The Synthesis of Superabsorbent Hydrogels Containing Naturally Derived Components and Their Use for the Environmental Cleanup of Dye Pollutants

Abstract

The utilization of naturally derived sustainable components is an increasing trend in polymer and materials science and technology. This paper reports on the synthesis of superabsorbent hydrogels (SAHs) containing acrylic acid (AA), styrene (St, 10 wt.% of AA) and poly(ethylene glycol diacrylate) (PEGDA) using one-pot radical suspension copolymerization. Cassava starch (CS, 1 wt.% of AA) was added to the copolymerization mixture as a natural hydrophilic gel modifier. Cardanol acrylate (CA) was used as a natural alternative to PEGDA, and its effect on SAH properties (gel content, swelling rate, swelling capacity, morphology, thermal stability, rheology and dye absorption) was investigated. The swelling capacity reached its maximum at 63,000% for the gel containing 50% of CA. CA also significantly increased SAH rigidity, as revealed by rheology tests that confirmed the gel-like behavior of all networks as well. The amphiphilic gels showed a remarkable binding capacity for the selected dyes, with 38.7 mg/g for crystal violet, 41.5 mg/g for Acridine Orange, 47.4 mg/g for Methylene Blue and 33.5 mg/g for Auramine Orange, indicating a promising application potential for the environmental cleanup of aqueous waste.

1. Introduction

Hydrogels are natural or synthetic networks composed of chemically or physically crosslinked polymers capable of swelling and retaining a large amount of water in their structure [1]. They represent a very important class of materials due to their usefulness in numerous applications, including contact lenses [2], drug delivery [3,4,5,6,7], wound dressing [8], tissue engineering [9], supercapacitors [1], soil hydration [10], hygiene products [11], heavy metal removal [12], and dye absorption [13,14,15,16]. Superabsorbent hydrogels (SAHs) were first introduced in 1978 in Japan from the crosslinking of starch and poly (acrylic acid) and were used as personal hygiene products for women [17]. They were further improved to be used as baby diapers in Germany and France in the 80s [17]. Most commercially available SAHs are made from petroleum-based vinyl monomers, which are challenging to degrade and pose a significant environmental threat. Recently, the market has been transferring to more bio-based alternatives to chemicals derived from fossil sources [18].

Starch is an inexpensive and renewable material often used in bio-based SAHs [19]. It could be derived from cassava (Manihot esculenta), a shrub considered as a reliable source since it can grow in harsh natural conditions [20]. The plant contains very high amounts of starch which contribute to about 70% of its dry matter [20]. Cassava starch is an excellent candidate for hydrogel production, owing to its higher solubility in water compared to other tuber starches, and in distinction to cereal starches the gel has better stability [20].

Cardanol [21] was recently considered to be a bio-based sustainable alternative to some typical crosslinkers for AA-based hydrogels, such as N,N-methylene-bis-acrylamide (MBA) and poly (ethylene glycol) diacrylate (PEGDA). It is the byproduct of the cashew industry, obtained as the major component of the cashew nut shell liquid (CNSL), a natural resin contained in the honeycomb-like shell of cashew nuts. Cardanol consists of a phenol substituted in the meta position with unsaturated C₁₅ linear alkyl chains averaging two double bonds per molecule [21]. Its use as a building block of naturally derived polymers is increasingly investigated [22,23,24]. It is also used in several applications, including plasticizers [25,26,27], coating crosslinkers [28,29,30], and surfactants [31,32]. Cardanol acrylate (CA) is obtained by attaching acrylate moieties to the cardanol molecule. This modification allows cardanol to self-polymerize and form oligomers [33] or be utilized with other vinyl monomers. Hu et al. [34] used CA as a reactive diluent in the synthesis of UV-curable polyurethane acrylate resins. They found that the addition of CA increased C=C conversion and improved thermal stability, viscosity, shrinkage, and hardness, among other properties [34].

An environmental concern that SAHs help tackle efficiently is dye removal from aqueous waste. Every year, about 800,000 tons of dyes are released in the environment [35]. They come essentially from effluents of many industries, such as textile, pulp and paper, food and beverages, and the plastic industry [36]. These dyes cause not only aesthetic problems but are also serious threats to human and aquatic health [37]. They increase chemical demand in oxygen and decrease light penetration, which hinders the photosynthesis process under water [38]. These dyes also show carcinogenic effects. For example, Auramine Orange (AO), widely used in the textile and paper industry, is among dyes identified as carcinogenic by the International Agency for Cancer Research [39]. Crystal violet (CV) is also highly utilized in industry and biomedical studies [40,41] and is shown to have cytogenic, genetic, mutagenic and carcinogenic effects [42]. Methylene Blue (MB) is highly used in the textile, cosmetic [43], and paper industry [44]. It persists in water and does not decompose under light.

Super absorbent hydrogels are an interesting alternative to many previously used materials for dye removal from effluent waters [45]. They are also the focus of this study, where SAHs were prepared using AA, styrene, and cassava starch, with CA used as a partial replacement for PEGDA. Acrylic acid imparts hydrophilicity and improves water absorption and swelling, while styrene introduces hydrophobic domains that reinforce the mechanical stability of the hydrogel structure and hydrophobic binding [46]. Cassava starch enhances the hydrogel’s environmental sustainability. Besides sustainability, CA brings unique binding attributes to amphiphilic networks due to its aromatic rings and long alkyl chain. Emulsion polymerization was chosen as the synthesis method to ensure uniform incorporation of hydrophilic and hydrophobic components. The aim of this study is to show how components such as cassava starch and styrene as well as partial replacement of PEGDA by CA affect gel formation, morphology, thermal, swelling, and rheological properties of the hydrogels. Their dye-binding capabilities are also explored.

