Hierarchically porous bacterial cellulose-derived biochar/PAA hydrogel beads for cationic dye adsorption: performance, mechanistic insights, and reusability

Abstract

The emergence of dye pollutants into aquatic environments necessitates the development of efficient, stable, and reusable adsorbents for wastewater treatment. In this study, hierarchically porous bacterial cellulose-derived biochar/poly(acrylic acid) (BCC-PAA) hydrogel beads were synthesized via gradual pH adjustment, constructing a three-dimensional network containing PAA-derived –COOH groups, oxygen-containing functionalities associated with the BCC-derived carbonaceous component and hydrogel network, and distinct BCC-derived carbonaceous domains. Physicochemical analyses revealed a highly porous texture, abundant oxygen-containing groups, and a pH-responsive surface with a point of zero charge of ∼5.5. The BCC-PAA beads achieved excellent adsorption capacities of 676.58 mg g⁻¹ and 725.74 mg g⁻¹ toward methylene blue (MB) and rhodamine B (RB), respectively. Adsorption kinetics were best described by the pseudo-first-order and Bangham models, indicating rapid surface adsorption followed by pore diffusion. Equilibrium data were well fitted by the Langmuir and Koble–Corrigan isotherms, suggesting predominantly monolayer adsorption with minor surface heterogeneity. Thermodynamic analysis revealed spontaneous and exothermic adsorption, with negative ΔG° values ranging from −26.77 to −26.88 kJ mol⁻¹ for MB and from −27.72 to −27.53 kJ mol⁻¹ for RB, and negative ΔH° values of −24.34 and −31.35 kJ mol⁻¹, respectively. Mechanistic evaluation suggested that adsorption may be governed by the synergistic effects of electrostatic attraction (E-attraction), π–π interactions, hydrogen-bonding (H-bonding), and pore filling phenomena. The beads retained 76.42% (MB) and 77.34% (RB) of their initial removal efficiencies after nine adsorption–desorption cycles. These findings demonstrate the potential of BCC-PAA hydrogel beads as a promising platform for further research on the efficient adsorption of cationic dyes from aqueous systems.

Bacterial cellulose-derived biochar offers a promising platform for the development of next-generation adsorbents for efficient, reliable, and reusable wastewater treatment.

1. Introduction

Despite progress since 2015, 1 in 4, or 2.1 billion, people worldwide lack access to safely managed drinking water.¹ The growing scarcity of clean water resources has become a critical global challenge due to rapid industrialization, population growth, and environmental degradation. Among various pollutants, synthetic dyes discharged from textile, printing, leather, and pharmaceutical industries represent a major class of hazardous contaminants in aquatic systems.² Wastewater containing dyes such as MB and RB is frequently discharged without adequate treatment, with reported concentrations ranging from a few mg L⁻¹ to hundreds of mg L⁻¹.³ These highly stable non-biodegradable dyes are resistant to light, heat, and microbial degradation, leading to their persistence in water bodies.

The emergence of dyes in water causes serious environmental and health crises. These persistent colorants can reduce light penetration and disrupt aquatic ecosystems, while certain dyes, particularly some azo dyes and their aromatic amine degradation products, have been associated with genotoxic and carcinogenic effects.⁴ Therefore, the development of efficient and sustainable technologies for dye removal is of paramount importance. Various treatment methods have been explored for wastewater remediation, namely membrane filtration,⁵ advanced oxidation processes,⁶ biological treatment,⁷ coagulation–flocculation,⁸ and electrochemical techniques.⁹ However, these approaches often suffer from limitations such as high operational cost, secondary pollution, energy demand, and incomplete removal performance.¹⁰ In contrast, adsorption has emerged as one of the most promising techniques due to its simplicity, high efficiency, cost-effectiveness, and adaptability to a wide range of pollutants.¹¹

The success of adsorption processes significantly depends on the selection of suitable adsorbent materials. Recently, polymeric hydrogels and carbonaceous materials have gained considerable attention. Poly(acrylic acid) (PAA) is a widely used synthetic polymer due to its abundant carboxylic groups, high hydrophilicity, pH-responsive behavior, and super swelling, which render strong electrostatic interactions with cationic dyes. However, pristine PAA hydrogels may exhibit limited mechanical stability¹² while incorporation of additional functional components can introduce complementary adsorption interactions beyond those primarily associated with the carboxyl-containing PAA network.¹³ Scientist reported various PPA composite hydrogel, namely melamine functionalized menta biochar/PAA (MB q_e = 638.43 mg g⁻¹, Crystal Violet (CV) 462.55 q_e = mg g⁻¹ and Safranin O (SO) q_e = 711.34 mg g⁻¹),¹⁴ CMC/PAA/montmorillonite hydrogels (MB, q_e = 361 mg g⁻¹),¹⁵ ZnO-Char-PAA nanocomposite hydrogels (RB q_e = 92.54%),¹⁶ cotton cellulose/PAA@LAPONITE^® hydrogel (CV, MB and Malachite Green (MG) 95%),¹⁷ curcumin functionalized biochar/PAA (MG q_e = 521 mg g⁻¹ and RB q_e = 741 mg g⁻¹),¹⁸ nanosilica/PAA hydrogel (Congo Red (CR) q_e = 555.56 mg g⁻¹).¹⁹

Biochar is a well-known natural-based adsorbent, particularly derived from biomass resources, and offers a complementary platform due to its porous architecture, large surface area, and rich functionalities such as –OH, –COOH, and carbonaceous material.²⁰ Biochar-based adsorbents such as peanut (RB 94%),²¹ rice straw (MB q_e = 539.35 mg g⁻¹),²² crab shell (MG q_e = 12 502 mg g⁻¹),²³ corn straw/red mud (acid and amino black q_e = 69.23 and 82.36 mg g⁻¹),²⁴ and sludge biochar (brilliant blue R q_e = 80.60 mg g⁻¹)²⁵ are deeply investigated for their efficacy in wastewater treatment. Among various precursors, bacterial cellulose-derived biochar has attracted considerable attention as a sustainable carbonaceous material owing to the high purity and nanofibrous structure of its bacterial cellulose precursor, as well as its interconnected network.^26,27 Upon carbonization, BC-derived biochar (BCC) develops a porous carbonaceous framework that can provide favorable sites for π–π interactions and hydrogen bonding, thereby contributing to adsorption performance.²⁸ Nevertheless, standalone biochar may exhibit aggregation issues and limited structural integrity.

To overcome these limitations, the scientist designed various combinations of biochar-polymeric hydrogel hybrids as an effective strategy to combine the advantages of both materials. The BCC incorporated polyethylenimine (PEI) and (metronidazole q_e = 691.32 mg g⁻¹ and tetracycline q_e = 960.77 mg g⁻¹) LTH decorated BCC/CS beads (2182 mg g⁻¹ for Vancomycine and 1845 mg g⁻¹ for Azythromycine) reveald autstanding performance.^29,30 It was hypothesized that the incorporation of ultra-light BCC into a pure PAA hydrogel network could generate a three-dimensional composite with enhanced porosity, induced carbonaceous domains and multifunctional adsorption sites. Despite recent advancement, studies on bacterial cellulose-derived biochar integrated with PAA hydrogel beads for simultaneous removal of multiple dyes remain undiscovered, particularly in understanding the adsorption mechanism through comprehensive kinetic and isotherm modeling.

In this context, the present study developed bacterial cellulose biochar-poly(acrylic acid) (BCC-PAA) hydrogel beads as an efficient and reliable adsorbent for the removal of cationic dyes MB and RB from aqueous solutions. The synthesized beads provided a synergistic adsorption platform that combined electrostatic interactions, hydrogen bonding, and π–π stacking within a porous hydrogel network. A systematic investigation was conducted to evaluate the effects of operational parameters including pH, contact time, adsorbent dosage, salinity effect, and initial dye concentration. The adsorption behavior was analyzed using kinetic models (Pseudo-First-Order (PFO), and Pseudo-second-order (PSO), Elovich, Bangham, and intra-particle diffusion (IPD)) and isotherm models (Langmuir, Freundlich, Temkin, and Koble–Corrigan (K–C)) to elucidate the governing mechanism. Furthermore, reusability and salt tolerance studies were performed to assess the practical applicability of the adsorbent.

