Development of a Chestnut Shell Bio-Adsorbent for Cationic Pollutants: Encapsulation in an Alginate Carrier for Application in a Flow System

Abstract

Melanin-based biosorbents (MiCS), derived from chestnut shells, were encapsulated in sodium alginate to obtain MiCS@Alg, useful in a column adsorption study. MiCS contains various acidic surface groups able to participate in the removal of cationic pollutants from aqueous solutions. The MiCS and MiCS@Alg were characterized by Fourier-transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), and Dynamic Light Scattering (DLS), while zeta potential and particle size analyses were performed to gain deeper insight into surface charge behavior. Batch adsorption experiments were carried out at three different temperatures, demonstrating that the adsorption kinetics followed a pseudo-second-order (PSO) model and that the Freundlich model best described the equilibrium data. The process was found to be endothermic and spontaneous, with maximum adsorption capacities of 300.2 mg g⁻¹ (BR2), 201.5 mg g⁻¹ (BY28) and 73.08 mg g⁻¹ (NH₃) on MiCS, and 189.3 mg g⁻¹ (BR2), 117.1 mg g⁻¹ (BY28) and 50.06 mg g⁻¹ (NH₃) on MiCS@Alg at 45 °C and compared with the unmodified chestnut shell. The MiCS and MiCS@Alg exhibited good adsorption performance, improved environmental compatibility, and greater reusability. Overall, these results highlight MiCS@Alg as a cost-effective, sustainable, and highly promising novel biosorbent for the removal of cationic pollutants (BR2, BY28, and NH₃) from water.

1. Introduction

Water pollution by synthetic compounds, particularly pharmaceutical pollutants and dyes containing amine functional groups, represents one of the most significant challenges in environmental protection. These compounds, including antidepressants (e.g., amitriptyline [1]), beta-blockers (e.g., atenolol), antibiotics (e.g., amoxicillin), and various cationic dyes (such as methylene blue [2]), are often not effectively removed by conventional wastewater treatment methods, leading to their accumulation in aquatic ecosystems and threatening human health and biodiversity. According to data from UNICEF and the World Health Organization (WHO), in 2020, two billion people lacked access to safe drinking water (WHO and UNICEF 2021) [3]. Yaseen and Scholz (2019) [4] report that dye concentrations in textile wastewater can range from 10 to 7000 mg L⁻¹, depending on the type of fabric and dye used during the manufacturing process. Consequently, the efficient treatment of dye-containing wastewater is essential before its discharge into aquatic environments [5]. Colored effluents are released by numerous industries, including textile, dyeing, food, pharmaceutical, paper, leather, and cosmetics sectors [6]. Among these pollutants, Safranin T (BR2) is a widely used azine dye. In addition to its application in the textile and leather industries, BR2 also finds use as a photosensitizer, biological stain, and fluorescent probe [7]. However, exposure to BR2 poses serious health risks, such as eye irritation, skin inflammation, and respiratory tract disorders. Owing to its extensive industrial use in textiles, plastics, paper, and cosmetics, the removal of BR2 from industrial effluents and wastewater is of urgent importance for ensuring a safer and healthier environment. Alongside dyes, ammonia (NH₃/NH₄⁺) represents another major contaminant in municipal and industrial wastewaters, originating from agriculture, fertilizer production, and domestic effluents. High concentrations of ammonia can cause eutrophication, oxygen depletion, and toxicity to aquatic organisms, thereby creating significant ecological and regulatory concerns.