2. Materials and Methods

2.1. Materials

Starch was extracted from cassava tubers sourced from Côte d’Ivoire. Cardanol (Cardolite, Harrisburg, PA, USA) was used as received. Styrene (St, 99%), 2,2′-Azobis(isobutyronitrile) (AIBN, 98%), and Acryloyl Chloride (99%) were supplied by Millipore Sigma (St. Louis, MO, USA) and were used without further purification. Triethylamine (TCI America, Portland, OR, USA), chloroform (J.T. Baker, Phillipsburg, NJ, USA), methanol (Burdick and Jackson, Muskegon, MI, USA), acetone (J.T. Baker, Phillipsburg, NJ, USA), diethyl ether (J.T. Baker, Phillipsburg, NJ, USA), sodium dodecyl sulfate (SDS, Sigma Aldrich, Milwaukee, WI, USA), potassium persulfate (KPS, Beantown Chemical, Hudson, NH, USA), acetic acid (Fisher Chemical, Pittsburg, PA, USA), sodium acetate anhydrous (Fischer Scientific, Pittsburg, PA, USA), NaCl (Macron, Philipsburg, NJ, USA), NaOH (Macron, Philipsburg, NJ, USA), Na₂CO₃ (Fisher Science Education, Nazareth, PA, USA), NaHCO₃ (Fischer Science Education, Nazareth, PA, USA), PEGDA (M_n = 250 Da, Aldrich, Milwaukee, WI, USA), and HCl (EMD Millipore Corporation, Billerica, MA, USA) were used as received. The following dyes (all from Aldrich, Milwaukee, WI, USA) were used in the binding experiments without further purification: Acridine Orange (AcrO, dye content ~90%), Auramine Orange (AO, dye content ~85%), Bromophenol Blue (BPB, dye content ~95%), crystal violet (CV, dye content ~98%), Methylene Blue (MB, dye content ~98%), and Methyl Orange (MO, dye content ~93%).

2.2. Methods

2.2.1. Synthesis of Cardanol Acrylate

Cardanol acrylate was synthesized according to the procedure shown in Scheme 1.

In a two-neck round bottom flask equipped with a reflux condenser, 30 g of cardanol (0.1 mol) were dissolved in 100 mL of acetone, and 14.08 mL of triethylamine (10.12 g, 0.1 mol) were added as an acid neutralizer. The mixture was cooled in an ice bath to 0 °C, and 8.41 mL of acryloyl chloride (9.05 g, 0.1 mol) was cautiously added dropwise through the neck to the reaction mixture. The reaction was carried out for 16 h at 55 °C under reflux. The reaction mixture was added dropwise to 100 mL of diethyl ether to precipitate triethylamine hydrochloride, a salt formed by the reaction between triethylamine and the hydrochloric acid produced during the reaction. The mixture was then rinsed with sodium carbonate and DI water until neutral, vacuum filtered, dehydrated with sodium sulfate, and concentrated using a rotary evaporator. The product was a yellow-orange liquid. The yield was 87.43%.

ATR-FTIR (ν, cm⁻¹): 3450 (w, residual OH stretch), 1746 (s, C=O stretch), 1635 (w, C=C stretch), 1401 (s, H-C=C bend), 1230 (s, C(=O)-O stretch), 1025 (m, O-C-C stretch), 983 (m, =C-H out-of-plane bend).

¹H NMR (600 MHz, CDCl₃): δ 7.29 (t, 1H, Ar-H), 7.14 (d, 1H, Ar-H) 7.07 (s, 1H, Ar-H); 6.61 (dd, J = 17.38 Hz, 1H, =CH₂); 6.33 (dd, J = 27.77 Hz, 1H, =CH-), 6.00 (dd, J = 10.46 Hz, =CH₂), (0.8–5.8 ppm, cardanol signals).

¹³C NMR (151 MHz, CDCl₃): δ 164.76 (C=O), 150.69 (Ar-C), 144.78 (Ar-C), 132.46 (=CH-), 129.23 (=CH₂), 128.19 (Ar-C), 121.47 (Ar-C), 118.77 (Ar-C).

2.2.2. Synthesis of Hydrogels

The emulsion polymerization was carried out using the one-pot technique. The organic phase contained 8 mg of AIBN (0.049 mmol), 0 or 1.1 mL of styrene (co)monomer (1 g, 9.6 mmol), 3 mL of toluene, and 0–8 mg of CA/PEGDA mixtures, which were first loaded in a pre-dried 50 mL test tube, capped with a rubber septum and purged with nitrogen. For the aqueous phase, 100 mg of starch was gelatinized in 5 mL of DI water at 90 °C. A total of 9.52 mL of acrylic acid (10 g, 138.8 mmol) was partially neutralized (70% neutralized) with NaOH (6 M, 16.2 mL). The starch solution and acrylic acid were mixed and sonicated at room temperature for 5 min. Then, SDS (80 mg, 0.406 mmol) was added, followed by the addition of potassium persulfate (72 mg, 0.266 mmol); sonicated for 3 min; and thoroughly mixed until full dissolution. This aqueous phase was then added with a syringe to the test tube containing the organic phase, immediately forming an emulsion. The mixtures in the sealed test tubes were vortexed for 1 min each and heated in an oil bath at 60 °C. Polymerization occurs within 20 min. The products were washed with 100 mL of methanol, followed by immersion in 3 DCM portions (100 mL each) over three days to remove all non-crosslinked products—after 24 h the initial 100 mL DCM portion was removed, and fresh DCM was added. The DCM fractions were combined and concentrated using a rotavapor.

2.2.3. Fourier Transform Infrared Spectroscopy (FTIR)

FTIR analysis was performed in Attenuated Total Reflectance (ATR) mode using a Perkin Elmer Frontier instrument (PerkinElmer, Shelton, CT, USA). All samples were air- or vacuum-oven-dried and analyzed in the medium infrared region (4000–550 cm⁻¹) at 32 scans. The transmittance spectra were converted to absorbance spectra with baseline correction and normalization through the instrument software.