The BCC-PAA hydrogel beads exhibited high adsorption capacity and rapid adsorption kinetics, which could be attributed to the combined contributions of surface functional groups and internal pore diffusion. The negatively charged carboxyl groups contributed to electrostatic attraction toward cationic dyes, while the carbonaceous domains of BCC provided potential sites for π–π interactions and hydrogen bonding. The findings of this work provided insights into the adsorption performance and possible mechanisms of dye removal using BCC-PAA hybrid hydrogel beads, highlighting their potential as a platform for further research on dye adsorption in aqueous systems.

2. Materials and method

2.1. Material

Required chemicals include acrylic acid (AA), 3,7-bis(dimethylamino) phenazathionium chloride (Methylene blue), 2,2′-azobis(2-methylpropionitrile) (AIBN), 9-(2-carboxyphenyl)-6-(diethylamino)-N,N-diethyl-3H-xanthen-3-iminium chloride (Rhodamine B), and N,N′-methylenebisacrylamide (MBA) were purchased from Sigma-Aldrich Co.

2.2. Synthesis

2.2.1. Synthesis of bacterial cellulose char

The bacterial cellulose (BC) was cultured according to the method reported by Salimi et al.³¹ The as-prepared BC membrane (200 ± 5 g, wet basis) was purified by boiling in 1 L of 1 M NaOH for 1 h to remove residual impurities and subsequently washed with deionized (DI) water until a neutral pH was achieved, as monitored using a pH meter. The purified membrane was kept refrigerated in its wet state until further use. The wet, cleaned membrane was cut into approximately 1 × 1 cm pieces and mechanically crushed to obtain a homogeneous BC slurry (paste) without the addition of extra water. A 5 g portion of the as-prepared wet BC slurry was then introduced into 100 mL of pre-cooled 63 wt% H₂SO₄. The sulfuric acid solution was precooled to approximately 3 °C before addition of the BC slurry to prevent BC degradation.

The temperature of the reaction mixture was continuously monitored and maintained at approximately 35 °C using a thermostatically controlled cooling bath equipped with a temperature probe during the hydrolysis process. After 1 h of hydrolysis, the reaction mixture was carefully introduced into approximately 500 mL of ice-cold DI water maintained at approximately 0 °C to quench the hydrolysis. The hydrolyzed BC was then collected by filtration using Whatman filter paper with a pore size of 2.5 µm and repeatedly washed with DI water until a neutral pH was reached. The material was maintained in a swollen, wet state prior to the subsequent carbonization step to avoid changes associated with drying.

For carbonization, 5 g of the wet hydrolyzed BC was dispersed in 50 mL of HCl, and the pH of the suspension was adjusted to 2 ± 0.05. The suspension was subsequently carbonized in a vacuum furnace at 200 °C for 9 h using a heating rate of 5 °C min⁻¹. The obtained bacterial cellulose char (BCC) was separated by centrifugation at 9000 rpm for 15 min, washed with DI water, and freeze-dried.²⁹

2.2.2. Synthesis of BCC char poly acrylic acid beads

0.2 g of ultralight BCC was dispersed in 10 mL of DI water containing 1.05 g of AA (not neutralized) and 0.046 g of MBA (2 mol% with respect to monomers) under an air atmosphere. The reaction was conducted at 35 °C with the addition of 0.049 g of AIBN initiator (2 mol% concerning monomers) till the successful synthesis of the hydrogel. A total of 3 mL of 0.1 M NaOH solution (0.30 mmol) was then added dropwise at 10 s intervals until bead formation was achieved, and the polymerization was allowed to proceed for an additional 30 min to ensure complete polymerization. The prepared beads were then washed with DI water and freeze-dried for further analysis. The synthesis process is summarized in Fig. 1.

Fig. 1. Synthesis of the adsorbent.

2.3. Batch adsorption experiment

The performance of BCC-PAA beads was assessed using a single-parameter adsorption test. The adsorption mechanism was elucidated using isotherm and kinetic model fitting. The impact of pH (2 to 10) and the mass of the adsorbent (1 to 5 mg) were examined over the periods of 2 to 90 minutes utilizing 10 mL of dye solution. The optimized parameters were further hired to assess the adsorbent's performance in a dye concentration range of 2 to 100 mg L⁻¹. For each test, the equilibrium concentrations were measured after separating the adsorbent using a Ultraviolet-visible (UV-VIS) spectrometer (SHIMADZU UV 3600, Japan) at the peak wavelengths of 664 nm (MB) and 553 nm (RB). The adsorbate uptake per gram of adsorbent (q_e) and removal efficacy (RE%) are determined.

1

2

3

C _e and C_i stand for saturation and initial concentration (mg L⁻¹). V (L) and m (g) show the volume of solution and adsorbent mass. The RE% denotes removal efficacy, while q_e (mg g⁻¹) and q_t (mg g⁻¹) designate adsorption capability at equilibrium and time t, respectively.

The reusability performance of BCC-PAA hydrogel beads was evaluated over 9 consecutive adsorption–desorption cycles to assess their practical applicability and structural integrity. The test was conducted using 2 mg of adsorbent, 10 mL of 50 mg L⁻¹ dye solution at the optimized pH, and a time duration of 90 min. After each test the adsorbent was washed using a 0.1 M HCl solution, neutralized using DI water, and dried in an air oven overnight.

Due to the challenges associated with obtaining real samples, the effect of ionic strength was investigated using NaCl and CaCl_2, FeCl₃ and Na₂SO₄ as representative mono- and divalent salts commonly found in wastewater. The experiments were conducted using 2 mg of adsorbent and 10 mL of 50 mg L⁻¹ dye solutions at the optimized pH, with salt concentrations ranging from 5 to 100 mM and a contact time of 90 min.

3. Characterization

A UV–Vis spectrophotometer (UV-2550, Shimadzu, Japan) was used to determine the dye concentration. Fourier-transform infrared spectroscopy (FTIR) was performed using a Nicolet 6700 spectrometer (Thermo Fisher Scientific, USA). The morphology of the adsorbent was examined using a scanning electron microscope (SEM; JSM-6010LA, JEOL, Japan), while thermogravimetric analysis (TGA) was conducted using a Mettler Toledo thermogravimetric analyzer (Switzerland). The zeta potential of the adsorbent was measured using a zeta potential analyzer (SZ-100, HORIBA, Japan). Energy-dispersive X-ray spectroscopy (EDS) was performed using a Numerix DXP-X10P system. The X-ray diffraction (XRD) pattern of the sample was obtained using a Bruker D2 PHASER X-ray diffractometer. The N₂ adsorption–desorption isotherm and pore-size distribution of the BCC–PAA beads were measured using a Micromeritics TriStar II 3020 surface area and porosity analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA).

3.1. FTIR

FTIR analysis was performed to investigate the surface functionalities of BCC and the BCC-PAA beads (Fig. 2a). The FTIR spectrum of BCC displayed a broad band at 3432 cm⁻¹, corresponding to O–H stretching vibrations of hydroxyl groups, indicating the presence of residual cellulose functional groups after carbonization and the hydroxyl groups on the biochar surface. Peaks at 2931 cm⁻¹ were assigned to C–H stretching, while the band at 1645 cm⁻¹ was attributed to conjugated C O stretching vibration and C C in the carbonized structure.³² The peaks at 1170 cm⁻¹ corresponded to aliphatic and aromatic C–O vibrations.³³ Additional peaks at 1113 cm⁻¹ and 1055 cm⁻¹ are assigned to vibrational modes of pentose and hexose structures originating from hemicellulose and confirmed ether linkages and polysaccharide structures.³⁴ The presence of aromatic regions and –OH and C O groups in BCC enables the establishment of H-bonding and π–π interactions with pollutants.

Fig. 2. FTIR spectra of BCC, BCC-PAA (a), TGA and DTG curves of BCC-PAA beads (b), and N₂ adsorption–desorption isotherm diagram (c).

