, quartz, and dolomite (Figure 8). The strongest peak appeared at 8.41°, which is a characteristic diffraction peak of the (110) crystal plane of palygorskite. The symmetrically sharp peak shape indicates that the palygorskite crystal form is relatively complete and highly crystalline. After activation, the characteristic dolomite peak disappears due to the strong acidic environment of the ferric chloride solution. The (110) peak shifts to 8.36°, while the (200), (130) and (400) peaks shift to the right (Table S2). Previous EDS analysis results have confirmed that Fe ions partially replace Si and Mg in the palygorskite crystal structure. The atomic radius of iron is slightly larger than that of silicon, but significantly smaller than that of magnesium [57]. A characteristic diffraction peak shifts towards a smaller angle as the lattice constant increases. This is typically caused by the addition of atoms with a larger radius than the host atom. The opposite occurs when atoms with a smaller radius than the host atom are introduced [58]. The replacement of silicon and magnesium in palygorskite crystals with iron causes the aforementioned change in peak position. After adsorbing the dye, the intensity of the characteristic diffraction peaks on the modified palygorskite (110) crystal plane decreases significantly, while the intensity of the other characteristic peaks remains largely unchanged. This is because a large number of dye molecules coat the surface of the palygorskite nanorods, forming a coating layer of a certain thickness, as confirmed by TEM results. Additionally, a strong new peak appears at 14.38°, which is likely caused by CR due to the aggregation and subsequent crystallization of numerous CR molecules on the palygorskite crystals.

According to BET testing, the specific surface area of Fe(III)-activated palygorskite is 153.32 m^2^ g^−1^ (Figure S6), and its adsorption capacity for the dye at a pH of 6 is 163.35 mg g^−1^ (0.234 mmol g^−1^). This means that each gram of the sample can adsorb 1.41 × 10^23^ CR molecules. Based on these data, it was found that a single dye molecule occupies an area of just 0.70 nm^2^ on the adsorbent. However, since CR molecules are much larger than the palygorskite rod nanopores (0.37 × 0.64 nm), only around two-thirds of the original surface area is accessible. The calculated area per molecule (0.46 nm^2^) is significantly smaller than the molecular dimensions of CR. This implies that the adsorbed dye molecules are not lying flat but rather form three-dimensional aggregates on the sample surface. This is consistent with the conclusion obtained from TEM.

2.6. Studies of Interacting Mechanisms

In this study, the ferric chloride solution had a concentration of 1 mol L^−1^. Based on the Ksp of Fe(OH)3 at 25 °C (4.0 × 10^−38^), the pH of the ferric chloride solution was calculated to be 0.47. Therefore, in addition to ion exchange, it can be concluded that a certain degree of acid activation occurred during the treatment process. Acid activation can remove carbonate impurities from palygorskite clay, unblock nanopores and increase its specific surface area [59]. We treated palygorskite with hydrochloric acid (pH = 0.47) and found that the adsorption capacity was 85.7 mg g^−1^ under the same adsorption conditions. This was much lower than the 163.35 mg g^−1^ achieved with the sample activated by ferric chloride. This suggests that the acidification effect during ferric chloride treatment alone does not explain the significant improvement in the adsorption capacity of the adsorbent.

During the experiment, we observed that the activated sample remained fully suspended in the solution after adsorption, exhibiting significantly better suspension performance than the unactivated sample. This suggests the need for a systematic study of the colloidal stability of the samples, as illustrated in Figure 9. The colloidal stability of the modified palygorskite improved considerably after treatment with ferric chloride (Figure 9a), as demonstrated by the slower aggregation rate of the nanorods compared to palygorskite. This facilitates its interaction with more CR aggregates. However, the sodium ions derived from the CR molecules in the system significantly reduce the colloidal stability of the two adsorbents. Interestingly, when the two adsorbents interact with the CR solution separately, the stability of the palygorskite particles deteriorated, whereas the stability of the activated palygorskite improved. This suggests that colloidal stability may play an important role in the high adsorption capacity of the particles. As the temperature increases, the introduction of sodium ions benefits the colloidal stability of both the unmodified and the modified adsorbent particles, but the unmodified sample was affected much more than the modified sample (Figure 9b,c). Once CR molecules were introduced, the colloidal stability of the modified sample improved significantly and continued to improve with increasing temperature (Figure 9d). This indicates that the CR anion is the decisive factor in the stability of the Fe(III)-activated palygorskite suspension.