2.2.4. Nuclear Magnetic Resonance (NMR)

¹H- and ¹³C NMR analyses were performed on a Bruker 600 MHz instrument (Bruker Corporation, Billerica, MA, USA) at 16 scans for ¹H and 256 scans for ¹³C in deuterated chloroform (CDCl₃) and DMSO-d₆ at room temperature. Chemical shifts were referenced to the residual peaks of CHCl₃ or DMSO at 7.26 ppm and 2.5 ppm for ¹H NMR and 77 ppm for ¹³C NMR using CDCl₃.

2.2.5. Gel Content

The DCM fractions at the end of the reaction were concentrated using a rotary evaporator and used to quantify the soluble polymer fraction and the gel fraction.

The soluble polymer fraction, $P_{sf}$, is obtained using the following formula:

$$P_{sf}(\%)={\displaystyle \frac{M_s-M_e}{M_s}}\times 100$$

(1)

where $M_s$ is the mass of the dry hydrogel after synthesis, and $M_e$ is the mass of the hydrogel after the extraction of the soluble polymer fraction and subsequent drying.

The gel fraction, which is the complementary percentage, is given by the following equation:

$$Gel (\%)=100- P_{sf}$$

(2)

2.2.6. Scanning Electron Microscopy

SEM imaging was performed using a JEOL IT 100 instrument (JEOL USA, Inc.; Peabody, MA, USA). Samples were sputter-coated with 22.5 nm Au/Pd to improve conductivity before imaging. Secondary electrons were detected in high vacuum at an acceleration voltage of 10 kV, an aperture size of 2, and a probe current of 50.

2.2.7. Swelling Rate

The swelling rate is the profile of the swelling capacity of a hydrogel against time [47,48]. To measure the swelling rate, approximately 100 mg of hydrogel samples were weighed and immersed in 100 mL of DI water. The samples were then removed from the water and gently wiped with lint-free paper, and the weight was recorded after 5, 10, 15, 30, 45, 60, 90, 120, 180, 300, 600 and 1440 min (24 h). The measurements were done in triplicates and averaged. The swelling capacity is given by Equation (3).

$$S(\%)={\displaystyle \frac{{(W}_s-W_d)}{W_d}}\times 100$$

(3)

where $S$ is the swelling capacity, $W_d$ is the weight of the dried hydrogel, and $W_s$ is the weight of the swelled hydrogel. The swelling-rate values are then fitted into a Voigt model [49] given by Equation (4).

$$S_t=S_e(1-e^{\raisebox{1ex}{$-t$}\!\left/ \!\raisebox{-1ex}{$r$}\right.})$$

(4)

where $t$ is the swelling time, $S_t$ is the swelling capacity at time $t$, $S_e$ is the equilibrium swelling capacity, and $r$ the rate parameter (time required to reach 0.63 of the equilibrium swelling) [50].

2.2.8. Maximum Swelling Capacity and Absorbance Under Load

Hydrogel samples of approximately 100 mg were immersed in 100 mL of deionized water and swelled for 24 h. After swelling, hydrogels were wiped gently with a lint-free paper and weighed to determine the maximum swelling capacity. For the quantification of the absorbance under load (AUL), the swelled samples were loaded in a system consisting of a custom 3D-printed PLA apparatus composed of an outer and inner cylinder supported by a stainless-steel spacer to let the excess water drip (Figure 1). After the hydrogel was loaded, a brass weight of 500 g was placed on top of the inner cylinder to allow the liquid to drip, the weight of the sample was recorded, and AUL was calculated. Both swelling capacity and AUL were all calculated using the previously mentioned Equation (3).

2.2.9. Salt and pH Sensitivity

Salt sensitivity was measured under the same conditions as the maximum swelling capacity (100 mg of hydrogel, 24 h) in different solutions. Hydrogels were swelled in a 0.9% NaCl solution for the determination of salt sensitivity. For pH sensitivity, buffer solutions of sodium acetate/acetic acid (pH = 5.05) and sodium carbonate/bicarbonate (pH 9.17) were used. All measurements were done in triplicates and averaged.

2.2.10. UV-Vis Spectroscopy

UV-Vis analyses were performed on a Shimadzu UV-1900i instrument (Shimadzu Scientific Instruments, Columbia, MD, USA) in 1 cm quartz cuvettes. The measurements were conducted with aqueous dye solutions at room temperature.

2.2.11. Dye Binding

Dry pieces of the PEGDA50%/CA50% hydrogel, weighing about 10 mg each, were immersed in 50 mL dye solutions in water at a concentration of 10 mg/L. The UV-Vis spectra of the solutions were recorded between 800 and 350 nm before and after immersion of the hydrogel for 24 h. The residual dye concentrations of the water solutions were calculated using the Bouguer–Lambert–Beer equation (Beer’s law) using the molar extinction coefficients of the dyes used in the binding experiments. They were established by UV-Vis spectroscopy of aqueous solutions at different concentrations (R > 0.99; Figure S1). Using the absorbance values from the UV-Vis measurements, Figure S2, the equilibrium dye uptake, $q_e$, was calculated by Equation (5).

$$q_e={\displaystyle \frac{{(C}_0-C_e)\times M}{m}}\times V$$

(5)

where $C_0$ and $C_e$ are the molar concentrations of the dye solutions before and after absorption by the hydrogel respectively. $M$ is the molar mass of the dye, $m$ is the mass of the dry hydrogel used, and $V$ is the volume of the dye solution.

2.2.12. Thermogravimetric Analysis

TGA measurements were carried out using the high-resolution TGA instrument Discovery 550 (TA Instruments, New Castle, DE, USA). Approximately 10 mg of the sample was loaded into a platinum pan and heated at room temperature to 600 °C. Onset and endset temperatures as well as maximum decomposition temperatures were obtained from the TGA and the derivative thermogravimetry (DTG) curves.

