Materials

DEX (molecular weight = 200,000–225,000 g mol^–1) was used as the base polymer for hydrogel synthesis and was purchased from Sigma-Aldrich. EPI, employed as the chemical crosslinking agent, was obtained from Sigma-Aldrich. N,N,N′,N′-Tetramethylethylenediamine (TEMED), used as an accelerator for the crosslinking reaction, and sodium hydroxide (NaOH), utilized to adjust and maintain alkaline reaction conditions, were purchased from Sigma-Aldrich. HNTs, used as inorganic reinforcing fillers for the preparation of nanocomposite hydrogels, were also obtained from Sigma-Aldrich. All chemicals were used as received without further purification.

Hydrogel synthesis

DEX-based hydrogels were synthesized via EPI-mediated crosslinking under alkaline conditions. Briefly, DEX (4 g) was dissolved in 25 mL of 2.8 M NaOH solution and stirred for 24 h to ensure complete dissolution and swelling of the polymer chains. For HNT-containing formulations, HNTs corresponding to 30 wt.% DEX were first dispersed in 4 mL of distilled water by stirring for 3 h and then added to the DEX solution. An additional 3 h stirring period was applied to promote homogeneous dispersion of HNTs within the polymer matrix.

After preparation of the DEX or DEX/HNT mixtures, EPI was added as the crosslinking agent at either 2 or 4 mL depending on the targeted crosslinking density. TEMED (2 mL) was then added to each formulation, and the reaction mixtures were stirred at 50–55°C until gelation occurred, typically within 15–20 min. Subsequent purification was carried out by immersing the hydrogels in distilled water for 24 h to remove unreacted species and residual reagents, followed by drying at room temperature, until constant weight was achieved. Formulation details of the prepared hydrogel samples are summarized in Table 1. The obtained hydrogels, including 2DEX, 2DEX-HNT, 4DEX and 4DEX-HNT, are shown in their synthesized and dried forms in Fig. 4.

Table 1.

Formulation details of DEX-based hydrogel systems.

Table 1 Long description

Formulations are listed by sample name with columns for DEX amount in grams, NaOH solution volume and molarity, HNT content in weight percent, and EPI amount in milliliters. Sample 2DEX has DEX amount 4 g, NaOH solution 25 mL at 2.8 M, HNT content shown as a dash, and EPI amount 2 mL. Sample 2DEX-HNT has DEX amount 4 g, NaOH solution 25 mL at 2.8 M, HNT content 30 wt.%, and EPI amount 2 mL. Sample 4DEX has DEX amount 4 g, NaOH solution 25 mL at 2.8 M, HNT content shown as a dash, and EPI amount 4 mL. Sample 4DEX-HNT has DEX amount 4 g, NaOH solution 25 mL at 2.8 M, HNT content 30 wt.%, and EPI amount 4 mL.

Figure 4.

(a) Synthesized and dried form of the low-crosslinked-density hydrogel (2DEX). (b) Synthesized and dried form of the low-crosslink-density DEX/HNT composite hydrogel (2DEX-HNT). (c) Synthesized and dried form of the high-crosslink-density hydrogel (4DEX). (d) Synthesized and dried form of the high-crosslink-density DEX/HNT composite hydrogel (4DEX-HNT).

Methylene blue adsorption experiments

Methylene blue adsorption experiments were conducted to evaluate the dye removal performance of the prepared hydrogel systems. A stock methylene blue solution was prepared by dissolving 1 g of dye in 1 L of distilled water. Standard solutions at concentrations of 10, 20, 30, 40 and 50 mg L^–1 were then prepared from this stock solution for calibration. Following 100-fold dilution to maintain absorbance within the linear range, absorbance measurements were recorded at 665 nm using a ultraviolet–visible (UV–Vis) spectrophotometer. A calibration curve was constructed in the form of y = ax + b, where y represents absorbance and x represents methylene blue concentration.

For adsorption experiments, dried hydrogel samples (2DEX, 2DEX-HNT, 4DEX and 4DEX-HNT, 50 mg each) were immersed in 25 mL of methylene blue solution with an initial concentration of 50 mg L^–1 and kept under shaking conditions for 24 h. Adsorption equilibrium was verified by monitoring residual methylene blue concentration at various contact times. Time-dependent measurements showed a progressive decrease in concentration during the early adsorption stage, followed by a plateau at later time points. Negligible variation between the final measurements confirmed that equilibrium was reached within 24 h.