Conventional biological treatments, such as nitrification–denitrification, are commonly employed but are often limited by high operational costs, long retention times, and sensitivity to pH and temperature fluctuations. In this regard, adsorption has emerged as a particularly attractive alternative due to its simplicity, efficiency, and potential for adsorbent regeneration. A wide variety of adsorbents [8] have been investigated for dye and ammonia removal, including zeolites, activated carbon [9], and biochar [10,11]. Among them, bio-based materials stand out as sustainable and cost-effective options. Within this category, melanin-based sorbents derived from chestnut shells (MiCS) and their alginate-modified derivatives (MiCS@Alg) show considerable potential, offering abundant functional groups and environmentally friendly synthesis routes. In the search for efficient, environmentally friendly, and cost-effective wastewater treatment solutions, the application of adsorption processes using biosorbents made from plant waste materials is gaining increasing importance. One such material is chestnut shell, an agro-industrial waste generated in large quantities during the peeling process, accounting for 10% of the total mass. In 2020, global chestnut production reached 2.322 million tons, distributed across many countries worldwide [12]. Chestnut shells are rich in lignocellulosic components and contain various functional groups (–OH, –COOH), enabling interaction with water contaminants. Like other agricultural wastes, chestnut shells have been evaluated for the adsorption of dye molecules and metal ions, either in their natural form or after carbonization and activation, with quite satisfactory results [13]. Their natural adsorption properties are often insufficient for the efficient removal of pollutants, especially those with specific functional groups such as amines. Main compounds in chestnut shell were found to be holocellulose 42.4 wt.% (27.8 wt.% of α-cellulose), 39.8 wt.% lignin, and extractive content was 3.2 wt.% [extraction agent MeOH/Water (95:5 v/v)] [14]. Therefore, chemical modification of natural biosorbents is necessary, offering an effective approach to enhancing their adsorption capacity. In this study, sodium alginate was used as a surface-modifying agent for melanin isolated from chestnut shells. By cross-linking with melanin particles, a higher number of functional groups (e.g., carboxyl and hydroxyl groups) was introduced, enabling the formation of microbeads with stronger and more selective interactions with cationic pollutants. Moreover, sodium alginate (Na-alg) is rich in carboxylate groups (–COOH), which facilitate chelation with a wide range of multivalent metal ions, forming stable “egg-box” structures [15,16]. This property makes it particularly suitable for the removal of heavy metals and cationic pollutants from water, such as dyes Astrazon Yellow 7GLL (BY28) and Safranin T (BR2). To improve its mechanical stability and adsorption efficiency, sodium alginate is frequently modified into various composite forms. For example, Na-alg has been successfully combined with chitin, humic acid, polyaniline, cellulose, and even carbon nanotubes, significantly enhancing its adsorption capacity and applicability in wastewater treatment. Recent studies have increasingly focused on the development of sorbents designed for the removal of carbon dioxide (CO₂), one of the major air pollutants that significantly affects climate change and human health. In this study [17], a mesoporous silica foam (MSF) was used as a support, prepared from coal fly ash (CFA), providing a cost-effective porous material while simultaneously enabling the recycling of silicoaluminate solid waste.

This research aimed to develop and characterize a new biosorbent derived from raw chestnut shell, initially in the form of isolated melanin, and subsequently encapsulated in alginate beads to obtain spherical microstructures to serve as a carrier material for the removal of cationic dyes from aqueous solutions. Adsorption capacity, sorption mechanisms, as well as the stability and reusability of the material, were examined through a series of experiments, with a particular focus on its potential application in the treatment of textile wastewater. In addition to its chemical advantages, the particle shape in the form of microgranules helps to overcome common issues associated with powdered biosorbents, such as small particle size, difficulty in separating the sorbent from the solution after adsorption, and limited potential for reuse. The prepared composites exhibited good stability in aqueous environments and a structure suitable for continuous pollutant removal processes. The materials characterization was performed to confirm the success of material purification and morphological changes. The adsorption performance of the new sorbent was evaluated through a series of batch experiments, including adsorption isotherm and kinetic analyses under varying temperature and pH conditions, with particular emphasis on interactions with cationic pollutants in aqueous systems.

2. Materials and Methods

2.1. Chemicals and Materials

Chestnut shells (Castanea mollissima) [15] were collected in Vršac, Serbia (Supplementary Materials S2.1), and the pericarp (outer shell) was separated and used in this study. Sodium alginate (Na-alg), copper sulfate (CuSO₄), calcium chloride (CaCl₂), Folin–Ciocalteu’s phenol reagent (2N, ACS), potassium hexacyanoferrate(III) (≥99%, ACS), ammonium hydroxide (~25%), gallic acid (C₇H₆O₅, 97.5–102.5%, p.a.), iron(III) sulfate pentahydrate (99.9%, trace metals), butyl alcohol (CH₃(CH₂)₃OH, ≥99.5%), potassium hydroxide (KOH, 90%, flakes), sodium hydrogen carbonate (NaHCO₃, ≥99.7%, ACS), sodium chloride (≥99%), sodium sulphate (≥99%), magnesium sulphate (≥99%), aluminium sulphate (99.99%), methylene blue (≥97%), and sodium benzoate (99%) were purchased from Sigma-Aldrich (Merck), Darmstadt, Germany. Ethanol, hydrochloric acid (HCl, 38% w/w, AG), and sodium hydroxide (NaOH) were obtained from Alkaloid, Skopje, Macedonia. Nitric acid (65% HNO₃) was purchased from Zorka Pharma a.d., Šabac, Serbia. Tannic acid (99%, ACS), potassium nitrate (KNO₃, ≥99.0%), and bovine serum albumin (≥98%, p.a.) were obtained from Merck, Darmstadt, Germany. All chemicals were of analytical grade and used without further purification. All solutions were prepared with distilled water purified using a Simplicity^® UV water purification system (Merck Millipore, Darmstadt, Germany). Dye solutions with different concentrations were prepared in deionized water. Astrazon Yellow 7GLL (C.I. Basic Yellow 28, BY28) and Safranin T (C.I. Basic Red 2, BR2) were purchased from Sigma-Aldrich, Darmstadt, Germany.