2.2.13. Differential Scanning Calorimetry (DSC)

DSC analysis was carried out using the calorimeter TA instrument Discovery 250 (TA Instruments, New Castle, DE, USA). The sample and the reference pan were heated from 20 °C to 150 °C at a rate of 10 °C/min, kept at an isotherm for 5 min, then cooled down to 0 °C at a rate of 10 °C/min, kept at an isotherm for 5 min, and finally heated again to 150 °C at a rate of 10 °C/min.

2.2.14. Rheology

Rheological properties were measured using the Discovery HR-3 rheometer (TA Instruments, New Castle, DE, USA) with a temperature-controlled lower Peltier plate geometry of 20 mm. Hydrogels with a diameter of 20 mm and thickness 5000 ± 300 μm were cut after DCM extraction and used for analysis. The measurements were performed on DCM-swollen hydrogels for practical reasons—to afford a uniform specimen geometry across networks and prevent slippage on the Peltier plate (Figure S3). DCM swelling does not expand the gels as much as the water but softens them enough to allow for a uniform cut, whereas water-swollen gels swell extensively (see Figure S4, for example) and cannot be uniformly cut. Prior to oscillatory frequency sweep experiments, oscillatory strain sweep measurements were first conducted to determine the linear viscoelastic region of the networks. The frequency sweep analyses were performed at 25 °C in the range of 0.01–250 rad/s. Experiments were carried out in triplicates, and the average value is reported.

3. Results and Discussion

3.1. Synthesis of Cardanol Acrylate

Cardanol acrylate (CA) was successfully synthesized via esterification of the cardanol phenolic hydroxyl group with acryloyl chloride in acetone under anhydrous conditions, using triethylamine as a stoichiometric organic base and producing a yellow-orange liquid with an 87.4% yield. In this reaction, triethylamine deprotonates the phenol to generate a more reactive phenoxide nucleophile while simultaneously scavenging the HCl formed during the acyl substitution as triethylammonium chloride, preventing acid-catalyzed side reactions and driving the equilibrium toward full conversion [34]. The use of an ice-bath during acryloyl chloride addition, followed by reflux at 55 °C for 16 h, ensured controlled exotherm management during the exothermic acylation step and complete consumption of the phenolic hydroxyl group, as confirmed by the near-total disappearance of the phenolic O–H stretch at ~3350 cm⁻¹ in the FTIR spectrum (Section 3.1.1) and by the disappearance of the phenolic proton in the ¹H NMR spectrum (Section 3.1.2). The purification protocol involving precipitation of triethylammonium chloride in diethyl ether, followed by alkaline washing with Na₂CO₃ and drying over Na₂SO₄, effectively removed ionic byproducts and residual acidity, as evidenced by the absence of any ammonium salt signals in the ¹H NMR spectrum of the purified product. Importantly, no radical inhibitor was required during the reaction, as the moderate reflux temperature (55 °C) under inert atmosphere was insufficient to trigger premature self-polymerization of the acrylate moiety. The dual reactive character of CA, bearing both an acrylate group available for free-radical copolymerization and the additional double bonds of the C₁₅ alkyl side chain capable of secondary crosslinking, makes it a particularly versatile bio-based crosslinker candidate, as discussed extensively by Caillol et al. [21,22,24]. These results are further confirmed by the following detailed FTIR and NMR characterizations.

3.1.1. FTIR of Cardanol Acrylate

Figure 2 presents the transmittance spectra of cardanol (black) and cardanol acrylate (orange). The broad peak at 3350 cm⁻¹, corresponding to intermolecular hydrogen bonds of the phenolic proton in cardanol, nearly disappears after the reaction, indicating successful substitution by the acrylate group. The strong peak at 1746 cm⁻¹ in the cardanol acrylate spectrum corresponds to the acrylate carbonyl bond. The small bump at 1635 cm⁻¹ is caused by the stretching vibration of the C=C bond in the acrylate group. The strong peak at 1401 cm⁻¹ is due to the bending vibration of the H-C=C bond in the acrylate group. The peak at 1230 cm⁻¹ suggests the presence of C(=O)-O bonds.

The ester O-C-C bond exhibits a strong signal at 1025 cm⁻¹. The peak at 983 cm⁻¹ indicates the presence of an out-of-plane bending vibration in the =C-H bond. These results are consistent with those reported by Hu et al. [34].

3.1.2. ¹H NMR of Cardanol Acrylate

Figure 3 shows the proton NMR spectra of cardanol and cardanol acrylate (CA). Differences are observed in the regions corresponding to the acrylate double bond and the aromatic ring. The phenolic proton 1 disappears due to substitution by the acrylate group.

The singlet and doublet at 6.65 ppm, representing protons a and a’; the doublet at 6.75 ppm, corresponding to proton b; and the triplet at 7.14 ppm, representing proton c in the cardanol structure, are all shifted downfield, appearing at 6.96 ppm (protons d and d’), 7.07 ppm (e), and 7.29 ppm (f). These shifts are attributed to the inductive effect of the additional oxygen atom from the acrylate group. Acrylate double-bond protons are identified based on coupling-constant differences. Proton g (trans to i and cis to h) exhibits the highest coupling constant (J = 27.77 Hz) and appears as a doublet of doublets centered at 6.33 ppm. Proton h (geminal to i and cis to g) has the smallest coupling constant (J = 10.46 Hz) and appears at 6.00 ppm, while proton i (geminal to h and trans to g) has a moderate coupling constant (J = 17.38 Hz) and appears at 6.61 ppm. These results are in accordance with those reported by Hu et al. [34].