For BCC-PAA, substantial transformation was observed, confirming successful incorporation of BCC onto PAA beads. The O–H stretching band shifted slightly to 3484 cm⁻¹ and broadened, reflecting the presence of abundant –OH functional groups enabling extensive H-bonding between the hydroxyl groups of BCC and carboxylic groups of PAA and the pollutant. The C O stretching of the –COOH groups appeared at 1741 cm⁻¹. The band at 1651 cm⁻¹ is assigned to overlapping stretching vibrations of C C and C O. The bands at 1545 cm⁻¹ and 1405 cm⁻¹ were attributed to asymmetric and symmetric stretching of –COO⁻ groups, indicative of partial ionization/interaction with BCC. Peaks at 1176 cm⁻¹ and 1020 cm⁻¹ correspond to C–O and C–O–C vibrations from both BCC and PAA chains.³⁵ These functionalities enable the formation of BCC-PAA beads, with strong H-bonding and electrostatic interactions and hydrophobic interactions between adsorbate and adsorbent, resulting from an integrated polymer-carbon framework.

3.2. Thermogravimetric analysis

The thermal decomposition behavior of the BCC-PAA beads was evaluated by TGA under an N₂ environment in the range of 30–800 °C. Fig. 2b shows the TGA and Differential Thermogravimetric Analysis (DTG) results. The TGA curve exhibits three distinct weight loss stages. The first stage at temperature >150 °C (with the maximum at 61 °C and ∼11.49% mass loss) is attributed to the removal of physically adsorbed water and residual moisture from the hydrogel matrix, which is assigned to the hydrophilic nature of both PAA and residual hydroxyl groups on BCC fibers.

The second stage, between ∼200 and 400 °C (with a maximum at 257 °C), with the highest mass loss at ∼61.74%, corresponds to the elimination of chemically H-bounded water molecules, decomposition of the PAA chains, and partial degradation of the remaining carbonized cellulose fibers. This result reflects the cleavage of carboxyl groups, C–C backbone scission, and the breakdown of the hydrogel network.³⁶ The third stage, in the range of 400–600 °C, with ∼28.75% mass loss (max at 398 °C), is attributed to the further carbonization of residual cellulose and the formation of stable carbonaceous structures. Beyond 600 °C, the mass stabilizes, indicating the formation of thermally stable carbon.

3.3. Surface area and porosity

The textural properties of the BCC-PAA beads were further investigated by N₂ adsorption–desorption analysis, with the corresponding pore-size distribution shown in Fig. 2c. The material exhibited a Brunauer–Emmett–Teller (BET) specific surface area of 8.32 m² g⁻¹, while the cumulative pore volume obtained from the Barrett–Joyner–Halenda (BJH) adsorption analysis was 0.045 cm³ g⁻¹. The moderate BET surface area is reasonable for the hydrogel-based structure, in which the carbonaceous component is integrated within a crosslinked polymeric network.

The BJH adsorption pore-size distribution revealed pores predominantly within the mesoporous range. The distribution extended from approximately 20 to 110 Å (2–11 nm), with the highest contribution occurring around 40–45 Å (4–4.5 nm). Thus, the BCC-PAA beads possess a substantial population of nanoscale mesopores. In addition, the t-plot analysis did not indicate a meaningful micropore volume, further suggesting that the accessible porosity is dominated by mesoporous rather than microporous domains. The mesoporous structure identified by BJH, together with the larger interconnected pores/voids observed in the SEM images, indicates a multiscale pore architecture. Accordingly, the BCC-PAA beads can reasonably be described as having a hierarchical porous structure, in which larger interconnected voids provide transport pathways while nanoscale mesopores contribute additional accessible internal surfaces. Such a pore architecture is expected to facilitate the diffusion of MB and RB molecules and improve their accessibility to the adsorption-active functionalities within the beads.

3.4. SEM and elemental composition

The morphology and internal structure of the crude PAA hydrogel and BCC-PAA beads are shown in Fig. 3. The PAA hydrogel (Fig. 3a) exhibited a homogeneous structure, with a relatively smooth surface and random micropores, without uniform porous architecture. The absence of porous structure limits the diffusion of adsorbate species and inhibits effective adsorption. In contrast, the BCC-PAA composite beads (Fig. 3b) displayed spherical morphologies with sizes in the 200–500 µm range and porous architecture. High-magnification cross-section micrographs of internal bead structure (Fig. 3c and d) revealed the development of a highly porous construction with interconnected pores, generating a 3D sphere during bead generation. The porosity of the beads provides an abundance of accessible sites for adsorption applications.

Fig. 3. SEM micrographs of PAA hydrogel (a), BCC-PAA beads (b), Internal porous structure of beads (c, d), BCC in composite (e, f), elemental mapping (g) and EDS results of the adsorbent (h).

At higher magnifications (Fig. 3e and f), the BCC fibers were evidently recognizable within the PAA matrix. Individual carbonized cellulose fibers appeared as elongated, entangled nanostructures with diameters in the range of 50–200 nm, forming a reinforcing network within the hydrogel. The intimate contact between BCC fibers and the polymer matrix suggests strong physical interactions (H-bonding between the –OH groups of BCC and –OH groups of PAA), which improves structural integrity while providing supportive interactions with adsorbate species. Elemental mapping and EDS analysis (Fig. 3g and h) show the composition of the BCC-PAA beads. The uniformly distributed carbon throughout the composite originates from both the BCC carbonaceous fibers and the polymer backbone, while oxygen is also homogeneously distributed, consistent with the oxygen-containing functionalities in PAA and residual oxygen in BCC. The atomic ratio (C: 59.20%, O: 40.80%) corroborates the high content of carbonaceous fibers and the carboxyl-rich polymer network. Overall, SEM and elemental analysis demonstrate that the BCC-PAA beads possess a hierarchical, porous architecture with well-dispersed carbon fibers embedded in the polymer network, strengthening robust adsorption.

3.5. Surface charge analysis

Surface charge analysis has been conducted to analyze the change in surface charge as a function of pH (from 2 to 10). Fig. 4a exhibits the result of the surface charge study. At the pH ranging from 2 to 4, the beads revealed a positive surface charge (28–11 mV). The carboxylic acid (–COOH) groups from PAA and residual hydroxyl groups from carbonized BCC fibers remain protonated in a highly and mildly acidic environment, yielding a positive charge. Partial deprotonation of –COOH groups by pH enhancement reduced the surface charge. Zeta potential becomes more negative, reflecting the formation of –COO⁻ groups. The deprotonation of these groups by pH enhancement reduced the charges and the point of zero charge attained at pH = 5.5 ± 0.05. Beyond pH 8–10, the beads exhibited a negative charge, and nearly all carboxyl groups are deprotonated in the ranges 8–10, so the surface remains highly negatively charged. The zeta potential reached its maximum negative value (−35 mV), which enhances E-repulsion and colloidal stability in aqueous dispersion. Compared to pure PAA, the presence of BCC fibers slightly modulated the zeta potential. pH-dependent surface charge plays a crucial role in applications such as adsorption and ion exchange.

Fig. 4. Zeta potential results of composite beads (a), XRD diffractogram of BCC and BCC PAA beads (b).

3.6. XRD analysis of the developed adsorbent beads

The crystalline structure of BCC and BCC-PAA hydrogel beads was analyzed using XRD method, and the obtained difractograms are presented in Fig. 4b. The BCC sample exhibits characteristic diffraction peaks at 2θ ≈ 17.73° (110), 24.86° (200), and 39.34°, which are associated with the partially ordered carbon structure derived from bacterial cellulose after carbonization. The broad nature of these peaks indicates that the material is predominantly amorphous with low crystallinity, a typical feature of biochar derivatives. After incorporation into the PAA hydrogel matrix, the BCC-PAA beads show a noticeable change in the diffraction pattern, with broadened peaks centered around 2θ ≈ 17.23° and 21.03° (short-range ordering of polymer chains). The shift and reduction in peak intensity suggest a decrease in crystallinity due to the integration of BCC into the polymeric network. The absence of sharp crystalline peaks further confirms the formation of a highly amorphous hydrogel. Overall, the XRD outcomes confirm that the BCC-PAA beads possess a predominantly amorphous structure with disrupted ordering after hydrogel development.

4. Adsorption experiment

4.1. Impact of pH on the adsorption efficacy

The optimal pH adjustment is carried out in the pH ranges from 3 to 10 using 10 mL of 50 mg L⁻¹ dye solution using 2 mg of adsorbent over a time period of 90 min. The surface charge and the pH-dependent dye structure impact adsorbate–adsorbent interactions. Fig. 6a depicts the influence of pH on dye adsorption. As can be observed, both dyes exhibited a climbing-declining pattern. The maximum RE% for MB (95.52%) and RB (98.61%) were attained at pH 7 and 5.5, respectively.