2.2. Extraction of Melanin from Integument Chestnut Shell (MiCS)

Pulverized chestnut shell (iCS, 70 g) was subjected to alkaline extraction using 0.5 M NaOH at a solid-to-liquid ratio of 1:1.15 (w/v), in a 1 L glass beaker with continuous stirring for 12 h. The mixture was then treated in an ultrasonic bath at 40 °C for 2 h. The resulting suspension was centrifuged for 15 min at 5000 rpm to separate the phases. The supernatant was further purified from proteins using the Sevag method. The obtained effluent was acidified to pH 2.5 using 2 M HCl and left to stand for 12 h. The precipitated melanin was then left at room temperature to dry for three days and subsequently used in the following experimental procedures. Following drying, the melanin precipitate was dissolved in an appropriate volume of 1.25% NaOH solution (200 mL) to attain a final pH < 8 [18]. The obtained alkaline extract was then treated with an equal volume of 8 wt.% CuSO₄·5H₂O solution [19] with continuous stirring for 30 min at room temperature. The resulting dark melanin precipitate was separated by centrifugation, thoroughly washed with distilled water, and treated with 200 mL 2.5% HCl, filtered, and washed with 250 mL deionized water (DW). The re-dispersion of wet products in 200 mL of DW using ultrasound for 10 min, and filtration with 250 mL of DW (two cycles applied), and freeze-dried (frozen at −50 °C for 2 h and lyophilized overnight at −70 °C) gave 15.6 g of product (yield 22.3 wt.%). All operation was performed in an inert atmosphere. The final water-insoluble melanin brown powder was used for further characterization and evaluation of its adsorptive properties. Residual Cu²⁺ ion was determined first by microwave digestion and atomic absorption spectroscopy (residual Cu²⁺ ion was found to be less than 0.01%).

2.3. Preparation of MiCS Encapsulated in Alginate Particles (MiCS@Alg)

The melanin obtained in powder form was used for the synthesis of MiCS@Alg, as illustrated in Scheme 1. Dissolution of sodium alginate (2.0 g) in 100 mL of deionized water was followed by the addition of 2.0 g MiCS, and stirring until a homogeneous dispersion was obtained [20]. Na-alg hydrogel beads were formed by dropping the produced dispersion into a CaCl₂ solution (2%, w/v). The resulting Na-alg beads were further immersed in the CaCl₂ solution for an additional 30 min to ensure complete cross-linking. After cross-linking, the beads were frozen at −50 °C for 2 h and lyophilized overnight at −70 °C. The resulting lyophilized beads were used for further characterization and adsorption experiments. The yield of MiCS@Alg was 33%. The results show that 77% of the free volume of MiCS@Alg was available for the filling/wetting with adsorption medium, providing a large free volume available for pollutant transport.

2.4. Characterization and Analysis

The structural and physicochemical properties of all samples were characterized using scanning electron microscopy (SEM), Fourier-transform infrared spectroscopy (FTIR), UV–Vis spectrophotometry, dynamic light scattering (DLS), and laser Doppler electrophoresis (LDE) with a Zetasizer Nano ZS and a Multi-Purpose Titrator MPT-2, as well as elemental analysis. Detailed results are provided in Supplementary Material S2.4.

Other characterization and analytical methods, such as the determination of specific density, total polyphenol content (TPC) in chestnut shells, total tannin content (TTC), protein analysis, Bradford assay, enzymatic deproteinization, and “dry-wet mass” method for porosity, are presented in the Supplementary Materials S2.4. The methods for determining total basic, total acidic, and carboxyl groups are provided in Supplementary Materials S2.4.1.

The collected filtrate, after iCS purification (Section 2.2. Extraction of melanin from integument chestnut shell (MiCS)), was analyzed for copper presence using atomic absorption spectrometry (AAS), Perkin Elmer PinAAcle 900T, PerkinElmer, Inc., Shelton, CT, USA.

Determination of chemical oxygen demand (COD) value: to assess adsorption success, the study referred to Serbian national emission limit values: COD = 200 mg O₂ L⁻¹ (satisfies the criteria prescribed by the Official Gazette of the Republic of Serbia, No. 67/2011, 48/2012, and 1/2016) [21]. The ISO 6060 method determined chemical oxygen demand (COD) using a Lovibond MD 600 spectrophotometer and Lovibond RD 125 sample heating apparatus [22]. Spectrophotometric tests, with mercury(II) sulfate masking chloride, were conducted for COD determination.