3.1.3. ¹³C NMR of Cardanol Acrylate

Figure 4 shows the carbon NMR spectrum of cardanol (blue) and cardanol acrylate (red). In the cardanol spectrum, saturated carbon signals appear between 14.24 ppm and 35.95 ppm and remain unaffected by acrylate substitution. Unsaturated aliphatic and aromatic ring carbons appear between 112.61 and 155.55 ppm. Upon substitution of the phenolic proton with the acrylate group, new peaks emerge between 112.58 and 164.76 ppm in the cardanol acrylate spectrum. Notable changes include three new peaks from acrylate carbons (g, h, and i) at 164.76 ppm, 132.46 ppm, and 129.23 ppm, respectively. The aromatic ring carbons shift: a from 155.55 ppm to 150.69 ppm (a’), b from 115.43 ppm to 118.77 ppm (b’), c from 112.61 ppm to 121.47 ppm (c’), d from 145.08 ppm to 144.78 ppm (d’), e from 129.52 ppm to 128.19 ppm (e’), and f from 121.13 ppm to 126.09 ppm (f’). These results also align with those previously reported by Hu et al. [34].

3.2. Synthesis of Hydrogels

Superabsorbent hydrogels based on partially neutralized acrylic acid (70%), styrene, and cassava starch were successfully prepared via oil-in-water suspension polymerization using a dual initiator system (AIBN in the organic phase and KPS in the aqueous phase) and SDS as a surfactant.

All formulations yielded reproducibly stable, coherent gel materials in yields ranging from ~91.3% (CA, 100%;

Table 1. Effects of crosslinking conditions on hydrogel yield, polymer soluble fraction (P_sf) and related gel content.

Hydrogel	PEGDA (mg)	CA (mg)	Yield (%)	${\mathit{P}}_{\mathit{s}\mathit{f}}\mathbf{(}\mathbf{\%}\mathbf{)}$	Gel (%)
Without starch	8 ± 0.1	0	93.63 ± 1.98	$0.40\pm 0.01$	99.60 ± 0.01
Without styrene	8 ± 0.1	0	94.91 ± 1.84	$0.51\pm 0.08$	99.49 ± 0.08
PEGDA, 100%	8 ± 0.1	0	91.70 ± 3.16	$0.51\pm 0.27$	99.49 ± 0.27
CA, 25%	6 ± 0.1	2 ± 0.1	91.89 ± 1.29	$1.29\pm 0.19$	98.71 ± 0.19
CA, 50%	4 ± 0.1	4 ± 0.1	94.50 ± 1.41	$1.55\pm 0.08$	98.45 ± 0.08
CA, 75%	2 ± 0.1	6 ± 0.1	95.08 ± 1.68	$0.33\pm 0.23$	99.67 ± 0.23
CA, 100%	0	8 ± 0.1	91.26 ± 2.25	$0.21\pm 0.19$	99.79 ± 0.19

) to ~95.1% (CA, 75%; Table 1), confirming the efficiency of emulsion polymerization as a method for incorporating both hydrophilic and hydrophobic components into a single crosslinked architecture.

The use of an oil-in-water emulsion approach allowed for the simultaneous polymerization of the water-soluble acrylic acid phase and the toluene-dissolved styrene/crosslinker phase, producing a phase-segregated but covalently interconnected amphiphilic gel, consistent with the semi-interpenetrating network (semi-IPN) concept described in the recent literature [51,52]. The system is referred to as a semi-IPN system since starch is not independently crosslinked. However, limited grafting of acrylic acid onto starch backbone chains via KPS-mediated hydrogen abstraction cannot be excluded under these conditions, consistent with observations reported by Ismail et al. [52]. The partial neutralization of acrylic acid to 70% with NaOH prior to polymerization is a well-established strategy in SAH synthesis that balances electrostatic repulsion within the growing network and promotes increased swelling while preventing excessive chain mobility that would otherwise reduce gel content [52,53]. Gelatinization of cassava starch at 90 °C before its incorporation facilitated the complete disruption of crystalline amylopectin domains and ensured homogeneous dispersion in the aqueous phase, which is essential to avoid phase separation during polymerization and to enable physical entrapment within the crosslinked network [54]. The gel content of all formulations exceeded 98% (Table 1), demonstrating a quantitative incorporation of both comonomers and crosslinkers into the insoluble/swellable network, irrespective of the PEGDA/CA ratio used. Furthermore, the hydrogels with a higher content of CA (75% and 100%) showed nearly total crosslinking, which demonstrates the efficiency of CA as a standalone crosslinker. The ¹H NMR analysis of methanol and dichloromethane (DCM) extracts, Figure 5 and Figure 6, confirmed that the extractable fractions consisted only of acrylic acid oligomers, residual SDS, and linear polystyrene, with no detectable crosslinker signal, indicating complete incorporation of both PEGDA and CA into the network. The formation of a stable emulsion within seconds of combining the organic and aqueous phases, followed by the onset of gelation within 20 min at 60 °C, is consistent with the previously reported rapid kinetics of persulfate-initiated radical polymerization of acrylic acid at moderately elevated temperatures [52] and with the accelerated polymerization behavior in emulsion systems described by Kovačič and Silverstein [51]. These results collectively confirm the successful formation of well-defined, highly crosslinked semi-IPN systems. The NMR analyses of the extracts gave further evidence of the incorporation of each component.

3.2.1. ¹H NMR of Methanol Extracts

Figure 5 shows the proton NMR spectra of methanol extracts, which were obtained after the first 24 h treatment by soaking the hydrogels formed.

The spectra show signals of acrylic acid (+) and SDS (*), as confirmed by signal overlaps in the Supporting Information ¹H NMR spectra (Figures S5 and S6). This suggests that only unreacted acrylic acid and SDS are released in the methanolic fractions. A further investigation was carried out after three days of DCM extraction.