After adsorption, aliquots withdrawn from each solution were diluted 100-fold with distilled water prior to analysis. Residual methylene blue concentrations were determined from the calibration curve, and the amount of adsorbed dye was calculated by comparing the initial and final concentrations.

Characterization and swelling studies

ATR-FTIR spectroscopy

The proposed crosslinking reaction mechanism between DEX and EPI under alkaline conditions, investigated by ATR-FTIR spectroscopy, is schematically illustrated in Fig. 5. In the presence of NaOH, hydroxyl groups of DEX are deprotonated to form alkoxide ions, which act as nucleophiles and attack the epoxide ring of EPI. This ring-opening reaction results in the formation of ether linkages, and subsequent intermolecular reactions lead to the formation of a three-dimensional crosslinked DEX network. Therefore, the formation of C–O–C ether bridges represents the key structural evidence expected to confirm the proposed mechanism.

Figure 5.

Crosslinking reaction of DEX with EPI under alkaline conditions.

Figure 5 Long description

Dextran is shown as a repeating sugar unit with multiple hydroxyl groups and a subscript n. A first reaction line shows dextran plus sodium hydroxide yields a dextran oxygen with a negative charge paired with sodium, plus water. A second reaction line shows this dextran oxygen with a negative charge paired with sodium plus epichlorohydrin, drawn with an epoxide ring and a chloromethyl group, yielding a dextran oxygen bonded to a three carbon chain that includes an opened epoxide and a chlorine substituent. A subsequent reaction line shows formation of a linkage between two dextran chains through an oxygen to carbon bond, with sodium chloride written as a byproduct. The final structure is labeled, crosslinked dextran and shows two dextran repeating units connected by an ether type oxygen bridge through a carbon chain. Reaction coefficients and physical state symbols are not shown.

The characteristic FTIR bands of DEX were clearly observed in all samples (Fig. 6). A broad absorption band centred at ∼3330–3340 cm^–1 corresponds to the stretching vibrations of hydroxyl (–OH) groups present in the DEX backbone. Following crosslinking, slight variations in the intensity and width of this band were observed, indicating partial consumption and involvement of hydroxyl groups in the crosslinking reaction, consistent with alkoxide formation and ether bond generation (Fig. 5). Bands observed at ∼2937 cm^–1 are attributed to aliphatic C–H stretching vibrations originating from both DEX and EPI segments.

Figure 6.

ATR-FTIR spectra of (a) low-crosslink-density DEX hydrogels (2DEX and 2DEX-HNT) and (b) high-crosslink-density DEX hydrogels (4DEX and 4DEX-HNT).

Figure 6 Long description

Four overlaid ATR FTIR transmittance spectra with a legend listing DEX, EPI, HNT and 2DEX HNT. The x-axis is labeled, Wavenumber (centimeter superscript minus 1), with tick labels at 4000, 3500, 3000, 2500, 2000, 1500, 1000 and 500. The y-axis is labeled, Transmittance (percent), with no numeric tick labels visible. Each spectrum contains multiple downward dips. Numeric annotations marking dip positions include 3337, 2937, 1648, 1450, 1154, 1021, 945, 894, 797, 694, 540 and 470. The image B showing four overlaid ATR FTIR transmittance spectra with a legend listing DEX, EPI, HNT and 4DEX HNT. The x-axis is labeled, Wavenumber (centimeter superscript minus 1), with tick labels at 4000, 3500, 3000, 2500, 2000, 1500, 1000 and 500. The y-axis is labeled, Transmittance (percent), with no numeric tick labels visible. Each spectrum contains multiple downward dips. Numeric annotations marking dip positions include 3330, 2937, 1640, 1450, 1077, 1033, 915, 797, 694, 540 and 470.