This interplay between colloidal stability and adsorption performance is intricately linked to the dynamic evolution of the nanorod network. The initial Fe(III) activation reduces inter-particle repulsion, enabling network precursor formation. The subsequent addition of CR, a large polyanionic dye, functions as a multivalent cross-linker and structural promoter. Specifically, the CR anion, with its multiple sulfonate and azo groups, bridges adjacent Fe(III)-primed nanorods through surface complexation between Fe(III) and -SO_3_^−^, and reinforces the network via π–π stacking between the aromatic backbones of adsorbed dye molecules. This charge neutralization and bridging effect accelerated the flocculation and densification of the nascent network into the large, coherent gel-like aggregates observed by TEM. This dye-induced network consolidation explains the concurrent decrease in suspension stability and the dramatic increase in adsorption capacity: the forming gel matrix efficiently traps and incorporates CR aggregates within its growing structure.

The interaction process between CR and activated palygorskite is proposed as follows: the two sodium ions in CR are first adsorbed onto the surface of palygorskite particles that have been treated with ferric chloride. This significantly weakens the surface electronegativity of the adsorbent. At the same time, some iron ions on the surface and pore ports of the palygorskite crystals are displaced and enter the solution, inducing the aggregation of nearby CR molecules. The decrease in surface electronegativity caused particle aggregation and sedimentation, and helps more CR aggregates approach the adsorbent surface. The combined effect of these two factors results in modified palygorskite having a much higher CR removal ability than palygorskite.

As the temperature increased, the CR removal capacity of the modified sample gradually decreased. This is because an increase in temperature reduces the tendency of particles to aggregate, thereby hindering the formation of rod-like crystal stacks capable of encapsulating CR. Meanwhile, it also impeded the self-association of dye molecules [60]. Additionally, the adsorption isotherm shows a gradual increase rather than plateauing (Figure 5). This is because, as the dye concentration increased, the number of CR molecules per unit volume also increased, as well as the amount of CR aggregates available for encapsulation. This situation is analogous to the solubilization of organic compounds by surfactant micelles.

Thus, the process is best described as a synergistic co-assembly. Fe(III) ions provide the primary cross-linking sites on the nanorods, while CR molecules act as secondary assembly promoters. Together, they drive the transformation of a palygorskite dispersion into a robust, three-dimensional hybrid gel network where the act of sequestration reinforces the very structure responsible for trapping. This self-reinforcing, gel-like trapping mechanism fundamentally distinguishes it from conventional adsorption on static surfaces.

To further clarify the interaction between CR and the activated sample, infrared spectroscopy was conducted and the results are shown in Figure 10. The palygorskite sample exhibits four peaks in the high-frequency region, positioned at 3614, 3583, 3552 and 3410 cm^−1^, respectively. These peaks are associated with the hydroxyl stretching vibration of various forms of water within the palygorskite crystals [61]. The absorption peak at 1655 cm^−1^ is attributed to the bending vibration of hydrogen bonds contained in zeolitic water [61]. The peak at 1196 cm^−1^ may be caused by the stretching vibration of the (Mg, Al)-O bond, forming a pair of clamp-shaped double peaks in the bands at 1030 cm^−1^ and 987 cm^−1^, which belong to the stretching vibrations of the Si-O-Si and Si-O bonds, respectively. These are also the main identification bands of palygorskite in this region [35]. Additionally, the peak at 1439 cm^−1^ indicates the presence of carbonate mineral impurities, as demonstrated by previous XRD spectra. After activation, the position and intensity of the aforementioned characteristic peaks remained unchanged, with only the peaks of the carbonate minerals disappearing. When CR was adsorbed onto the activated sample, new peaks at 3030, 2927, and 2853 cm^−1^ (C-H stretching) were assigned to the adsorbed CR molecules, confirming their intact incorporation into the gel network. Additionally, new peaks appeared in the intermediate frequency range of 1600–1300 cm^−1^ (Figure 10), which are clearly caused by CR molecules that have been adsorbed onto the adsorbent. Compared to CR, the positions of these new peaks remained largely unchanged. The C=C stretching vibration of the benzene ring occurs at 1600–1520 cm^−1^. The two peaks located at 1500–1480 cm^−1^ represent the N=N stretching vibration of the azo group in CR. The shift in the C-H in-plane bending peak from 1447 cm^−1^ to 1458 cm^−1^ was interpreted as evidence of π–π stacking between the aromatic rings of CR and the siloxane surface of palygorskite. In addition, the disappearance of the S=O asymmetric stretching peak at 1322 cm^−1^ (sulfonate group of CR) after adsorption was attributed to surface complexation between -SO_3_^−^ and Fe(III) sites. These FTIR observations collectively indicated that surface complexation (Fe-O-S), hydrogen bonding (N-H···O, O-H···O), and π–π stacking are jointly responsible for the strong retention of CR.