3.2.2. ¹H NMR of DCM Extracts

Figure 6 shows the DCM extracts which were obtained after three additional days of immersion of the hydrogels in fresh DCM portions. These extracts are composed of acrylic acid (+) (Figure S5), SDS (*) (Figure S6) and linear polystyrene (▫) at 6.3–7.2 ppm and 1–2.5 ppm (Figure S7). The samples without starch did not release any non-crosslinked poly(styrene), which indicates that the absence of starch may enhance styrene incorporation into the network. In contrast, the absence of styrene shows prominent unreacted acrylic acid peaks, which might be interpreted as an enhanced styrene/acrylic acid copolymerization tendency. No crosslinker signal (PEGDA or CA) was identified in the spectra, confirming the full incorporation of the crosslinkers into the hydrogel networks.

3.3. Gel Content

Table 1 shows the gel yield, the soluble polymer fraction ($P_{sf}$), and the gel content. All formulations showed a high gel content between 98 and 100%. The gel content remains very high (99.41–99.76%) for the first three hydrogels crosslinked with PEGDA only (without starch, without styrene, and 100% PEGDA). It then decreases slightly to ~98.5% and ~98.4% for the hydrogels crosslinked with mixtures of CA and PEGDA (25% CA and 50% CA) with further increases to nearly 100% when the CA content increased (75% CA and 100% CA). The decreasing gel content at 25% and 50% of CA in the mixture could be attributed to steric effects and differences in reactivities between the two crosslinkers. The CA structure is composed of a phenyl-substituted acrylate group with alkenes in its long C₁₅ side chain (Scheme 1). The phenyl group has a resonance effect that can enhance the reactivity of CA compared to PEGDA. And this mismatch in reactivity, together with the steric differences, could cause a minor loss in crosslinking efficiency at certain ratios.

3.4. Scanning Electron Microscopy

SEM images of the hydrogels are shown in Figure 7, revealing their morphology. All hydrogels exhibit a homogeneous surface and highly porous inner structure. Images at a higher magnification (500×) are provided in the Supporting Information (Figure S8). It was seen that macropores with diameters of several hundred microns were randomly distributed in the hydrogel. Conversely, micropores of ≤10 microns were uniformly distributed in the networks (Figure 7). The pore-formation mechanism can be attributed to two parallel processes. The formation of macropores, as seen in Figure 7 and Figure S8, can be caused by gas evolution during the AIBN decomposition at the initiation stage [55]. Pore sizes of this origin can vary between 1 and 3000 μm [56]. The micropore formation, however, could be attributed to emulsion templating. In this process, droplets of the dispersed phase act as templates during copolymerization [53]. The (co)monomers (co)polymerize around these droplets, forming a crosslinked network. After that, the droplets are removed by solvent washing and drying, leaving behind voids (pores) corresponding to the droplet size, with diameters ranging from 1 to 30 μm [53]. Previous studies [52] reported similar porous structures in hydrogels synthesized via emulsion polymerization of acrylic acid, poly(styrene), and starch reinforced with bentonite.

3.5. Swelling Rate

One of the most important properties of superabsorbent polymers is their swelling rate, which shows how quickly the hydrogel can absorb and retain water in its structure [56]. To compare swelling rates, the values are fitted to a Voigt model [49,50], and the rate parameter, which is the value at which the swelling ratio reaches 63% of the hydrogel’s maximum swelling capacity, was used as a baseline for a comparison of the swelling rates. Figure 8a shows the swelling profile of all the hydrogels. Figure 8b–h show that after 2 h, the Voigt model describes the hydrogels swelling behavior well. The slow swelling rate in the initial period marks the penetration of the first wave of water molecules and the hydration of the amphiphilic segments in the networks that expand and open the space for the rapid water absorption in the second stage. All hydrogels show a strong correlation, with R² > 0.93, meaning that 93% of the values can be fitted in the model. It was observed that the equilibrium swelling capacity is smaller for the hydrogels without starch and without styrene. The other networks produced with a mixture of crosslinkers (25% CA, 50% CA, and 75% CA) have higher equilibrium swelling capacity values, $S_e$, than the hydrogels crosslinked with pure PEGDA and CA (Table S1).

The rate parameter (r) values (Table S1) are lined up in the following increasing order: without styrene < without starch < 100% PEGDA < 50% CA < 100% CA < 75% CA < 25% CA. The rate parameters show that the absence of styrene and starch causes hydrogel to reach swell equilibrium faster. This could be due to the relative hydrophobicity of both materials compared to acrylic acid. The other hydrogels show no significant differences.

3.6. Maximum Swelling Capacity and Absorbance Under Load (AUL)

Figure 9 shows the maximum swelling capacity (blue bars) and the absorbance under load (orange bars). Swelling capacity was measured as the amount of water retained in the hydrogel structure after 24 h. The swelling capacity increases from 53,000% for 100% PEGDA to 63,000% for 50% CA. It then decreases back to 55,000% for 100% CA. This notable similarity in the swelling capacity of 100% PEGDA and 100% CA deserves further discussion. Most probably, two processes are at play here: (1) The PEG segment is short but hydrophilic and adds the poly(acrylic) segment in sufficient swelling. (2) CA is hydrophobic but notably longer than PEG and generates longer intercrosslink segments upon crosslinking. Wider spacers between crosslinks (pores) would match the swelling of 100% PEGDA despite increased hydrophobicity. It is well known that a higher crosslinking density is negatively correlated with swelling capacity [57]. This means that differences in reactivity between CA and PEGDA would promote a looser network for mixtures of the two crosslinkers, thus enhancing their swelling capacity. A possible synergistic interaction between CA and PEGDA cannot be excluded as well. The synergy between the two in the mixed-composition hydrogels is based on the hydrophilic contribution of the PEG units and the spatial crosslink extension by the CA blocks. The hydrogels without starch and styrene show notably lower swelling capacities (i.e., higher crosslinking density) compared to the other hydrogels. Most probably, starch, being inactive in radical polymerization, acts as an inert spacer between the radically formed crosslinks. In addition, styrene could be involved in several reaction pathways, none of them producing crosslinks, homopolymerization, copolymerization with acrylic acid or copolymerization with CA, forming linear polymers or copolymers that would increase the intrinsic hydrophobicity of the semi-IPNs with a concomitant reduction in their swelling capacity. Absorbance under load (AUL) measures a hydrogel’s ability to retain absorbed fluid under applied pressure. AUL is considered a measure of the mechanical strength of a hydrogel, since logically, stronger hydrogels can withstand more pressure [58]. In this study, when 500 g of load (339.5 N/m²) was applied to the hydrogels, the load-free swelling capacity decreased only slightly across the entire line of networks, with the loss in swelling being between 5 and 11% of the original swelling capacity of the hydrogels. This shows that all hydrogels are mechanically strong regardless of their degree of crosslinking.