For the low-crosslinked DEX hydrogel (2DEX), the appearance of bands of increased intensity in the 1200–1000 cm^–1 region is associated with C–O and C–O–C stretching vibrations, indicating the formation of ether linkages between DEX chains via EPI (Fig. 6a). In particular, the band observed near 1020 cm^–1 is highly indicative of newly formed C–O–C bridges between glucose units, providing direct spectroscopic confirmation of the crosslinking mechanism (Fig. 5). The increased intensity of this band compared to that of pristine DEX reflects successful network formation. A weak band at ∼1750 cm^–1 may be attributed to carbonyl-containing structures formed during side-reactions under alkaline conditions or associated with residual species, although its contribution remains limited and does not dominate the reaction pathway. The band near 880–925 cm^–1 corresponds to vibrations of anomeric carbon atoms in the DEX structure, confirming the preservation of the polysaccharide backbone after chemical modification.

In addition, the band at ∼1440–1450 cm^–1 observed in some hydrogel formulations can be attributed primarily to CH₂ bending (scissoring) vibrations associated with the DEX backbone and EPI-derived segments. The relatively high intensity of this band in the crosslinked and HNT-containing hydrogels suggests changes in the local chemical environment of the polymer chains following crosslinking and halloysite incorporation. This observation further supports the successful structural modification of the DEX network.

Upon incorporation of HNTs (2DEX-HNT), additional spectral features appeared in the FTIR spectra (Fig. 6a). The bands observed at ∼1118–1120 cm^–1 are assigned to the Si–O–Si stretching vibrations of halloysite, confirming the successful incorporation of HNT into the polymer matrix. Furthermore, the deformation band near 916 cm^–1 corresponds to inner-surface hydroxyl groups of the HNTs. The overlap of DEX-derived C–O–C vibrations with the Si–O–Si bands in the 1200–1000 cm^–1 region suggests strong spectral superposition and indicates close interfacial contact between the inorganic nanotubes and the organic polymer network. Importantly, the presence of halloysite-specific bands without disruption of ether-related peaks confirms that HNT incorporation occurs without interfering with the fundamental crosslinking mechanism.

In the high-crosslinked hydrogels (4DEX), characteristic aliphatic C–H stretching bands observed at 2939 and 2877 cm^–1 and the C–O–C stretching band near 1020 cm^–1 confirm the presence of the DEX-based crosslinked network (Fig. 6b). Although direct comparison of band intensities between Fig. 6a and Fig. 6b is limited by spectral scaling and presentation, the retention of ether-related bands in the 1200–1000 cm^–1 region supports EPI-mediated crosslink formation. Therefore, the FTIR spectroscopy results indicate successful formation of crosslinked DEX networks without overinterpreting relative intensity differences between separately presented spectra.

The incorporation of HNT into the high-crosslinked system (4DEX-HNT; Fig. 6b) resulted in pronounced Si–O–Si bands at 1118–1120 cm^–1 while maintaining the characteristic DEX- and EPI-related vibrations. The simultaneous observation of C–O–C bands and halloysite-specific Si–O–Si vibrations confirms the formation of a chemically crosslinked DEX network reinforced with inorganic nanotubes.

Overall, the ATR-FTIR spectra (Fig. 6) provide strong experimental support for the EPI-mediated crosslinking mechanism (Fig. 5). Formation of ether bridges, the preservation of the DEX backbone and the successful incorporation of HNTs are clearly evidenced by the observed spectral features, validating the proposed structural model of the DEX/HNT nanocomposite hydrogels.

Scanning electron microscopy

SEM was employed to investigate the surface morphologies of the DEX-based hydrogels prepared with different crosslinking densities (2DEX and 4DEX) and their corresponding nanocomposite hydrogels containing 30 wt.% HNTs. SEM analysis was performed to evaluate the effects of crosslinking degree and HNT incorporation on surface homogeneity, porous architecture, morphological variations and matrix–filler interactions.

SEM images of the 2DEX sample (Fig. 7a) show a distinct porous morphology that was formed because of chemical crosslinking. EPI promoted the formation of crosslinks between DEX chains, leading to a porous network with a relatively uniform pore distribution. The observed structural regularity suggests effective network formation and overall structural integrity. No pronounced surface defects or large-scale roughness were detected; instead, only mild micro-scale roughness accompanied by a porous texture was observed, indicating that the 2DEX hydrogel exhibits a smooth and comparatively homogeneous surface.