Fig. 6. The impact of pH (a) (2 mg adsorbent; 10 mL of 50 mg L⁻¹ dye solutions; 90 min), and adsorbate dosage (b) on adsorbent efficacy (10 mL of 50 mg L⁻¹ dye solutions; 90 min; optimized pH), time (c) (2 mg adsorbent; 10 mL of 50 mg L⁻¹ dye solutions, optimized pH) and concentration (d) (1 mg adsorbent; 10 mL of dye solutions; 90 min; optimized pH) dependent adsorption. Values are reported as mean ± SD, with error bars representing the standard deviation (n = 3).

RB has two pK_a values (Fig. 5). At a pH lower than 4.2 the molecule exists in its cationic form (RB⁺). The protonated amine groups (ammonium) are repelled by the positive adsorbent surface. By enhancing the pH, the surface charges reduced, which lowered the adsorbate–adsorbent repulsion and enhanced the adsorption. After pH > 4.2, the –COOH group of RB will be deprotonated, and the molecule will transform into its zwitterionic neutral form (RB^±). In this range up to point of zero charg (pH_PZC = 5.5), the surface charge will be reduced in the positive range. The repulsion between the cationic ammonium group (RB) and the bead surface will reduce while H-bonding and charge-assisted H-bonding (CAHB) enhance the adsorption, and the higher adsorption is observed at pH ∼5.5 with RE% = 98.61% (Fig. 6a). Adsorption between positive active sites and –COO⁻ groups of RB plays a critical role at 4.2 < pH < 5.5. At a pH higher than pH_PZC, the surface exhibits a negative surface charge, which effectively attracts RB ammonium groups, while the negative COO⁻ group will be repelled. While at this range, both repulsion and adsorption are enhanced by the adsorbent surface charge, no significant changes in adsorption performance are observed at pH > 5.5, while minor enhancements can be attributed to CAHB.³⁷

Fig. 5. Potential interaction between MB and RB molecules with the BCC-PAA beads in various pH (Isoelectric point (IEP)).

At pH < 2 the MB molecules exist in their protonated cationic (MB²⁺) form that strongly repels the adsorbent's positively charged active site. At 2 < pH < 7, the amine groups will remain protonated, and the molecules will exist in protonated form while still repelling the adsorbent's positive charges (Fig. 5). Further pH enhancement reduces repulsion with MB molecules and adsorbs them more efficiently by the adsorbent. At the pH_PZC = 5.5, the surface carries zero net charge, and minor E-interactions exist between the adsorbent and the adsorbate. Beyond pH = 5.5, the negative active site can adsorb positive MB²⁺ molecules and reaches its maximum at pH = 7 with RE% = 95.52% (Fig. 6a). At pH > 7 the second amine group will be deprotonated, which drastically reduces the adsorption capacity, while further enhancement in surface charge in the negative range once more enhances the adsorption performance by electrostatic forces (Fig. 6a).

4.2. Impact of adsorbent dosage on the adsorption efficacy

The impact of adsorbent dosage on the RE% and q_e of MB and RB was evaluated under optimized conditions (10 mL of 50 mg L⁻¹ dye solution for 90 min, pH = 5.5 and 7 for RB and MB). As illustrated in Fig. 6b, increasing the adsorbent dosage from 0.1 to 0.5 g L⁻¹ results in a significant improvement in RE%. For MB, the RE% increases from 88.63% at 0.1 g L⁻¹ to 98.92% at 0.5 g L⁻¹. RB shows a similar trend, rising from 93.30% to complete removal (100%) over the same dosage range. This enhancement is accredited to the increase in available active sites, allowing more dye molecules to be effectively adsorbed from the solution. The sharp increase in RE% at lower dosages (0.1–0.2 g L⁻¹) indicates that adsorption is initially limited by the availability of active sites, whereas the system approaches saturation at higher dosages.

Conversely, the q_e decreases notably with increasing dosage. For MB, q_e declines from 443.16 mg g⁻¹ at 0.1 g L⁻¹ to 98.92 mg g⁻¹ at 0.5 g L⁻¹, while RB displays a reduction from 466.49 mg g⁻¹ to 99.68 mg g⁻¹. This behavior can be elucidated by the fixed amount of dye molecules over a larger mass of adsorbent, leading to underutilization of available adsorption sites at higher dosages. Overall, 0.1 g L⁻¹ (1 mg) of adsorbent was selected as the optimum dosage for subsequent experiments, as it provided relatively high removal efficiencies while maintaining the highest adsorption capacity among the investigated dosages.

4.3. Time and concentration dependence adsorption

The effect of contact time on the adsorption of MB and RB onto BCC-PAA hydrogel beads was investigated under optimized conditions (2 mg adsorbent, 10 mL 50 mg L⁻¹ dye solution at optimized pH (pH = 5.5 and 7 for RB and MB)), and the results are presented in Fig. 6c. Both dyes exhibit a classic multi-stage adsorption behavior, consisting of a rapid initial uptake followed by a gradual approach to equilibrium. In the initial stage (0–10 min), adsorption proceeds rapidly due to external surface adsorption and film diffusion. For MB, the removal efficiency increases sharply from 0 to 67.57%, with adsorption capacity reaching 168.94 mg g⁻¹, while RB shows a smoother increase to 52.32% and 130.80 mg g⁻¹. This difference indicates faster diffusion of MB molecules into the swelled adsorbent due to its lower steric hindrance. In the intermediate stage (10–30 min), the adsorption rate slows as the process becomes controlled by intra-particle diffusion and the gradual occupation of internal active sites. During this period, MB removal increases from 67.57% to 89.24% (q_t ≈ 223.09 mg g⁻¹), while RB rises from 52.32% to 90.64% (q_t ≈ 226.60 mg g⁻¹), indicating enhanced diffusion of dye molecules into the porous structure of the hydrogel beads. Beyond 30 min, the system reached a near-equilibrium phase, where adsorption occurs at a much slower rate due to saturation of available sites. Equilibrium was attained after 50 min. At 90 min, MB exhibited a removal efficiency of 95.53% with an adsorption capacity of approximately 238.81 mg g⁻¹, while RB achieved an RE% of 98.62% with a corresponding q_e of approximately 246.54 mg g⁻¹. The minimal increase in RE% after 60 min (<3%) suggests that equilibrium had been effectively established.

The effect of C_i on the RE% and q_e of BCC-PAA hydrogel beads was investigated in the range of 2–100 mg L⁻¹ under optimized conditions (1 mg adsorbent, 90 min). The results are presented in Fig. 6d. At low C_i (2–8 mg L⁻¹), both dyes exhibited complete or near-complete removal (RE ≈ 100%), with q_e increasing proportionally from 20 to ∼80 mg g⁻¹ due to the excess availability of adsorption sites relative to the number of dye molecules. As the concentration increased to intermediate levels (10–40 mg L⁻¹), a gradual decline in RE% was observed. For MB, RE% decreased from 94.67% at 10 mg L⁻¹ to 88.39% at 40 mg L⁻¹, while RB decreased from 99.73% to 94.55%. Meanwhile, q_e increased to 353.58 and 378.19 mg g⁻¹ for MB and RB, respectively, due to the increased concentration gradient and greater availability of dye molecules. At higher concentrations (50–100 mg L⁻¹), the decline in RE% became more pronounced, reaching 67.66% for MB and 72.57% for RB at 100 mg L⁻¹, while the corresponding q_e values reached 676.58 and 725.74 mg g⁻¹.

To evaluate the effect of BCC incorporation, pristine PAA hydrogel was tested as a control under identical experimental conditions over the same concentration range. Although complete removal of both dyes was observed at 2–4 mg L⁻¹, the adsorption performance of pristine PAA declined more markedly with increasing C_i. At 100 mg L⁻¹, pristine PAA exhibited RE% values of 25.80% for MB and 21.91% for RB, corresponding to q_e values of 258.04 and 219.15 mg g⁻¹, respectively. Under the same conditions, BCC-PAA achieved substantially higher q_e values of 676.58 mg g⁻¹ for MB and 725.74 mg g⁻¹ for RB. These comparative results indicate that the incorporation of BCC markedly improved the adsorption performance of the PAA hydrogel, particularly at higher dye concentrations.