3.7. Salt and pH Sensitivity of Hydrogels

All hydrogels showed similar responses to changes in the medium—saline, basic or acidic. In all cases, a loss in swelling capacity was observed: ~87.7% in a saline environment, ~83.9% at pH 5 and ~79.9% at pH 9 (Figure 9 and Figure 10). These changes can be attributed to the balance of electrostatic interactions, osmotic pressure, and ion-exchange dynamics within the amphiphilic gels [59,60]. When the polymer network absorbs pure water, a fraction of its acidic functional groups most probably undergoes partial dissociation, releasing hydrogen ions (H⁺) into the interior solvent while generating conjugate anions (COO⁻) that remain covalently bound to the network [61]. The partial neutralization of the hydrogel (70%) also introduces Na⁺ counterions that further enhance swelling. When the surrounding solution consists of water molecules and different ions, the networks behave differently, showing that the swelling behavior of the gel is highly dependent on the ionic composition of the external solution.

The saline solution contains Na⁺ and Cl⁻ ions, the acidic medium (pH 5) has an excess of H⁺ and negative counterions, while the basic solution (pH 9) has an excess of HO⁻ and positive counterions. Diffusion of these charged species in the interior of the networks leads to changes in the ionic equilibrium between the fixed negative charges in the hydrogel and the positive charges in the solution, known as the Donnan equilibrium [62]. The resulting partial neutralization of charges leads to reduced swelling, which is clearly seen comparing the data in Figure 9 and Figure 10.

3.8. Thermogravimetric Analysis

TGA experiments revealed a slight weight loss between 100 and 150 °C, probably due to the removal of molecularly bound residual water [63] and two major degradation onsets for all hydrogels (Figure 11).

They occur in the range of 301–319 °C (5–11% weight loss) and 400–405 °C (31–36% weight loss) (Figure S9). During the first degradation stage, the decomposition of non-crosslinked starch most probably occurred (the starch used in this study thermally breaks down at ~318 °C; Figure S10). However, it represents only 1% relative to the mass of acrylic acid contained in the semi-IPN and could not cause the entire percentage loss observed in this temperature interval. Acrylic acid and its derivatives (homopolymers, copolymers and crosslinking segments) are the major components of all hydrogels under this investigation, and they are probably the major causes for the two major decompositions (Figure 11). This assumption is supported by previously published data [64]. The lack of styrene in one of the hydrogels does not markedly affect the decomposition profile compared with the 100% PEGDA hydrogel (Figure 11). Poly(styrene) has a maximum temperature of decomposition positioned between 390 and 420 °C [65], coinciding with the temperature of the second decomposition event in Figure 11. Due to the small styrene/acrylic acid ratio (1:10) it may not be sufficiently high to significantly affect the thermogram.

3.9. Differential Scanning Calorimetry

Figure 12 shows the DSC thermograms of the PEGDA/CA-crosslinked hydrogels, with no observable thermal transitions between 0 and 150 °C, an absence that could be attributed to the highly crosslinked nature of the networks.

T_g values of poly(styrene), poly(acrylic acid) and their copolymers are at or below 126 °C [66]. Crosslinking of the copolymers generally increases T_g, but it generally remains below 150 °C [67]. The absence of a detectable glass transition suggests a highly restricted segmental motion in the interpenetrating (co)polymer chains and crosslinked segments.

3.10. Rheology

The effect of CA addition on the viscoelastic properties (storage modulus and loss modulus) of the PEGDA/CA networks was investigated through rheological measurements (Figure 13). While the analysis was performed on DCM-swollen gels, their rheological response would still be influenced by their composition. Similar to other hydrogels, they do not respond instantaneously to an applied force; instead, their response depends on time [68,69].

This means that when a force is applied, part of the energy is stored reversibly; this is the elastic component known as storage modulus [70]. Some of this energy is dissipated as heat; this is the viscous component known as loss modulus [71]. A preliminary strain sweep is necessary to identify the linear viscoelastic region at which frequency sweep measurements should be conducted [72]. This region was identified at 0.5% strain (Figure S11). For all hydrogels, the storage modulus is parallel and higher to the loss modulus. This is the sign of a predominantly elastic gel behavior [72]. The storage modulus of most hydrogels that contain CA is significantly higher than that of the pure PEGDA hydrogel, revealing that the addition of cardanol increases the rigidity of the hydrogel. This increase in rigidity could be attributed to several factors—the benzene ring in CA and the multiple double bonds in its side chain that might increase crosslinking density. The loss modulus quantifies the viscous response, measuring the energy dissipated as heat during deformation [73]. The results show a similar trend, with loss-modulus values increasing with CA content compared to the 100% PEGDA gel. It would be reasonable to assume that the increased presence of long CA chains positively affects the fluid-like characteristics of the networks containing this compound.