2.5. Adsorption Experiments

The adsorption behavior of the new materials was investigated using the batch and column methods (MiCS@Alg). Briefly, varying masses of the adsorbent (1, 2.5, 5, 7.5, and 10 mg) were added into 10 mL glass conical flasks containing 8 mL of dye solution with an initial concentration of 30 mg L⁻¹. The flasks were then placed on a digitally heated magnetic stirrer (Thermo Scientific, Waltham, MA, USA) and stirred at a constant rate for 90 min to reach equilibrium. In an experiment of ammonia removal, plastic centrifuge tubes of 50 mL and varying masses of the adsorbent (5, 12.5, 25, 37.5, and 50 mg) were used. The shaking was provided using the Thermo Scientific Thermal mixer and Blocks. The temperature was maintained at 25 °C, 35 °C, or 45 °C, depending on the experimental condition, while the pH was adjusted to the optimal value of 7.5 for dyes removal and pH 4 for ammonia (Section 3.3.1 Adsorption isotherm study). pH was adjusted using either 1 mol L⁻¹ HCl or 1 mol L⁻¹ NaOH solution. After equilibrium was reached, the MiCS@Alg and MiCS adsorbents were separated either by filtration through quantitative Whatman filter paper or centrifugation for 15 min at 11,000 rpm, respectively. The residual concentration of pollutants in the filtrate was determined using a UV–Vis spectrophotometer in absorption peak intensity at 518 nm and 435 nm (λmax of BR2 and BY28, respectively), as shown in Scheme 2. The residual ammonia concentration in aqueous samples was determined using the standard boric acid absorption–titration method following alkaline distillation (modified Kjeldahl procedure) [23]. To describe the experimental data and determine the nature of the adsorption process, commonly used isotherm models, as well as various kinetic models adopted from relevant literature, were employed [2]. Ionic strength influences on adsorption efficiency were performed in the presence of KNO₃ (0.01, 0.05, and 0.1 mol dm⁻³). Each experiment was run in triplicate, and mean values were calculated. The chemical structure of the dyes is presented in Figure 1. The design of adsorption experiments is presented in the Supplementary Information, in the section on Supplementary Materials S2.7. Response Surface Methodology (RSM).

In order to test the applicability of MiCS and MiCS@Alg adsorbents, a higher pollutant concentration—e.g., C_i of 200 mg dm⁻³—of the dyes and ammonia was used, and adsorption experiments were performed at the following conditions: C_i = 30 mg L⁻¹, m = 1 mg, V = 8 mL, T = 25 °C, and t = 90 min.

Competitive adsorption experiments were performed in a two-component system using Ca²⁺, Al³⁺, Cl⁻, SO₄²⁻, methylene blue (MB), and sodium benzoate at 30 mg dm⁻³ of initial concentration.

For column adsorption (Supplementary Materials S2.5), data analysis was simplified using the Bohart–Adams and Yoon–Nelson models. The test solution was prepared by passing BR2 and BY28 dyes or ammonia (5 mg L⁻¹) through model water adjusted to pH 7.5 for the dyes and pH 4 for ammonia, at controlled flow rates of 0.5, 1.0, and 1.5 mL min⁻¹.

2.6. Desorption Experiments

Desorption and reuse of adsorbents were examined using a batch method to evaluate their efficiency in repeated adsorption of BY28, BR2 dyes, and NH₃ from wastewater. The experiments were performed according to the procedure described in [24]. Each test was carried out separately using 0.1 M HCl as the desorption agent for 3 h under agitation at 150 rpm to regenerate the surface of MiCS and MiCS@Alg adsorbents loaded with pollutants. Before reuse, each adsorbent was thoroughly rinsed with distilled water. This procedure was repeated for five consecutive cycles.

3. Results

3.1. Characterization of Materials

In this work, an attempt is made to isolate melanin and encapsulate it into Na-alg matrix to obtain applicable adsorbents in both batch and flow systems. Chemical and elemental analysis (Table S2) of pulverized chestnut shell (iCS) showed the presence of holocellulose, mainly α-cellulose, and klason lignin. After chestnut shell purification—i.e., melanin isolation from cellulose and lignin—significant increases in the total basic group, from 0.28 to 0.71 mmol g⁻¹, the total acidic group, from 1.32 to 3.06 mmol g⁻¹, and the carboxyl group, from 0.89 to 2.14 mmol g⁻¹, indicate high potential of MiCS for cationic pollutants removal. Recording of the NMR spectra of the MiCS material (dissolved in DMSO-d6) was not possible, since melanin possesses a stable population of free radicals distributed across its aromatic structures, as confirmed by the Electron Spin Resonance (ESR) spectrum [25].

3.1.1. SEM Analysis

SEM micrographs of iCS, MiCS, and the surface of MiCS@Alg are given in Figure 2.