Overall, MB exhibited faster initial adsorption, whereas RB showed a slightly higher final adsorption capacity on BCC-PAA. Across the investigated concentration range, RB generally maintained a higher RE% than MB, particularly at elevated C_i, indicating its stronger retention within the BCC-PAA bead matrix. The comparison with pristine PAA further demonstrated that BCC incorporation substantially improved the adsorption performance of the hydrogel, particularly at higher dye concentrations. These results indicate that the adsorption behavior of BCC-PAA influenced by the structural and functional characteristics introduced by the BCC component. The observed behavior is consistent with the contribution of electrostatic attraction, hydrogen bonding, possible π–π interactions, and diffusion into the porous bead structure.

4.4. Kinetic study

To elucidate the kinetic mechanism (rate-limiting step) of MB and RB adsorption into BCC-PAA beads, the adsorption data were fitted into nonlinear PFO, PSO, Elovich, Bangham, and IPD models. Table 1 and Fig. 7 show the kinetic results.

Table 1. Kinetic study results for the adsorption of MB and RB into BCC-PAA beads.

Parameter		Parameter (Unit)		MB	RB
PFO		q t	(mg g−1)
		q cal	(mg g−1)	231.941	249.157
		k 1	(min−1)	0.124	0.072
		R 2		0.995	0.990
		P-value		4.394 × 10−23	8.648 × 10−20
		F-value		10 931	3694
PSO		q cal	(mg g−1)	261.805	299.646
		k 2	(g mg−1 min−1)	6.012 × 10−4	2.562 × 10−4
		R 2		0.982	0.974
		P-value		6.673 × 10−19	1.091 × 10−16
		F-value		2757	1327
Elovich		α El	(mg g−1 min−1)	103.405	45.075
		β EI	(g mg− 1)	0.020	0.014
		R 2		0.938	0.964
		P-value		3.160 × 10−15	9.733 × 10−16
		F-value		818	969
Bangham	q t = qm[1 − exp(−kbtn)]qm	q m	(mg g−1)	230.898	0.832
		k b		0.111	0.013
		N		1.057	0.061
		R 2		0.996	0.999
		P-value		3.072 × 10−21	2.045 × 10−21
		F-value		7355	1023
IPD	q t = kid × t0.5 + C	k id	(mg g−1 min0.5)	23.093	28.749
		C	(Mg g−1)	62.218	24.001
		R 2		0.753	0.855
		P-value		46.817	83.137
		F-value		8.032 × 10−6	2.894 × 10−7

Fig. 7. Kinetic and isothermal fitting plots of PFO (a), PSO (b), Elovich (c), Bangham (d), IPD (e), Langmuir (f), Freundlich (g), K–C (h) and Temkin (i) for the adsorption of MB and RB into BCC-PAA beads.

4.4.1. Pseudo-first-order (PFO) model

For MB the PFO model shows an excellent fit (R² = 0.995) with a high F-value, indicating strong statistical significance. The calculated adsorption capacity (q_cal = 231.94 mg g⁻¹) is reasonably close to expected experimental values (238.81 mg g⁻¹), suggesting that physisorption processes govern MB adsorption, particularly at the initial stages. The higher rate constant (k₁ = 0.124 min⁻¹) implies rapid adsorption due to strong E-attraction between the MB ammonium molecules and -COO⁻ on the BCC-PAA beads (Fig. 7a). Similarly, RB follows the PFO model (R² = 0.990), slightly lower than MB. The lower rate constant (k₁ = 0.072 min⁻¹) indicates slower adsorption kinetics, which is attributed to the steric hindrance of RB. The strong fit of PFO suggests that physisorption plays a vital role in the adsorption process (Fig. 7a).³⁸

4.4.2. Pseudo-second-order (PSO) model

For MB the PSO model exhibits a slightly lower correlation (R² = 0.982) compared to PFO. However, the calculated adsorption capacity (q_cal = 261.80 mg g⁻¹) suggests that chemisorption also contributes, involving electron sharing and exchange between MB molecules and functional groups (–COOH and –OH) on the adsorbent. The relatively low k₂ value (6.012 × 10⁻⁴) indicates that although chemisorption occurs, it is not the sole rate-controlling mechanism (Fig. 7b). For RB, the PSO model shows a lower fit (R² = 0.974) compared to the PFO model, indicating that chemisorption is not the dominant mechanism. The lower k₂ value (2.562 × 10⁻⁴) further confirms slower interaction kinetics.³⁹

4.4.3. Elovich model

For MB the Elovich model provides a moderate fit (R² = 0.938), indicating adsorption on a heterogeneous surface. The high α value (103.405 mg g⁻¹ min⁻¹) suggests a high initial adsorption rate, while the β value (0.020 g mg⁻¹) reflects moderate surface coverage and activation energy. This supports the presence of mechanical interaction between MB and BCC-PAA beads (Fig. 7c). RB shows a better fit to the Elovich model (R² = 0.964), indicating stronger heterogeneity effects during adsorption.⁴⁰ The lower α value (45.075 mg g⁻¹ min⁻¹) confirms a slower initial adsorption rate compared to MB. This behavior suggests that RB adsorption is more influenced by surface heterogeneity and the energy variation of adsorption sites (Fig. 7c).

4.4.4. Bangham model

The Bangham model demonstrates an excellent fit (R² = 0.996), indicating that pore diffusion plays a significant role in MB adsorption. The value of n ≈ 1 suggests that adsorption is not purely diffusion-controlled but involves a combination of film diffusion and pore diffusion mechanisms. Similarly, RB exhibits the best fit with the Bangham model (R² = 0.999), strongly confirming that pore diffusion is the central mechanism.⁴¹ The very low n value (0.061) indicates significant resistance to diffusion within pores due to the larger molecular size of RB and steric constraints within the hydrogel matrix (Fig. 7d).

4.4.5. Intra-particle diffusion (IPD) model

The IPD model shows a relatively poor fit (R² = 0.753), indicating that intra-particle diffusion is not the sole rate-limiting step for MB for the whole adsorption interval. The high intercept (C = 62.218 mg g⁻¹) suggests a strong boundary layer effect, meaning surface adsorption occurs prior to diffusion into pores (Fig. 7e). For RB, the fit is also weak (R² = 0.855), confirming that intra-particle diffusion alone cannot describe the adsorption process. However, the higher k_id value (28.749 mg g⁻¹. min^−0.5) compared to MB (23.093 mg g⁻¹. min^−0.5) indicates that once diffusion begins, RB transport within pores occurs at a comparatively constant rate, though still limited by initial external resistance. The higher correlation in the initial intervals for both dyes indicates that the IPD phenomenon controls adsorption in the early adsorption stage (Fig. 7e).⁴²

The kinetic analysis reveals that the adsorption of both dyes onto BCC-PAA hydrogel beads is governed by a multi-step mechanism, involving surface adsorption, boundary layer diffusion, and pore diffusion, which is supported by the superior fitting of the PFO and Bangham models for both MB (R² = 0.995 and 0.996) and RB (R² = 0.990 and 0.999), respectively, indicating that physisorption and pore diffusion are dominant mechanisms. In contrast, the relatively lower R² values of the PSO model (0.982 for MB and 0.974 for RB) and Elovich model (0.938 for MB and 0.964 for RB) suggest that chemisorption plays a secondary role. The faster adsorption of MB is evidenced by its higher rate constant (k₁ = 0.124 min⁻¹) compared to RB (k₁ = 0.072 min⁻¹), reflecting more rapid interaction with the negatively charged surface functionalities. Furthermore, the Bangham model parameter (n ≈ 1.057 for MB vs. 0.061 for RB) indicates that RB experiences greater pore diffusion resistance, confirming its more diffusion-controlled behavior due to steric hindrance. The poor fitting of the IPD model for MB (R² = 0.753) and the significant intercept values (C = 62.218 mg g⁻¹ for MB and 24.001 mg g⁻¹ for RB) demonstrate that IPD is not the main rate-limiting step, although it may meaningfully govern the adsorption in the initial steps.

4.5. Isotherm study

To assess the mechanism of the adsorption of MB and RB into BCC-PAA beads, the isothermal analysis has been carried out by fitting the adsorption data into non-linear Langmuir, Freundlich, Temkin, and Koble-Carrigan isotherm models.⁴³Table 2 and Fig. 7 show the attained results.

Table 2. Isotherm results for the adsorption of MB and RB into BCC-PAA beads.