SEM images of the 2DEX-HNT nanocomposite hydrogel (Fig. 7b) demonstrate that HNT addition induces pronounced changes in surface morphology. Incorporation of HNTs increased surface roughness and formed more complex microstructures. HNTs appeared to form localized aggregates within the matrix, creating micro-hierarchical arrangements. This morphological evolution suggests a reinforcing effect of the nanotubular filler and indicates increased structural robustness of the hydrogel network.

For the 4DEX hydrogel (Fig. 8a), a higher crosslinking density is expected due to the increased EPI content. Increasing the crosslinker amount led to the formation of a tighter and more ordered network structure, resulting in a more compact porous architecture (Fig. 8a). Compared to the 2DEX hydrogel, the 4DEX sample exhibited smaller pores that were distributed more uniformly. Increased crosslinking density is therefore associated with improved structural integrity and potentially greater mechanical stability. A slight increase in surface roughness was also observed, consistent with the formation of a denser polymer network.

Figure 7.

SEM images of (a) the 2DEX hydrogel and (b) the 2DEX-HNT hydrogel.

Figure 8.

SEM images of (a) the 4DEX hydrogel and (b) the 4DEX-HNT hydrogel.

SEM images of the 4DEX-HNT nanocomposite hydrogel (Fig. 8b) reveal that the combined effects of increased crosslinking density and HNT incorporation substantially reduced pore visibility and pore size at the surface. Increased surface roughness, attributed to the presence of HNTs, contributes to the formation of a more heterogeneous and complex microstructure. With intensified EPI-driven crosslinking, the polymeric network becomes more compact, a structural feature commonly associated with improved mechanical stability. Consequently, the synergistic effect of crosslinking and HNT reinforcement indicates that the 4DEX-HNT hydrogel exhibits a more stable and mechanically robust architecture compared to non-filled and/or less-crosslinked systems. The SEM images also suggest the presence of localized HNT agglomerations within the hydrogel matrix. Such aggregation behaviour is commonly observed in nanocomposite systems due to interparticle interactions between nanotubes. While partial agglomeration may reduce the effective dispersion of HNTs, it can also contribute to the formation of heterogeneous microdomains within the polymeric network. These microstructural heterogeneities may influence both mechanical stability and adsorption performance. From a mechanical perspective, HNT agglomerates can act as localized reinforcement domains, potentially enhancing stiffness in specific regions. From an adsorption standpoint, the presence of clustered nanotubes may create regions with increased local surface functionality, which can facilitate dye retention through combined surface adsorption and confinement effects. However, excessive aggregation might limit the accessibility of active sites and reduce the overall efficiency of the system. Therefore, the adsorption behaviour observed in the present study is probably governed by a balance between HNT dispersion and localized aggregation within the hydrogel network.

Nitrogen adsorption–desorption (BET) analysis

To further investigate the influence of HNT incorporation and crosslinking density on the porous characteristics of the DEX-based hydrogel systems, nitrogen adsorption–desorption analyses were performed. The nitrogen adsorption–desorption isotherms of pristine HNT, 2DEX-HNT and 4DEX-HNT are presented separately in Fig. 9a–c. Figure 10 presents the corresponding pore-size distribution curves, whereas specific surface area, pore-volume and pore-size parameters are summarized in Tables S4–S6.

Figure 9.

Nitrogen adsorption–desorption isotherms of (a) pristine HNT, (b) 2DEX-HNT nanocomposite hydrogel and (c) 4DEX-HNT nanocomposite hydrogel samples. STP = standard temperature and pressure.