3.11. Dye Binding

One advantage of hydrogels over traditional adsorbents, such as activated carbon, is their swelling capacity, which increases surface contact and enhances dye-absorption efficiency. The 50% CA hydrogel was used for all dye-absorption measurements because of its highest swelling capacity and AUL. Figure 14 illustrates dye-absorption results, expressed as both percentages and mg of dye per mg of gel.

The differently charged dyes can interact with the hydrogel through electrostatic attraction/repulsion with the acrylate groups/anions and π-π interactions with the aromatic rings from styrene and CA. The difference in absorption behavior can be explained by the differences in dye structures and their charges (Figure 15). Cationic dyes (Acridine Orange, Auramine Orange, crystal violet, and Methylene Blue) exhibited substantial absorption, ranging from ~41.5 mg/g (AcrO), ~33.5 mg/g (AO) and ~38.7 mg/g (CV) to ~47.4 mg/g (MB) (Table S2). The CV binding capacity of the 50% CA hydrogel places this SAH in the middle range of other hydrogels containing naturally derived sustainable components—it is superior to some (28 mg/g [74] and 35 mg/g [75]) and inferior to others (~67 mg/g [13]). The significantly lower absorption of Auramine Orange is due to its relatively good solubility in water (10 mg/mL [76]) thus shifting the distribution equilibrium towards the liquid phase. Nonetheless, the observed binding is still significantly higher than the values achieved in previously published studies (9.225 mg/g [77]). The enhanced absorption of cationic dyes could be attributed to favorable interactions between carboxylate groups in the hydrogels and positive functionalities in cationic dyes in combination with π-π interactions of the aromatic groups in the dyes with the aromatic rings in styrene and CA (co)polymers (Figure 15).

There is no drastic difference between the binding capacity of 100% PEGDA and 100% CA hydrogels—33.51 ± 2.03 vs. 36.93 ± 3.74, respectively (values in mg dye/g hydrogel; Table S3). Table S3 also shows no notable correlation between the composition of the other hydrogels and their affinity to CV.

Additional information on the binding mechanism is provided by ATR FTIR analyses of the dry hydrogel (50% CA) and the vacuum-oven-dried dye complexes of the same network with the investigated six dyes compiled in Table S4 and Figure S12. The carbonyl vibration band of the -COOH groups in the network undergoes the most notable shifts (Table S4; purple arrows in Figure S12). The fact that all of them shift in the direction of the -COO⁻Na⁺ carbonyls hints at the possible deprotonation of -COOH and the formation of a -COO⁻ cationic dye complex.

Anionic dyes (Methyl Orange and Bromophenol Blue), in contrast, showed no absorption. In fact, they even showed artificially negative values (Table S2). However, a visual inspection of the hydrogels confirmed that some amount of these dyes was absorbed by the hydrogels (MO and BpB; Figure S4 and Figure 15). This seemingly puzzling behavior could be explained by the specifics of the binding procedure used. The gels are introduced in the aqueous dye solution in dry form. Upon contact with the water, they begin to swell, preferentially absorbing water molecules. Some of the dye molecules are also absorbed, but most of them stay outside the crosslinked structures due to the repulsive electrostatic force of the -COO⁻ groups. Consequently, their concentration in the water outside the networks increases, causing the intensity of the absorption maxima to increase, thus falsely reporting the presence of more dye than initially dissolved. Few binding experiments with water-preswollen hydrogels were performed to verify this hypothesis. CV was used as the reference for positive dyes (Figure S13 and Table S5). It can be seen that the amount of the absorbed CV is in the same range as those absorbed in the dry gel (Tables S2 and S3), while both MO and BpB, in particular, again show false negative binding tendencies (Tables S2 and S5). Their values are smaller than those observed with the dry 50% CA hydrogel, indicating that after the preferential water uptake, the unfavorable dye distribution equilibrium still affects the binding process.

4. Conclusions

In this study, SAHs based on acrylic acid, styrene, PEGDA and cassava starch were efficiently synthesized with high yields. Infrared and NMR spectroscopy confirmed the successful synthesis of CA, which was used as an additional semi-natural modifier and potential crosslinker. Crosslinking was achieved at different ratios of CA to PEGDA, producing networks with different degrees of crosslinking density. The synthesized materials were extensively characterized by different analytical techniques. The NMR spectra of methanol and DCM extracts showed no crosslinker leakage, suggesting a good incorporation of all components into the network. SEM micrographs revealed a highly porous structure for all hydrogels, with micropores ranging from 1 to 30 μm, created by emulsion templating, and large macropores formed by nitrogen gas evolution during AIBN decomposition. The swelling capacity of the networks was within the typical SAH range, reaching a maximum of 63,000%. AUL measurement demonstrated a high resistance to stress for all gels. The hydrogels displayed high sensitivity to the salinity and the pH of the aqueous media. TGA tests showed that the SAH decomposition pattern is mostly affected by the thermal stability of the acrylic components. DSC analyses could not detect an observable glass transition within the 0–150 °C temperature interval, a sign for the rigid nature of the hydrogels. Rheology analysis showed typical gel behavior, with CA-containing hydrogels showing higher rigidity. Finally, dye-binding experiments demonstrated that SAHs have a good binding capacity towards cationic dyes, ranging from 47.4 to 33.5 mg/g in the following order MB > AcrO > CV > AO.

Overall, this study demonstrated that acrylic acid, styrene, and cassava starch could be successfully incorporated into stable SAHs by one-pot emulsion polymerization and either PEGDA or CA, a naturally based crosslinker. The process leads to the formation of SAHs containing more sustainable components (cassava starch and cardanol derivatives) than the oil-based PEG. A cost comparison of the crosslinker precursors also reveals the price advantage of cardanol vs. PEG [78,79]. The new SAHs have enhanced structural stability and swelling capacity, properties that can be potentially useful for the environmental cleanup of industrial waste effluents.