SEM was employed to investigate the surface morphology of raw chestnut fruit shells, the isolated melanin, and the Na-alg-based formed beads. As shown in Figure 2a, irregularly shaped and sized fragments with a rough and porous surface were observed. The particles exhibited a layered and heterogeneous texture, which is characteristic of lignocellulosic plant residues rich in lignin, cellulose, and hemicellulose, as presented in the study [13]. The pronounced surface damage can be attributed to the mechanical grinding applied as part of the pretreatment using a laboratory mill. In Figure 2b, the surface of the MiCS displays narrowed slits and a wavy, layered structure with an amorphous appearance [12]. At higher magnifications, a compact amorphous morphology with low porosity becomes evident, which is typical of melanin derived from plant-based materials [26]. Figure 2c presents a cross-section of the formed beads MiCS@Alg, showing a homogeneous and smooth surface, indicating successful alginate cross-linking. Similar to the melanin sample, the internal structure appears dense and non-porous, with no pronounced cavities, suggesting uniform distribution within the gel matrix. At higher magnification, traces of dispersed particles, most likely melanin, can be observed in the central region of the bead. Figure 3 presents SEM micrographs of bead cross-sections, along with their morphological characteristics and diameters. The bead diameters were determined using a digital optical microscope (Delta Optical Smart 5.0 MP PRO) equipped with HiView (HiRISE) software, version 1.5.0.

SEM micrographs (Figure 3a), taken at magnifications from 500 µm down to 10 µm, reveal the highly porous internal architecture of the MiCS@Alg beads. The cross-sections reveal interconnected macropores with irregular walls, while higher magnifications reveal a rough, layered morphology with micro-fissures, indicating an enhanced surface area that is favorable for adsorption. A complementary analysis (Figure 3b) was performed on a large number of beads to obtain accurate size measurements, and the size-distribution characteristics were determined by processing digital images of the samples. The synthesized composites appeared as small black beads, forming nearly monodisperse spheres of varying diameters. SEM observations confirmed that the lyophilized beads retained a spherical shape with a rough surface and no visible signs of collapse, with an average diameter of 2.394 ± 0.260 mm. Volume shrinkage caused by water evaporation during lyophilization inevitably led to slight shape deformation, and the observed irregularities in sphericity can be attributed to this effect. Similar findings have been reported for Argan-nutshell-based beads used for methylene-blue removal [6]. The result of the “dry-wet mass” method revealed that the determined bead porosity was about 76.2%. The density, determined to be 0.19 g mL⁻¹, was calculated following the procedure reported by Popovic et al. [27].

3.1.2. FTIR Analysis

FTIR spectra of the iCS, MiCS, Na-alg, and MiCS@Alg are given in Figure 4.

The broad absorption band at 3255 cm⁻¹ indicates the presence of O–H groups, characteristic of hydroxyl moieties in cellulose, hemicellulose, and lignin. The bands in the region between 2929 and 2850 cm⁻¹ correspond to symmetric and asymmetric C–H stretching vibrations of methyl and methylene groups, commonly found in lignocellulosic structures. A moderate peak at 1734 cm⁻¹, observed in the spectrum of the raw sample (iCS), is assigned to ester and carbonyl (C=O) groups, which are mainly associated with hemicellulose and/or pectin [28]. The disappearance of this peak in the MiCS sample after melanin extraction suggests the removal or hydrolysis of the ester group in the course of melanin isolation. In any case, a significant number of carboxylic groups are still present in MiCS (Table S3). The iCS spectrum also displays a prominent band at 1014 cm⁻¹, which can be attributed to C–O stretching vibrations of phenolic compounds, indicating the presence of natural pigments. Similar melanin spectra have been reported in a study comparing standard and extracted pigments [29]. In addition, bending vibrations of aliphatic groups and aromatic ring deformations were observed at 1435 and 1281 cm⁻¹, respectively, confirming the presence of lignin-related structures. In the MiCS sample (blue line), the O–H stretching band shifts and decreases in intensity (from 3255 to 2929 cm⁻¹), indicating chemical modification of the material and a reduction in free hydroxyl groups. The transmission at 1605 cm⁻¹ represents aromatic C=C stretching vibrations (skeletal stretching), suggesting an increased content of aromatic structure after treatment. The band at 1435 cm⁻¹ is also related to deformation vibrations of aliphatic moieties. Following acid extraction, the intensity of the peak at 1085 cm⁻¹, associated with cellulosic material, is significantly reduced in the MiCS@Alg spectrum, confirming successful cellulose removal from the matrix. The Na-alg spectrum showed characteristic asymmetric and symmetric vibrations of the carboxylate anion at 1605 and 1405 cm⁻¹, respectively.