Figure 9 Long description

Line graph in image A with the legend text, HNT. The horizontal axis label is, Relative Pressure, P over P subscript 0, with tick labels 0.0, 0.2, 0.4, 0.6, 0.8, 1.0. The vertical axis label is, Volume at STP, cc g superscript minus 1, with tick labels 0, 1, 2, 3, 4. Two plotted traces are shown with square markers. One trace rises from about 0.2 at relative pressure about 0.1 to about 1.9 at about 0.5, then to about 2.6 at about 0.9, then drops to about 2.0 near about 0.95, then rises to about 4.0 at 1.0. The other trace rises from about 0.1 at about 0.1 to about 0.4 at about 0.4, then to about 1.0 at about 0.5, then stays near about 1.0 to about 0.7, then rises to about 1.8 at about 0.8 and to about 2.0 at 1.0. Line graph in image B with the legend text, 2DEX HNT. The horizontal axis label is, Relative Pressure, P over P subscript 0, with tick labels 0.0, 0.2, 0.4, 0.6, 0.8, 1.0. The vertical axis label is, Volume at STP, cc g superscript minus 1, with tick labels 0, 1, 2, 3, 4, 5, 6. Two plotted traces are shown with circular markers. One trace rises from about 0.1 at relative pressure about 0.1 to about 1.0 at about 0.4, then to about 1.8 at about 0.7, then to about 2.0 at about 0.9, then rises to about 5.4 at 1.0. The other trace rises from about 0.0 at about 0.1 to about 0.6 at about 0.4, then to about 1.0 at about 0.7, then to about 1.6 at about 0.9 and to about 2.0 at 1.0. Line graph in image C with the legend text, 4DEX HNT. The horizontal axis label is, Relative Pressure, P over P subscript 0, with tick labels 0.0, 0.2, 0.4, 0.6, 0.8, 1.0. The vertical axis label is, Volume at STP, cc g superscript minus 1, with tick labels 0, 1, 2, 3, 4, 5. Two plotted traces are shown with triangle markers. One trace rises from about 0.1 at relative pressure about 0.1 to about 0.9 at about 0.3, then to about 1.6 at about 0.7, then to about 2.0 at about 0.9, then rises to about 4.7 at 1.0. The other trace rises from about 0.0 at about 0.1 to about 0.4 at about 0.3, then to about 0.9 at about 0.7, then to about 1.0 at about 0.9 and to about 1.8 at 1.0.

Figure 10.

Pore-size distribution curves of HNT, 2DEX-HNT and 4DEX-HNT samples determined from nitrogen adsorption–desorption analysis.

The adsorption–desorption isotherms obtained for all investigated samples displayed mesoporous characteristics with hysteresis behaviour, suggesting restricted pore structures and interconnected diffusion pathways within the hydrogel matrices. The pristine HNT displayed a typical adsorption–desorption profile associated with its tubular morphology and accessible lumen structure (Fig. 9a). In comparison, the 2DEX-HNT and 4DEX-HNT nanocomposite hydrogels (Fig. 9b,c) exhibited modified isotherm profiles, suggesting that incorporation of HNT into the crosslinked DEX network altered pore accessibility and nitrogen diffusion within the hydrogel architecture. The presence of hysteresis loops further indicates that nitrogen adsorption was influenced not only by external surface adsorption but also by diffusion limitations within partially accessible porous domains.

The BET specific surface areas of 2DEX-HNT and 4DEX-HNT were 1.572 and 1.618 m² g^–1, respectively. For comparison, pristine HNT exhibited a BET specific surface area of 1.586 m² g^–1. Hence, the HNT-containing hydrogel systems exhibited highly comparable accessible surface areas. Therefore, the incorporation of HNT into the DEX network did not affect the total BET specific surface area. However, notable differences were observed in pore-volume and pore-size distribution parameters, demonstrating the significant influence of crosslinking density on pore accessibility and internal network structure.

Pristine HNT showed relatively broad pore-width characteristics, with DH and BJH adsorption pore diameters centred at ∼31 Å. This behaviour is consistent with the intrinsic tubular morphology and accessible lumen structure of HNTs. In contrast, both 2DEX-HNT and 4DEX-HNT hydrogels showed reduced pore diameters, generally clustered at ∼22 Å, indicating partial confinement of porous domains within the polymeric network. Reduced pore accessibility after hydrogel formation suggests that DEX crosslinking partially covered the HNT surfaces or restricted access to nanotubular cavities.

Crosslinking density exerted a more evident influence in the highly crosslinked 4DEX-HNT system. Compared with 2DEX-HNT, the 4DEX-HNT hydrogel exhibited lower micropore area and lower cumulative pore-volume values as determined using the DR, HK and SF methods. This reduction indicates that increasing EPI content resulted in a denser and more compact polymeric network, limiting gas diffusion pathways and decreasing the accessibility of porous regions within the hydrogel matrix. These findings are consistent with the SEM observations, in which the highly crosslinked hydrogels displayed more compact and less visible porous architectures.