Model		Parameter	Unit	MB	RB
Langmuir		q e	mg g−1	676.58	725.74
		q cal	mg g−1	820.679	878.915
		k L	L mg−1	0.197	0.195
		R L		0.048	0.049
		R 2		0.984	0.987
		P-value		1.041 × 10−16	3.559 × 10−17
		F-value		1336	1559
Freundlich	q e = kfCe1/n	k F	(mg g−1)(L mg−1)1/n	202.465	191.403
		1/n		0.4081	0.405
		R 2		0.947	0.943
		P-value		6.170 × 10−13	9.249 × 10−13
		F-value		381.45007	359
Temkin		b T	kJ mol−1	136.220	145.842
		A T	L mg−1	4.493	4.421
		R 2		0.935	0.938
		P-value		2.293 × 10−12	1.769 × 10−12
		F-value		315.014	327.169
K–C		A KC	Ln mg1−n g−1	164.850	174.83
		B KC	(L mg−1)n	0.198	0.196
		n KC		0.981	0.975
		R 2		0.983	0.985
		P-value		4.329 × 10−15	1.577 × 10−15
		F-value		828	968

4.5.1. Langmuir isotherm

The Langmuir model provided a strong correlation with the experimental MB adsorption data (R² = 0.984, F = 1336), indicating that the model adequately describes the observed equilibrium behavior. The Langmuir constant (k_L = 0.197 L mg⁻¹) is consistent with favorable adsorption, while its magnitude indicates appreciable affinity between MB and the adsorbent. Although the q_cal = 820.68 mg g⁻¹ does not fully match the experimental q_max = 676 mg g⁻¹, this difference reflects the deviation between the model-derived and experimental capacities.⁴⁴ B similarly follows the Langmuir model (R² = 0.987, F = 1559), with higher estimated adsorption capacity (q_cal = 878.92 mg g⁻¹), indicating that the Langmuir model adequately describes the equilibrium adsorption behavior of RB. The comparable k_L value (0.195 L mg⁻¹) suggests similar affinity to MB. The slightly higher capacity may be associated with differences in the interactions of RB with the carbonaceous and polymeric substrate of the composite, although the specific contributions of individual interactions requires further analysis (Fig. 7f).⁴⁵

4.5.2. Freundlich isotherm

The Freundlich model exhibits a moderate fit (R² = 0.947), indicating adsorption on a heterogeneous surface (Fig. 7g). The 1/n = 0.408 (<1) suggests advantageous adsorption and strong interaction between MB and the adsorbent. The RL values (0.048 and 0.049 for MB and RB) denote favorable adsorption. The relatively high k_F value (202.47 (mg g⁻¹)(L mg⁻¹)^1/n) supports the presence of energetically diverse active sites. For RB, the Freundlich model also indicates promising adsorption (R² = 0.943, 1/n = 0.405) and suggests the presence of heterogeneous active sites. However, the slightly lower k_F value (191.40) compared to MB indicates marginally weaker interaction strength on heterogeneous sites.⁴⁶

4.5.3. Koble–Corrigan (K–C) isotherm

The Koble–Corrigan model provided a very good fit for MB (R² = 0.983), with n_KC = 0.981, indicating that the equilibrium adsorption behavior was well described by this model and approached the Langmuir limit. Similarly, RB showed a strong fit (R² = 0.985), with n_KC = 0.975, also suggesting behavior close to the Langmuir limit. These results, together with the other isotherm analyses and the physicochemical characteristics of the BCC-PAA beads, suggest that the equilibrium adsorption behavior may involve contributions from both surface affinity and adsorbent heterogeneity (Fig. 7h).⁴⁷

4.5.4. Temkin isotherm

The lower correlation of the Temkin model (R² = 0.935) suggests that adsorbate–adsorbent interactions and heat of adsorption effects are present but not dominant. The Temkin constant (b_T = 136.22 kJ mol⁻¹) indicates relatively strong interactions, while the AT value (4.493 L mg⁻¹) reflects moderate binding affinity. RB exhibits a slightly better fit (R² = 0.938), with a higher b_T value (145.84 kJ mol⁻¹), indicating stronger interaction energy compared to MB.⁴⁸ This may be associated with differences in the interactions of RB with the functional groups and carbonaceous domains of the BCC-PAA matrix. In conjunction with the Bangham kinetic analyses, which indicated a contribution from diffusion to the overall adsorption process, the Temkin results suggest that adsorption may involve both mass-transfer effects and energetically significant adsorbate–desorbent interactions (Fig. 7i).

The isotherm analysis suggests that the adsorption of both dyes onto BCC-PAA hydrogel beads is predominantly described by Langmuir-type behavior, as suggested by the results of the Langmuir model for MB (R² = 0.984) and RB (R² = 0.987), along with their high adsorption capacities (820.68 and 878.92 mg g⁻¹, respectively). The strong agreement with the Koble–Corrigan model (R² = 0.983 for MB and 0.985 for RB; n_KC ≈ 1) further supports that while Langmuir-type adsorption dominates, may indicate contributions from surface heterogeneity. The Freundlich constants (1/n = 0.408 for MB and 0.405 for RB) indicate favorable adsorption and suggest presence of energetically diverse active sites, though their lower R² values (0.947 and 0.943) suggest the weaker description of the equilibrium data. Additionally, the statistical results of the Temkin model (R² ≈ 0.935–0.938) and relatively high b_T values (136.22–145.84 kJ mol⁻¹) suggest that variations in adsorption energy may contribute to the overall adsorption process. Overall, these findings suggest that adsorption equilibrium behavior is predominantly Langmuir-type, with strong affinity toward active sites, accompanied by secondary contributions from surface heterogeneity and intermolecular interactions.

4.6. Thermodynamic evaluations

The spontaneity, feasibility, and energetic nature of the adsorption process were evaluated through thermodynamic analysis. The adsorption experiments were performed at temperatures ranging from 298–313 K (1 mg adsorbent, using 10 mL of 2–100 mg L⁻¹ dye solutions at optimized pH over 90 min). Thermodynamic parameters including ΔG°, ΔH°, and ΔS° were determined using the Van't Hoff equation,⁴⁹ and the thermodynamic equilibrium constant (K_e°) was obtained by converting the Langmuir equilibrium constant (K_L) into a dimensionless parameter according to eqn (4)–(6). The obtained results are summarized in Fig. 8a and b and Table 3.

4

5

6

where the K_L stances for the Langmuir constant (L g⁻¹), [adsorbate]^o is standard concentration of the adsorbate (1 mol L⁻¹) and γ exhibit activity coefficient of adsorbate (dimensionless).

Fig. 8. Thermodynamic graphs of vant, Hof (a) and ΔG (b), reusability result of BCC-PAA beads in MB and RB adsorption (c) (2 mg of adsorbent, 10 mL 50 mg L⁻¹ dye solution over 90 min at pH of 6 for RB and 7 for MB), effect of Ca²⁺ (d), and Na⁺ (e) concentration on adsorption efficacy, the comparision of mono di and trivalent cations and mono and divalent anions on adsorption efficacy (f) (2 mg of adsorbent, 10 mL 50 mg L⁻¹ dye solution over 90 min and 20 mM salt concentration).

Thermodynamic outcomes.

Thermodynamic data of MB
Temperature	298 K	303 K	308 K	313 K
K L (L mg−1)	0.197	0.165	0.143	0.121
K e°	63.01 × 103	52.77 × 103	45.73 × 103	38.70 × 103
ΔH° (kJ mol−1)	−24.34	—	—	—
ΔS° (J mol−1 K−1)	8.11	—	—	—
ΔG° (kJ mol−1)	−26.77	−26.78	−26.870	−26.88

Thermodynamic data of RB
Temperature	298 K	303 K	308 K	313 K
K L (L mg−1)	0.195	0.153	0.125	0.104
K e°	93.40 × 103	73.28 × 103	60.11 × 103	50.00 × 103
ΔH° (kJ mol−1)	−31.35	—	—	—
ΔS° (J mol−1 K−1)	−12.26	—	—	—
ΔG° (kJ mol−1)	−27.72	−27.59	−27.55	−27.53

The adsorption equilibrium constants declined gradually with growing temperature for both dyes. For MB, K_L reduced from 0.197 to 0.121 L mg⁻¹, while for RB it decreased from 0.195 to 0.104 L mg⁻¹ between 298 and 313 K. A similar trend was detected for K_e°, which weakened from 63.01 × 10³ to 38.70 × 10³ for MB and from 93.40 × 10³ to 50.00 × 10³ for RB. This temperature-dependent decline confirms that adsorption becomes less favorable at elevated temperatures, indicating an exothermic adsorption process. The ΔG° values remained negative throughout the studied temperature range, confirming the spontaneous nature of adsorption for both dyes. RB exhibited slightly more negative ΔG° values (−27.72 to −27.53 kJ mol⁻¹) compared with MB (−26.77 to −26.88 kJ mol⁻¹). Furthermore, RB showed consistently higher equilibrium constants than MB at all temperatures. The more negative ΔG° values and higher equilibrium constants indicate that RB adsorption is thermodynamically more favorable than MB adsorption, which is consistent with its stronger adsorption affinity.