The spectrum of MiCS@Alg shows multiple absorption peaks, including a broad band around 3255 cm⁻¹, assigned to O–H stretching vibrations of hydroxyl groups in the alginate and melanine structure [30]. The peak at 1710 cm⁻¹, resulting from C=O stretching, is related to the carboxyl groups of the melamine structure. An additional intense peak at 1605 and 1415 cm⁻¹ further confirms the presence of carboxylate anion. The bands at 1281, 1085, and 1014 cm⁻¹ correspond to C-H and C-O deformation vibrations, characteristic of polysaccharide structures. Finally, the peaks at 997 and 815 cm⁻¹ can be attributed to C–O–C vibrations and glycosidic linkages in the ring structures of alginate.

Natural materials MiCS showed structural complexity of constituent material, which indicates that the adsorption mechanisms are very demanding. From that point of view, it is of utmost importance to quantify surface functionalities to discuss the relation between adsorption performance versus adsorbent properties. Melanin, lignin, and tannins, present in the chestnut shell and MiCS, contain an appropriate number and type of functional groups as potential binding sites for specific pollutants. The establishment of a relationship between adsorbent properties/functionalities versus adsorption performance can be based on the quantitative determination of surface functionalities. The results from total basic and acidic groups, and carboxyl groups content determination using standard volumetric methods are given in Table S3.

3.1.3. UV-Vis Analysis

UV-Vis spectra of iCS and MiCS are given in Figure 5.

On the basis of the obtained UV–Vis spectra, an intense absorption in the UV region can be observed with maxima at 279 nm (iCS) and 283 nm (MiCS), corresponding to π–π* transitions of aromatic chromophores present in the melanin structure. Following these maxima, the absorbance gradually decreases toward longer wavelengths, displaying a continuous “tail” up to 800 nm without distinct peaks, which represents a typical characteristic of melanin and melanin-like materials known for their so-called “featureless broadband absorption” profile [31,32]. A comparison of the samples shows that MiCS exhibits slightly more pronounced absorption in the UV region, indicating a higher concentration or better dispersion of melanin pigments compared to the iCS sample. The obtained results clearly confirm that the analyzed samples exhibit the spectral characteristics of melanin, with the identification of this pigment further corroborated by its distinctive continuous absorption behavior in the visible region. A similar spectrum was reported in the study [32], where melanin was extracted from the marine sponge-associated actinomycete Micromonospora fulva HV6.

3.1.4. DLS and ZETA Analysis

The zeta potential and average particle radius were determined to better understand surface electrostatics as a function of pH. The zeta potential of the isolated melanin is −36.6 ± 2.9 mV at a pH of 10.5, while the untreated chestnut bark exhibits a value of −25.9 ± 1.04 mV at a pH of 12.8 (Figure 6). These results indicate a pronounced negative surface charge of both materials, which can significantly influence their ability to interact with cations in solution, specifically Astrazon Yellow 7GLL and Safranin T dyes. In relation to these results, a series of preliminary adsorption experiments at initial pH 6, 7, and 8 were performed, and the optimal results were obtained at pH 7.5. Higher pH affects the increase in solubility of MiCS.

The stability of the colloidal dispersions of the iCS and MiCS samples was investigated as a function of pH by determining their hydrodynamic radius using DLS analysis (Figure 6). The results clearly indicate that the aggregation behavior of the particles depends on the pH of the medium. At pH 2, the MiCS sample exhibited a huge hydrodynamic radius (4952 nm), suggesting strong particle aggregation in the acidic environment due to reduced zeta potential and destabilization of interparticle repulsion. In contrast, the iCS sample showed a significantly smaller radius at the same pH (997 nm), indicating somewhat better dispersibility.

As the pH increased to 4, a decrease in the hydrodynamic radius was observed for both samples. Although MiCS still exhibited a larger radius compared to iCS (1852 nm versus 384 nm), a significant reduction in aggregation relative to pH 2 was evident. The lowest values of hydrodynamic radius were recorded in the pH range of 6 to 10, where both samples displayed relatively small dimensions (below 500 nm), indicating the highest colloidal stability. Within this pH range, MiCS exhibited a slightly smaller radius compared to iCS, suggesting that melanin extraction contributes to enhanced dispersion stability, likely through the reduction in interparticle interactions caused by changes in surface functional groups. At higher pH values (12 and 13), a slight increase in hydrodynamic radius was observed for both samples, with aggregation more pronounced in iCS (821 nm at pH 12.6). In contrast, MiCS maintained a moderate size (612 nm), further confirming the stabilizing effect of melanin extraction. As shown in Figure 6, MiCS exhibits better dispersibility and a lower degree of aggregation across a broad pH range. The most favorable pH for stable dispersion of both samples lies within the 6–10 range, which is particularly important for potential applications in aqueous systems, including wastewater treatment. In this range, the average particle size of iCS was 786 nm at pH 5.8, while MiCS measured 612 nm at pH 10.5.