Furthermore, the relatively restricted pore accessibility observed in 4DEX-HNT may contribute to increased structural stability and stronger confinement of dye molecules during adsorption. This interpretation is consistent with the methylene blue adsorption results, for which the highly crosslinked HNT-containing hydrogel showed the greatest adsorption capacity despite its comparatively low swelling capacity. Therefore, the BET results suggest that adsorption performance in these systems is controlled not only by swelling behaviour, but also by the combined effects of pore architecture, diffusion pathways and matrix compactness.

Overall, the nitrogen adsorption–desorption analysis showed that HNT incorporation and crosslinking density influenced the porous architecture of DEX-based hydrogels. While the BET specific surface area values remained highly comparable, the pore-size and pore-volume parameters revealed differences in pore accessibility and network compactness. These findings provide additional insight into the structure–transport–adsorption relationships governing the performance of DEX/HNT nanocomposite hydrogels.

Swelling behaviour of the hydrogels

Hydrogels are polymeric materials capable of absorbing and retaining large amounts of water, which makes them highly relevant for a wide range of industrial and environmental applications (Kaplan & Güner, Reference Kaplan and Güner2000a; Guo et al., Reference Guo, Bae, Fang, Li, Zhao and Yu2020). The swelling ratio is a key parameter that reflects the water-uptake capacity of hydrogels and their response to environmental conditions (Kaplan & Güner, Reference Kaplan and Güner2000b; Ganji et al., Reference Ganji, Vasheghani and Vasheghani2010; Lv et al., Reference Lv, Wu and Shen2019).

In this study, the swelling behaviours of four different hydrogel formulations were investigated as a function of crosslinking density and HNT incorporation. The percentage swelling ratios of the hydrogels were calculated and presented graphically to compare the effects of crosslinking and HNT in addition to swelling performance.

Gravimetric swelling data recorded at various immersion times are provided in Table S1. These data provide a quantitative comparison of water absorption behaviour for all hydrogel formulations over extended swelling durations.

Figure 11a presents the time-dependent swelling behaviours of the low-crosslinked hydrogels (2DEX and 2DEX-HNT), and Fig. 11b shows the swelling profiles of the highly crosslinked hydrogels (4DEX and 4DEX-HNT).

Figure 11.

(a) Swelling behaviour (% swelling vs time) of the 2DEX and 2DEX-HNT hydrogels and (b) swelling behaviour (% swelling vs time) of the 4DEX and 4DEX-HNT hydrogels.

The swelling ratio of the low-crosslinked hydrogel (2DEX) was higher than that of the high-crosslinked hydrogel (4DEX), which can be attributed to the lower crosslinking density and higher chain mobility of the polymer network (Fig. 11). The maximum swelling ratios were ∼685% for 2DEX and ∼603% for 4DEX, indicating that increased crosslinking density restricted polymer chain expansion and reduced water-uptake capacity.

Comparison of 2DEX-HNT and 4DEX-HNT suggests that the influence of HNT incorporation on swelling was more evident in the less densely crosslinked matrix, resulting in higher overall water uptake for 2DEX-HNT. These findings indicate that crosslink density and inorganic filler incorporation significantly influence solvent transport behaviour by altering network tightness and chain mobility. Similar diffusion-controlled swelling behaviour in crosslinked polymer systems has been previously reported in the literature (Kaplan & Güner, Reference Kaplan and Güner2000a; Can et al., Reference Can, Kavlak, Kırcı and Güner2003).

Diffusion kinetics analysis

To further elucidate the swelling mechanism, diffusion kinetics were evaluated using the empirical power-law model applied to the initial swelling region, as given in Equation 1:

(1)\begin{equation}\frac{{M_t}}{{M_\infty }} = {\text{k}}{t^n}\end{equation}

where ${M_t}$ represents the amount of water absorbed at time $t$, ${M_\infty }$ corresponds to the equilibrium swelling amount, ${\text{k}}$ is the swelling rate constant and $n$ is the diffusion exponent that characterizes the solvent transport mechanism within the polymer network.

The diffusion exponent $(n)$ was obtained from the slope of the linear plot of ln $({M_t}/{M_\infty })$ vs ln $(t)$ using the initial 60% swelling data (Fig. 12).

Figure 12.