The negative enthalpy values of −24.34 kJ mol⁻¹ for MB and −31.35 kJ mol⁻¹ for RB confirm the exothermic nature of the adsorption process. The magnitude of ΔH° for both dyes falls within the range generally associated with physical adsorption, suggesting that non-covalent interactions predominantly govern the adsorption process. The slightly higher absolute enthalpy value observed for RB indicates stronger interactions with the BCC-PAA beads compared with MB, which contribute to its greater thermodynamic favorability. The entropy change showed dissimilar trends for the two dyes. A positive ΔS° value of 8.11 J mol⁻¹. K⁻¹ was obtained for MB, demonstrating a slight increase in randomness at the solid–liquid interface during adsorption. This behavior attributed to the release of hydrated water molecules from the adsorbent surface and/or the dye molecules into the bulk solution, resulting in a net increase in system disorder. In contrast, RB exhibited a negative ΔS° value of −12.26 J mol⁻¹ K⁻¹, indicating a decrease in randomness at the adsorption interface. This suggests that RB molecules become more orderly arranged upon adsorption onto the BCC-PAA surface. The opposite signs of ΔS° for MB and RB confirm that the two dyes behave differently with the adsorbent surface.

Overall, the thermodynamic outcomes reveal that the adsorption of MB and RB onto BCC-PAA beads is spontaneous and exothermic. The reduction in K_e° with increasing temperature indicates that lower temperatures favor adsorption. The ΔH° values suggest that adsorption is primarily governed by physical interactions, while the contrasting ΔS° values reveal differences in the molecular arrangement of MB and RB molecules during adsorption. These findings are in good agreement with the equilibrium adsorption behavior observed in the isotherm and kinetic studies.

4.7. Reusability test

Fig. 8c presents the results of the reusability test over nine adsorption–regeneration cycles. Both dyes exhibited a gradual decrease in removal efficiency with increasing cycle number. For MB, the removal efficiency decreased from 95.53% in the first cycle to 88.02% after five cycles and further to 76.42% after the ninth cycle, corresponding to an overall decrease of approximately 19.1 percentage points. Similarly, RB removal decreased from 98.62% initially to 91.97% after five cycles and 77.34% after nine cycles, corresponding to an overall decrease of approximately 21.6 percentage points. Thus, both dyes retained more than 75% of their initial removal efficiency after nine cycles, although a gradual loss in adsorption performance was observed.

The observed decrease may result from a combination of incomplete desorption of the previously adsorbed dye, progressive occupation or blocking of adsorption sites, and possible changes in the hydrogel structure during repeated acid-assisted regeneration. However, the present reusability data alone do not allow these individual contributions to be distinguished. Therefore, the decline in adsorption performance is discussed as an overall loss in regeneration efficiency rather than being assigned to a specific mechanism. Nevertheless, the retention of more than 75% of the initial removal efficiency after nine cycles indicates that the BCC-PAA beads maintain a considerable fraction of their adsorption performance during repeated use.

4.8. Impact of salt concentration on adsorption

The presence of coexisting ions can significantly influence the adsorption performance of materials in real wastewater systems. The results of CaCl₂ and NaCl concentrations are shown in Fig. 8d and e, respectively. The presence of both salts leads to a gradual decline in RE% for both dyes; however, the magnitude of this effect depends on the type and concentration of the ions. In the case of Na⁺, the adsorption of MB shows only a slight decrease from 95.53% at 0 mM to 90.01% at 100 mM, corresponding to an overall reduction of approximately 5.5%. Similarly, RB removal decreases from 98.62% to 87.46%, representing a loss of about 11%. The relatively small decline indicates that adsorption remains highly effective even at elevated Na⁺ concentrations. This behavior can be attributed to the electrostatic shielding effect, where increasing Na⁺ concentration compresses the electrical double layer and reduces the electrostatic attraction between the negatively charged adsorbent surface and ammonium groups on MB and RB. Despite this, the limited reduction suggests that non-electrostatic interactions, such as π–π interactions and hydrogen bonding within the biochar-polymer matrix, maintain adsorption efficiency.

In contrast, the presence of Ca²⁺ exerts a significantly stronger influence on adsorption performance. For MB, the removal efficiency decreases from 95.53% to 75.75%, indicating a higher reduction of ∼20% across the studied concentration range. For RB removal, it drops sharply from 98.62% to 64.30%, corresponding to a total loss of approximately 34%. This marked decline is attributed to the higher charge density of Ca²⁺ ions, which enhances electrostatic screening and more effectively compresses the electrical double layer. Additionally, Ca²⁺ ions can compete with dye molecules for active sites and potentially form ionic bridges with negatively charged functional groups (–COO⁻), thereby blocking adsorption sites and reducing dye uptake.

To further assess the influence of higher-valence cations, FeCl₃ was evaluated at 20 mM as a representative trivalent ion. The presence of Fe³⁺ resulted in removal efficiencies of 86.32% and 81.65% for MB and RB, respectively, compared with 91.75% and 92.00% for Ca²⁺ and 93.98% and 96.49% for Na⁺ at the same concentration. Thus, Fe³⁺ produced the greatest reduction in adsorption, particularly for RB, consistent with its higher charge density and stronger interaction with carboxylate groups within the BCC-PAA network. These results further support the increasing inhibitory effect of multivalent cations on dye adsorption by reducing the availability of active sites.

The effect of a divalent anion was additionally examined using Na₂SO₄ (20 mM) and compared with NaCl. Na₂SO₄ reduced the removal efficiencies to 90.53% for MB and 90.11% for RB, compared with 93.98% and 96.49%, respectively, in the presence of NaCl. This moderate reduction can be attributed to changes in ionic strength and electrostatic interactions within the adsorption system, which may shield the cationic dye species and partially reduce their electrostatic attraction toward the anionic active sites on the adsorbent surface. Overall, the results indicate that cation valence has a more pronounced effect on adsorption performance than anion valence under the investigated conditions.

Comparatively, the inhibitory effect of the cations increased with increasing valence, following the order Fe³⁺ > Ca²⁺ > Na⁺ under the investigated conditions. This trend associated with stronger electrostatic screening and competitive interactions of multivalent cations with negatively charged functional groups on the BCC-PAA surface. In comparison, replacing Cl⁻ with the divalent SO₄²⁻ ion resulted in a moderate reduction in removal efficiency, indicating that anion identity also influences adsorption, although its effect was less pronounced than that of cation valence.

Overall, the results demonstrate that the adsorption efficiency of BCC-PAA beads was influenced by the type and valence of the studied ions, with the tested multivalent cations producing a more pronounced effect under the experimental conditions.

4.9. Mechanism of adsorption

To discover the adsorption mechanism of dye molecules into the BCC-PAA beads, the potential interactions were depicted in Fig. 9.

Fig. 9. Ponential interaction between dye molecules and BCC-PAA beads.

The adsorption mechanism of MB and RB onto BCC-PAA hydrogel beads is governed by a multi-step and multi-interaction process, as suggested by the kinetic, isotherm, salt effect, and structural analyses (FTIR). Initially, dye molecules are rapidly transferred from the bulk solution to the surface of the adsorbent via liquid film diffusion, driven by a strong concentration gradient. His is consistent with the good fit of the PFO model (R² ≈ 0.995 for MB and 0.990 for RB), suggesting that physisorption may dominate the early stage of adsorption. Subsequently, adsorption may progress through pore diffusion, as suggested by the IPD model and the strong agreement with the Bangham model (R² ≈ 0.996–0.999), suggesting that dye molecules may penetrate the porous biochar-hydrogel network. As adsorption proceeds, the process gradually approaches equilibrium, where pore filling and surface saturation may become increasingly important.