3.2. Elemental Analysis

The elemental composition of the pulverized chestnut shells and MiCS is presented in Supplementary Table S2. Compared to the raw chestnut shells, MiCS exhibited a slightly increased carbon content (49.7 wt.%) and decreased hydrogen (3.90 wt.%) and oxygen (43.21 wt.% vs. 41.83 wt.%) contents, indicating a mild concentration of organic components and partial removal of oxygen- and hydrogen-containing functional groups during the preparation process. The nitrogen content remained essentially unchanged, suggesting the preservation of proteinaceous or nitrogen-containing constituents. Overall, these changes reflect subtle modifications in the chemical composition of the material resulting from the MiCS preparation process.

3.3. Batch Adsorption Experiments

3.3.1. Adsorption Isotherm Study

Analysis of adsorption equilibrium data provides valuable information about the interaction mechanisms between the adsorbent and adsorbate. Analyzing the zeta potential measurement at different pH (Figure 6a) and the results from the study of the pH influences on dyes and ammonia adsorption efficiencies onto MiCS and MiCS@Alg (Figure S1), it was deduced that pH 7 was optimal for dyes removal, while pH 4 was optimal for ammonia removal. Low decrease in adsorption capacity of both MiCS and MiCS@Alg adsorbent indicates the significance of deprotonated phenol groups in dyes removal. Moreover, a wider pH range, from 6 to 8, could be used in processes of dye removal, while a narrower pH range, from 4 to 5, could be applied for ammonia removal without a significant decline in adsorption efficiency. This pH range can possibly be used in processes of dye removal, which falls in the range of most natural water, while one that could be applied in processes of ammonia removal could be found in wastewater from the mining industry.

In this study, adsorption experiments were conducted at three different temperatures to determine the adsorption capacity of MiCS and MiCS@Alg. The experimental data were fitted to both the Langmuir and the Freundlich isotherm models [33], which describe monolayer and multilayer adsorption on surfaces with a defined number of active sites, respectively. The corresponding model nonlinear parameters and correlation results are presented in

Table 1. Results of nonlinear Langmuir and Freundlich model for pollutant adsorption (Ci = 30 mg L⁻¹, m_a = 1–10 mg, V = 8 mL, T = 25–45 °C, t = 90 min).

Sample			Langmuir			Freundlich
	Pollutant	T (°C)	qm (mg g−1)	KL (L mg−1)	R2	KF (mg g−1) (L mg−1)1/n	n	R2
MiCS	BR2	25	291.1 ± 21.5	0.2166 ± 0.061	0.9635	56.21 ± 2.47	1.714 ± 0.08	0.9912
		35	295.7 ± 22.2	0.2247 ± 0.064	0.9632	58.36 ± 2.62	1.715 ± 0.08	0.9905
		45	300.2 ± 22.6	0.2471 ± 0.075	0.9601	62.73 ± 2.98	1.709 ± 0.08	0.9912
	BY28	25	192.5 ± 13.4	0.1864 ± 0.03	0.9829	35.67 ± 0.96	1.873 ± 0.045	0.9974
		35	198.2 ± 13.8	0.1942 ± 0.03	0.9827	37.65 ± 1.36	1.882 ± 0.074	0.9933
		45	201.5 ± 14.3	0.2154 ± 0.03	0.9857	40.40 ± 1.26	1.895 ± 0.056	0.9960
	NH3	25	71.26 ± 5.46	1.9554 ± 1.15	0.7565	38.35 ± 3.04	3.563 ± 0.501	0.9354
		35	71.94 ± 5.52	2.0242 ± 1.19	0.7558	38.98 ± 3.06	3.569 ± 0.503	0.9361
		45	73.08 ± 5.85	2.2128 ± 1.32	0.7519	40.28 ± 3.13	3.596 ± 0.504	0.9361
MiCS@Alg	BR2	25	180.5 ± 9.40	0.2351 ± 0.106	0.8829	43.09 ± 3.14	2.175 ± 0.17	0.9757
		35	184.3 ± 10.5	0.2338 ± 0.116	0.8654	44.37 ± 3.71	2.193 ± 0.205	0.9674
		45	189.3 ± 10.8	0.2853 ± 0.141	0.8741	48.91 ± 3.68	2.141 ± 0.187	0.9755
	BY28	25	109.3 ± 8.96	0.4981 ± 0.28	0.7919	39.28 ± 3.33	2.736 ± 0.301	0.9564
		35	115.2 ± 9.45	0.4566 ± 0.27	0.7872	40.24 ± 3.51	2.706 ± 0.304	0.9545
		45	117.1 ± 9.60	0.4873 ± 0.30	0.7775	42.20 ± 3.67	2.747 ± 0.316	0.9529
	NH3	25	49.78 ± 5.15	4.1601 ± 2.78	0.7809	34.24 ± 2.28	5.622 ± 1.00	0.9309
		35	49.65 ± 5.14	4.5621 ± 3.17	0.7762	34.62 ± 2.24	5.726 ± 1.01	0.9333
		45	50.06 ± 4.99	4.7934 ± 3.17	0.7882	35.13 ± 2.23	5.807 ± 1.02	0.9338