(a) Diffusion analysis of 2DEX and 2DEX-HNT hydrogels using ln $({M_t}/{M_\infty })$ vs ln $(t)$ plots. (b) Diffusion analysis of 4DEX and 4DEX-HNT hydrogels using ln $({M_t}/{M_\infty })$ vs ln $(t)$ plots.

Figure 12 Long description

A line graph with the title text, ln(t). The horizontal axis has tick labels 0, 1, 2, 3, 4, 5, 6, 7, 8, 9. The vertical axis label reads, ln(M subscript t over M subscript infinity). The vertical axis tick labels read 0.5, 0, minus 0.5, minus 1, minus 1.5, minus 2, minus 2.5, minus 3, minus 3.5. A legend lists, 2DEX, 2DEX-HNT, Linear fit (2DEX), Linear fit (2DEX-HNT). Two curved lines rise from about minus 3.2 near t equals 0 to about 0.1 near t equals 8. Two straight fitted lines run diagonally upward across the plot. Two text boxes near the bottom show line equations and R superscript 2 values; the text is too small to read accurately. The image B showing a line graph with the title text, ln(t). The horizontal axis has tick labels 0, 1, 2, 3, 4, 5, 6, 7, 8, 9. The vertical axis label reads, ln(M subscript t over M subscript infinity). The vertical axis tick labels read 1, 0.5, 0, minus 0.5, minus 1, minus 1.5, minus 2, minus 2.5, minus 3, minus 3.5. A legend lists, 4DEX, 4DEX-HNT, Linear fit (4DEX), Linear fit (4DEX-HNT). Two curved lines rise from about minus 3.2 near t equals 0 to about 0.2 near t equals 8. Two straight fitted lines run diagonally upward across the plot. Two text boxes near the bottom show line equations and R superscript 2 values; the text is too small to read accurately.

The diffusion exponent $(n)$ values obtained from the slope of the diffusion plots were 0.416 for 2DEX, 0.390 for 2DEX-HNT, 0.433 for 4DEX and 0.381 for 4DEX-HNT. The diffusion exponent $(n)$ was calculated from the slope of the linear plot of ln $({M_t}/{M_\infty })$ vs ln $(t)$ using the initial 60% swelling data. This approach allows determination of the dominant transport mechanism governing water uptake in crosslinked polymer networks (Kaplan & Güner, Reference Kaplan and Güner2000a). All $n$ values fall within the range of $n$ ≤ 0.5, indicating that the swelling behaviour of all hydrogel formulations follows Fickian diffusion (Kaplan & Güner, Reference Kaplan and Güner2000b). In this diffusion regime, water transport within the polymer network occurs faster than polymer chain relaxation. The calculated diffusion parameters are summarized in Table 2.

Table 2.

Diffusion kinetic parameters (n, k and R ²) obtained from power-law analysis of the swelling behaviour of DEX-based hydrogels.

Table 2 Long description

Diffusion kinetic parameters are listed for four samples, 2DEX, 2DEX-HNT, 4DEX, and 4DEX-HNT, with columns for n, k, R2, and diffusion type. For 2DEX, n is 0.416, k is 0.042, and R2 is 0.976, with diffusion type, Fickian. For 2DEX-HNT, n is 0.390, k is 0.050, and R2 is 0.978, with diffusion type, Fickian. For 4DEX, n is 0.433, k is 0.038, and R2 is 0.971, with diffusion type, Fickian. For 4DEX-HNT, n is 0.381, k is 0.055, and R2 is 0.977, with diffusion type, Fickian.

The coefficients of determination (R ² = 0.971–0.978) indicate that the power-law model adequately describes the early-stage swelling behaviour of the prepared hydrogels. In the absence of HNT, increasing crosslinking density resulted in a slight increase in the diffusion exponent, from 0.416 for 2DEX to 0.433 for 4DEX. By contrast, in the presence of HNT, increasing crosslinking density led to a slight decrease in $n$, from 0.390 for 2DEX-HNT to 0.381 for 4DEX-HNT. Moreover, the lower $n$ values observed for HNT-containing hydrogels compared with their corresponding non-filled systems suggest that HNTs increase diffusion tortuosity within the polymer matrix. Comparable diffusion-controlled swelling behaviour in crosslinked polymer systems has been widely reported in the literature (Kaplan & Güner, Reference Kaplan and Güner2000a; Can et al., Reference Can, Kavlak, Kırcı and Güner2003).