The isotherm results further suggest that the adsorption behavior may be reasonably described by Langmuir-type behavior, as indicated by the strong fits of the Langmuir model (R² ≈ 0.984–0.987), while the Freundlich (R² ≈ 0.943–0.947) and Koble–Corrigan (R² ≈ 0.983–0.985) models suggest that surface heterogeneity may also contribute to the overall adsorption behavior. These results suggest that adsorption may initially favor relatively similar adsorption sites, while contributions from surface heterogeneity and different adsorption energies may become more pronounced at higher concentrations.

At the molecular level (Fig. 9), the adsorption is driven by a combination of E-attraction, π–π stacking, H-bonding, and secondary interactions. Under optimised pH conditions, the BCC-PAA surface is predominantly negatively charged (pH_PZC = 5.5) due to the deprotonation of –COOH groups into –COO⁻, which promotes strong E-attraction with the cationic dye molecules (MB and RB). The aromatic structures of both dyes interact with the carbonaceous domains of biochar via π–π stacking interactions, enhancing adsorption stability, particularly for RB due to its larger conjugated system. Additionally, hydrogen bonding occurs between the oxygen-containing functional groups (–OH, –COOH) of the adsorbent and heteroatoms (N, S) present in the dye molecules. The presence of Yoshida-type hydrogen bonding (π–H interactions) further strengthens the interaction between aromatic rings and functional groups.

The influence of ionic composition provides further insight into the adsorption behavior. Under the investigated conditions, the tested multivalent cations produced a greater reduction in dye removal than Na⁺, with the inhibitory effect generally increasing in the order Fe³⁺ > Ca²⁺ > Na⁺. This trend suggests that cation valence and charge screening may influence the adsorption process, particularly through interactions with the carboxylate groups of the PAA component. The additional effect observed with SO₄²⁻ compared with the tested chloride conditions also suggests that anion identity may influence adsorption. However, the specific contributions of electrostatic interactions, ion competition, charge screening, and possible ion bridging requires further assesment.

Overall, the adsorption of MB and RB onto BCC-PAA beads is governed by a synergistic mechanism. This multi-interaction pathway explains the high adsorption capacity, strong affinity, and robustness of the developed adsorbent system.

4.10. Comparision with previous reports

The adsorption performance of the synthesized BCC-PAA beads was benchmarked against documented adsorbents using a normalized multi-criteria approach (Fig. 10 and Table 4). Five crucial performance variables, namely q_max, C_i, adsorbent dosage, equilibrium time, and optimized pH, were nominated for comparison. All variables were normalized to a 0–1 scale, with elevated values signifying stronger adsorption performance. Direct normalization was employed for adsorption capacity since higher q_max values indicate superior adsorption potential. Conversely, inverse normalization was applied for C_i, adsorbent dosage, and equilibrium time, as lower values are commonly favored for technical and economic objectives. The pH variable was normalized to align with neutral conditions (pH 7), which implies the stronger relevance of adsorbents' operation under environmentally pertinent conditions (Tables S1 and S2).

Fig. 10. Normalized radar plots comparing the adsorption performance of BCC-PAA beads with previously reported adsorbents for MB and RB removal.

Table 4. Comparision of previous studies.

	Dye	q max (mg g−1)	Ref
BCC-PAA hydrogel bead	MB	676	This work
Wood pinus caribaea	MB	149	50
Pea shells ZnCl2	MB	246.9	51
Polyacrylamide/cellulose nanocrystals	MB	326	52
Hing gum-based hydrogel	MB	284	53
Melamin@Char/PAA gel	MB	638	14
CaCO3-encapsulated CA	MB	397.9	54
BCC-PAA hydrogel bead	RB	725	This work
Chitosan mesoporous adsorbent	RB	217	55
Gum ghatti-cross-linked- polyacrylamide	RB	421	56
CS graft poly (AA-acrylamide-methyl propane sulfonic acid) hydrogel	RB	556	57
PVA/carboxymethyl chitosan hydrogel	RB	15	58
Mg–Al LDH/NaCMC/sodium alginate	RB	59	59
(BIAPEHB)/P (AA-co-AM) composite	RB	312	60

Fig. 10 demonstrates that BCC-PAA beads unveiled the best-balanced and optimal overall performance for both MB and RB removal among the compared adsorbents. The exceptional performance was principally ascribed to the remarkably high adsorption capacities (676.58 mg g⁻¹ for MB and 725 mg g⁻¹ for RB), which significantly surpass previous adsorbents. Furthermore, the developed beads achieved these adsorption capacities at comparatively low adsorbent dosage (0.1 g L⁻¹) and relatively moderate time interval (1.5 h), demonstrating efficient occupation of active sites within rapid adsorption kinetics. The radar graph further indicates that the dominance of BCC-PAA beads is attributable to a synergistic combination of adsorption efficacy, time-saving operation, and applicability under a near-neutral environment. These findings confirm the strong potential of the developed biocomposite beads as high-performance adsorbents for the removal of studied dye contaminants from aqueous systems. The superior adsorption performance of BCC-PAA beads can be attributed to the synergistic integration of BCC and PAA hydrogel network, which provides abundant functional groups (–COOH/–COO⁻), carbonaceous domains and a porous structure that facilitates dye diffusion and interaction. Additionally, the coexistence of multiple interaction mechanisms, including E-attraction, π–π interactions, and H-bonding, enhances the overall adsorption capacity.

5. Conclusion

In this study, BCC-PAA hydrogel beads were successfully synthesized and evaluated as an efficient adsorbent for the removal of cationic dyes, namely methylene blue and rhodamine B, from aqueous solutions. The integration of bacterial cellulose-derived biochar within a polyacrylic acid hydrogel matrix resulted in a porous structure enriched with carbonaceous domains, enhancing adsorption performance. Under optimized conditions (low adsorbent dose, short contact time, and controlled pH), the beads exhibited high removal efficiencies, achieving maximum capacities of 676.58 mg g⁻¹ (MB) and 725.74 mg g⁻¹ (RB) in batch experiments.

Kinetic analysis suggested that the adsorption process follows a multi-step mechanism, with rapid initial surface adsorption governed by film diffusion, followed by intra-particle diffusion into the porous structure. The good fits obtained with the PFO model (R² ≈ 0.995 for MB and 0.990 for RB) and the Bangham model (R² ≈ 0.996 and 0.999, respectively) suggest that physisorption and pore diffusion may contribute substantially to the adsorption process, while chemisorption may play a secondary role. Isotherm analysis suggested that the adsorption behavior was predominantly consistent with Langmuir-type behavior, as indicated by the Langmuir model (R² ≈ 0.984 and 0.987), while the Freundlich and Koble–Corrigan models suggested possible contributions from surface heterogeneity. Considering the isotherm, kinetic, and characterization results, the adsorption behavior may involve a combination of electrostatic interactions, π–π interactions, and hydrogen bonding, which may contribute to the observed affinity toward both dyes.

The adsorbent exhibited relatively good tolerance toward ionic conditions, although its adsorption performance decreased to varying degrees in the presence of mono-, di-, and trivalent ions, with the greatest reduction observed for Fe³⁺ under the investigated conditions. Reusability studies demonstrated that the beads retain substantial adsorption efficiency after multiple cycles, maintaining ∼76.42% (MB) and ∼77.34% (RB) removal after nine cycles, confirming acceptable regeneration capability.

Overall, the BCC-PAA hydrogel beads exhibit high adsorption capacity, favorable kinetics, and reasonable reusability, competing with many previously reported hydrogel-based adsorbents. These properties highlight their strong potential as affordable and efficient materials for dye-contaminated wastewater treatment, particularly in systems where high adsorption capacity and multifunctional interaction mechanisms are essential.

Author contributions

E. M.: methodology, conceptualization, experimental, data analysis, investigation, validation, writing – original draft, review & editing, F. A. C.: data validation, review & editing, A. B.: conceptualization, methodology, data validation, writing – review & editing, funding acquisition, supervision, instrumental facilities.

Conflicts of interest

The other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data can be made available on request.

Supplementary information (SI): comparative adsorption performance data for BCC-PAA beads and previously reported adsorbents, including normalized performance parameters used for multi-criteria comparison. See DOI: https://doi.org/10.1039/d6ra05858b.