Previous results on adsorption kinetics and thermodynamics suggest that CR may have formed hydrogen bonds with iron-activated palygorskite. It is highly likely that the amino groups of CR formed hydrogen bonds with adsorbed water on the adsorbent’s surface, as well as with zeolitic water in the nanochannels. An increase in adsorption temperature weakened hydrogen bonding, hindering the adsorption process. Meanwhile, colloidal stability experiments have confirmed that increasing the temperature significantly enhances the stability of the activated sample in suspension, which is consistent with other studies on the temperature-dependent dispersion stability of colloidal particles [62]. The stability of the particle suspension improves because the formation of nanorod aggregates is unfavourable, which affects the effective encapsulation of dye molecules. Therefore, as the temperature increases, iron chloride-treated palygorskite’s ability to remove CR gradually weakens.

The pH of the solution was 6.0 before adsorption and 6.5 after. This indicates that the liquid phase primarily comprises deprotonated CR molecules. Solid-state UV results (Figure 11) showed that CR adsorbed onto Fe(III)-treated palygorskite exists in multiple forms, including neutral, ammonium and two azo forms [63]. Washing with deionized water did not alter the shape or position of the aforementioned characteristic peaks, only their intensity was reduced. This suggests that the process removed some of the outermost CR aggregates from the adsorbent, but at a rate of only 22.1%, which is far from exceeding the aforementioned adsorption capacity. This implies that a significant quantity of dye aggregates remain within the tightly packed, three-dimensional area formed by the palygorskite nanorods. It also indicates that Fe^3+^ activation of palygorskite can help form three-dimensional, nanorod-like aggregates that capture the target substance. The modified palygorskite prepared in this study exhibits a higher adsorption capacity for CR than most other modified palygorskite materials reported in the literature (Table 5), demonstrating its potential for use in the treatment of dye-contaminated wastewater.

Based on the experimental evidence and the schematic illustration in Figure 12, the synergistic adsorption and gelation mechanism of CR onto Fe(III)-activated palygorskite is proposed as a three-stage co-assembly process in which Fe(III) ions play a dual role throughout. In stage (a), Fe(III) ions anchored on the nanorod surfaces serve as both cross-linking nodes—neutralizing surface charges and bridging adjacent nanorods to form loose three-dimensional network precursors—and as specific adsorption sites that provide strong binding affinity for anionic groups via surface complexation. In stage (b), the introduced CR molecules acted as multivalent cross-linkers: their sulfonate groups (-SO_3_^−^) form surface complexes with Fe(III) sites (orange arrows), their amino groups participate in hydrogen bonding with surface hydroxyls and zeolitic water (green dashed lines), and the aromatic backbones of adjacent CR molecules undergo π–π stacking with each other and with the siloxane surface (purple dotted curves), thereby bridging neighboring nanorods and reinforcing the nascent network. In stage (c), these cumulative cooperative interactions drove the progressive consolidation of the nanorod network into a dense, three-dimensional gel-like architecture, within which CR aggregates are physically entrapped in the interstitial pores. This dye-induced gelation not only enhanced adsorption capacity by creating abundant confined spaces, but also markedly improved the structural integrity and settleability of the adsorbent, facilitating facile separation. In summary, Fe(III) fulfilled a dual role as both the architectural cross-linker that constructed the gel framework and the functional binding site that captured dye molecules, while the pollutant itself (CR) acts as a synergistic structural promoter, a concept we term “pollutant-induced gelation”.