The observed Fickian diffusion behaviour indicates that water transport within the prepared hydrogels was predominantly governed by concentration-driven diffusion rather than polymer chain relaxation. This finding is consistent with the compact crosslinked network structure formed through EPI-mediated crosslinking. The lower $n$ values obtained for the HNT-containing systems further suggest that HNTs increased the tortuosity of diffusion pathways within the hydrogel matrix. Accordingly, the diffusion behaviour in these hydrogels appears to be influenced by the combined effects of network architecture and HNT incorporation.

Methylene blue adsorption performance

The methylene blue adsorption performance of the prepared hydrogel systems was evaluated to investigate the effects of crosslinking density and HNT incorporation on dye removal efficiency. Removal efficiency (%) was additionally calculated to provide a clearer comparison of dye adsorption performance among the hydrogel systems. The adsorption results are summarized in Table 3, including the final methylene blue concentration, the amount of adsorbed dye and the removal efficiency after 24 h of contact time.

Table 3.

Methylene blue adsorption performance of DEX-based hydrogels after 24 h.

Table 3 Long description

Measurements are reported for four samples, 2DEX, 2DEX-HNT, 4DEX, and 4DEX-HNT, across absorbance, final methylene blue concentration in milligrams per liter, amount of adsorbed dye in milligrams, and removal efficiency in percent. For 2DEX, absorbance is 0.050, final concentration 0.383, adsorbed dye 0.293, and removal efficiency 99.2. For 2DEX-HNT, absorbance is 0.025, final concentration 0.199, adsorbed dye 0.756, and removal efficiency 99.6. For 4DEX, absorbance is 0.048, final concentration 0.368, adsorbed dye 0.330, and removal efficiency 99.2. For 4DEX-HNT, absorbance is 0.021, final concentration 0.169, adsorbed dye 0.830, and removal efficiency 99.6.

Removal efficiency for methylene blue was calculated according to Equation 2:

(2)\begin{equation}{\text{Removal efficiency }}\left( \% \right){\text{ }} = {\text{ }}\left( {\left( {{C_0} - {C_{\text{e}}}} \right)/{C_0}} \right){\text{ }} \times {\text{ }}100\end{equation}

where C ₀ is the initial methylene blue concentration and C _e is the equilibrium methylene blue concentration after 24 h of adsorption.

The HNT-containing hydrogels exhibited higher methylene blue adsorption capacities than the corresponding non-filled hydrogels (Table 3). Among all samples, 4DEX-HNT showed the highest adsorbed dye amount (0.830 mg), followed by 2DEX-HNT (0.756 mg). By contrast, the non-filled hydrogels showed lower adsorption values, with 4DEX and 2DEX adsorbing 0.330 and 0.293 mg of methylene blue, respectively. These results indicate that HNT incorporation plays an important role in improving the adsorption performance of DEX-based hydrogels.

Increased dye retention in HNT-containing hydrogels may arise from the hollow tubular morphology and dual-surface chemistry of HNTs, which provide additional interaction sites for cationic methylene blue molecules. Electrostatic interactions, hydrogen bonding and physical confinement within the nanotubular structure may contribute to dye retention within the hydrogel matrix.

Interestingly, although lower-crosslinked hydrogels exhibited higher swelling capacities, the highly crosslinked HNT-containing hydrogel showed the highest methylene blue adsorption performance. This result suggests that adsorption performance was not governed solely by swelling capacity. Instead, the combined effects of network compactness, pore architecture, diffusion pathways and HNT-derived interaction sites played a decisive role in dye retention. Therefore, the superior performance of 4DEX-HNT can be attributed to an optimized balance between crosslinking density and HNT incorporation.

Overall, these findings demonstrate that the adsorption behaviour of DEX/HNT nanocomposite hydrogels is controlled by the combined influence of crosslinking density and inorganic filler incorporation. While crosslinking density regulates network compactness and solvent transport, HNTs contribute additional interaction sites and nanotubular domains for methylene blue retention. This structure–property–performance relationship highlights the potential of DEX/HNT nanocomposite hydrogels as tuneable materials for aqueous dye remediation.