The gel-like structure enhanced the mechanical stability and porosity of the aggregate, which improves dye entrapment and increases the accessible surface area for adsorption. This structure also aided in preventing adsorbent disintegration during regeneration cycles, thereby supporting reusability. While the gel-like structure enhanced adsorption, it may complicate CR desorption. Future studies will explore mild elution methods (e.g., alkaline or organic solutions). If regeneration is inefficient, the spent adsorbent can be solidified for safe disposal or thermally treated for potential energy recovery.

3. Conclusions

In this study, an Fe(III)-activated palygorskite was developed, exhibiting significantly enhanced efficacy for CR sequestration compared to the pristine clay. The adsorption capacity increased remarkably by 95.4–277% across a pH range of 4–10. The adsorption process followed pseudo-second-order kinetics and was determined to be spontaneous and exothermic based on thermodynamic analysis.

Beyond these performance metrics, the key advancement of this work lies in the elucidation of a novel synergistic assembly and trapping mechanism, moving beyond the paradigm of simple surface adsorption. Characterization, particularly TEM analysis, revealed that Fe(III) activation initiated the formation of loose, three-dimensional associations among palygorskite nanorods. Most importantly, the introduction of CR dye acted as a crucial trigger, dramatically promoting the consolidation and growth of these associations into large, coherent gel-like aggregates (hundreds of nanometers in size). This direct observation indicates a co-assembly process where the pollutant molecule actively participates in constructing the trapping matrix.

Therefore, the exceptional removal performance is attributed to a dynamic trapping mechanism within a hybrid gel network. Fe(III) ions serve as primary cross-linkers on the nanorod surfaces, while the anionic CR molecules function as secondary assembly promoters and structural consolidators. Together, they facilitate the formation of a robust, porous network that physically entraps dye aggregates within its interstitial spaces. This mechanism is corroborated by the low desorption efficiency and the evolution of colloidal stability during adsorption.

In conclusion, this work successfully fabricates a high-performance adsorbent and, more fundamentally, demonstrates a pollutant-induced gelation strategy using a modified nanoclays. It provides fresh insights into designing intelligent, gel-inspired hybrid materials where the target contaminant co-contributes to the formation of the stable matrix that ensures its own immobilization, offering a promising avenue for advanced wastewater treatment.

4. Materials and Methods

4.1. Materials

Palygorskite clay in powdered form (<50 μm) was sourced by Anhui Mingguang Rare Minerals Ltd. Co., Mingguang, China. Congo red (C.I. No. 22120) and iron(III) chloride (FeCl_3_, ≥99.9%) were purchased from Sigma-Aldrich (Shanghai, China). All other chemicals were analytical grade and used without further purification. Deionized water (resistivity: 18 MΩ·cm) was employed in all experiments.

4.2. Preparation of Fe(III)-Activated Palygorskite

To prepare the gel-inspired adsorbent, palygorskite (10.0 g) was dispersed in an aqueous FeCl_3_ solution (1 mol L^−1^, 100 mL). This suspension was then subjected to continuous magnetic agitation at 25 °C for 24 h to facilitate iron incorporation and potential nanorod cross-linking. After that, it was poured into a centrifuge tube for centrifugation using an H1850 high-speed centrifuge (Changsha Xiangyi Centrifuge Co., Ltd. Changsha, China) at 10,000 rpm for 10 min. The supernatant was decanted, and the solid was resuspended in deionized water. This mixture was agitated using a glass rod and subsequently isolated by centrifugation. The resulting product was dried at 80 °C to constant weight and finally ground into a fine powder for subsequent use (Figure S7).

4.3. Adsorption Study

All adsorption experiments were performed in batch mode under standardized conditions: 0.05 g of adsorbent and 25 mL of dye solution. The initial pH was adjusted as required with 0.1 M HCl or NaOH, using a calibrated pH meter (Mettler Toledo 320-S). After solid–liquid mixing, the suspension was placed in a thermostatic oscillator and shaken at 120 rpm and the desired temperature. The suspension was then centrifuged at 10,000 rpm for 10 min. The resulting supernatant after adsorption was measured using a 7200 ultraviolet spectrophotometer at a maximum wavelength of 498 nm to determine the dye concentration. The adsorption capacity of the adsorbent for CR was calculated using the following equation:

$$[eqn]$$

In the equation, qe (mg g^−1^) is the equilibrium adsorption capacity, Co and Ce (mg L^−1^) are the initial and equilibrium concentrations of the adsorbate, respectively, in which L is the volume (L) of the solution and m is the mass (g) of the adsorbent. Different initial CR concentrations were employed for the pH-dependent and kinetic studies according to their specific experimental objectives. A relatively high concentration (1 g L^−1^) was used in the pH experiments to ensure a readily measurable adsorption capacity across the entire pH range, thereby allowing reliable evaluation of pH effects. In contrast, a lower concentration (300 mg L^−1^) was selected for the kinetic studies to better capture the rate-limiting behavior and avoid premature equilibrium, which is consistent with common practice in adsorption kinetics investigations. All adsorption experiments were repeated three times and the data were presented with error bars in the figures.

4.4. Desorption Study

A batch method was employed for the desorption study. Briefly, the adsorbent (0.05 g) was equilibrated with 25 mL of dye solution (1 g L^−1^) at 30 °C for 24 h. Solid–liquid separation was achieved by centrifugation. The spent adsorbent was collected, washed with deionized water, and then mixed with 25 mL of fresh deionized water. Desorption was allowed to proceed at 30 °C for 24 h under agitation. The concentration of dye released into the supernatant (i.e., the residual CR concentration in the desorption solution at equilibrium) was measured spectrophotometrically after centrifugation.

The efficiency of desorption was evaluated using Equation (12):

$$[eqn]$$

In this equation, Edes is the desorption efficiency (%), Qdes is the mass of dye desorbed (mg), calculated from the residual dye concentration in the desorption solution, and Qads is the mass of dye originally adsorbed (mg). Reported values are the average of three independent replicates.

4.5. Characterization Techniques

The surface morphology of the samples was examined using a Hitachi S-4800 scanning electron microscope (SEM) (Hitachi High-Technologies Corporation, Tokyo, Japan). Prior to imaging, samples were sputter-coated with a thin gold film and analyzed at an acceleration voltage of 20 kV. Transmission electron microscopy (TEM) was conducted on a Thermo Scientific FEI/Talos L120C instrument (Thermo Fisher Scientific, Waltham, MA, USA) operated at 120 kV. Phase composition and crystal structure were analyzed by X-ray diffraction (XRD) on a Philips diffractometer (Philips, Amsterdam, Netherlands) with Cu Kα radiation (40 kV, 40 mA), scanning from 5° to 70° (2θ) with a step size of 0.03° and a counting time of 0.5 s per step. The specific surface area and pore structure of the modified palygorskite were determined from N_2_ adsorption–desorption isotherms measured at 77 K using a Micromeritics 3Flex analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA). Samples were degassed under vacuum at 200 °C for 6 h prior to analysis. Diffuse reflectance UV-Vis-NIR spectra were recorded on a Shimadzu S-4100 spectrophotometer (Shimadzu Corporation, Kyoto, Japan) over the range of 200–800 nm, using barium sulfate as a reflectance standard. Surface charge properties were assessed via zeta potential measurements performed with a Malvern Zetasizer Nano ZS (Malvern Panalytical Ltd, Malvern, UK). Fourier-transform infrared (FTIR) spectra were obtained in the range of 4000–400 cm^−1^ at a resolution of 4 cm^−1^ (32 scans per sample) using a Thermo Nicolet NEXUS TM spectrometer (Thermo Fisher Scientific, Madison, WI, USA) with the KBr pellet technique. Background scans were collected and automatically subtracted from sample spectra.

The concentration of Fe^3+^ ions in solution was quantified by inductively coupled plasma optical emission spectrometry (ICP-OES, PerkinElmer DV2100 series) (PerkinElmer Inc., Waltham, MA, USA). To prevent hydrolysis, the supernatant was acidified with a trace of hydrochloric acid before measurement.

Colloidal stability was evaluated by monitoring the time-dependent transmittance at 900 nm using a 7200 UV-Vis spectrophotometer (Shanghai Spectrum Instruments Co., Ltd., Shanghai, China). Briefly, suspensions were prepared by dispersing 50 mg of sample in 25 mL of various media (deionized water, dye solution, or salt solution) at 15,000 rpm for 3 min, then immediately transferred to a cuvette for transmittance recording.