. The results of nonlinear modelling using Langmuir and the Freundlich isotherms are presented in the Supplementary, in Figure S2. Data fitting with other adsorption isotherms (Dubinin-Radushkevich, Tempkin, Sips, Toth, etc.) [34] provides modelling data at lower statistical validity.

The equilibrium adsorption data were analyzed using both the Langmuir and the Freundlich isotherm models (Table 1). The Langmuir model assumes monolayer adsorption on a homogeneous surface, whereas the Freundlich model accounts for adsorption on heterogeneous surfaces with the possibility of multilayer formation. In this study, the Freundlich model exhibited a superior fit, with R² values ranging from 0.752 to 0.997, confirming that adsorption predominantly occurs on heterogeneous surfaces, consistent with the structural complexity of the synthesized adsorbents. The Langmuir equilibrium constants (K_L) also supported these trends. For BR2, MiCS exhibited values up to 0.2471 L mg⁻¹, whereas BY28 showed moderately lower constants (0.2154 L mg⁻¹). For NH₃, however, K_L values were much higher (2.21 L mg⁻¹ for MiCS and 4.79 L mg⁻¹ for MiCS@Alg at 45 °C), suggesting a strong affinity of NH₃ molecules for active sites despite the lower q_m values, which may reflect steric limitations and weaker multilayer formation compared to dye molecules. The Freundlich constants (K_F) confirmed these observations. The highest K_F was obtained for BR2 adsorption on MiCS (62.73 mg g⁻¹ (L mg⁻¹)¹/ⁿ at 45 °C), confirming its strong adsorption capacity. For NH₃, K_F values were also substantial (40.28 mg g⁻¹ (L mg⁻¹)¹/ⁿ for MiCS and 35.13 mg g⁻¹ (L mg⁻¹)¹/ⁿ for MiCS@Alg at 45 °C), indicating favorable adsorption but less pronounced than for dyes. The Freundlich intensity factor (n) ranged from 1.70 to 5.80 across all systems, confirming favorable adsorption conditions, with NH₃ on MiCS@Alg showing the highest intensity (n = 5.81) (Table 1). Overall, the Freundlich model provided a more accurate description of the adsorption behavior, highlighting the heterogeneous surface nature of both adsorbents and the potential for multilayer adsorption, as also in the work [35]. Comparable adsorption capacities have been reported in the literature, such as the removal of BR2 using bentonite (269 mg g⁻¹) and the removal of BY28 with zinc oxide-chitosan composites [36,37]. Similarly, adsorption of ammonium ions has been extensively studied, with related findings reported in recent works [10] further validating the observed trends in this study. In order to test the applicability of both MiCS and MiCS@Alg adsorbents at higher initial pollutant concentrations at 200 mg L⁻¹ of dye and ammonia pollutants, adsorption experiments in a batch system were performed. The preliminary results confirmed the high potential of produced adsorbents, and the following adsorption capacities were obtained: 954.4 mg g⁻¹ for BR2, 686.2 mg g⁻¹ for BY28, and 176.4 mg g⁻¹ for NH₃ on MiCS; 628.4 mg g⁻¹ for BR2, 445.8 mg g⁻¹ for BY28, and 116.2 mg g⁻¹ for NH₃ on MiCS@Alg at 25 °C. In the case of BY28 dye, methyl groups contribute to the steric inhibition of both the approach and the establishment of electrostatic interactions. On the other hand, BR2 dye, with its higher aromatization and charge dispersibility in the phenazine ring, helps in adaptable access to the carboxylate anion. Without exception, the contribution of amino to the establishment of hydrogen bonding with carboxylate anions must not be neglected.

In summary, the produced adsorbents MiCS and MiCS@Alg showed significantly higher adsorption capacities in relation to iCS (preliminary results indicated between 32 and 36% of those obtained for MiCS and MiCS@Alg, as shown in Table 1). Use of MiCS as an ammonia adsorbent provides additional benefits, as the enriched materials contain approximately 10 wt.% nitrogen, which could be used in agriculture as fertilizer. Also, the obtained results refer to the possibility of MiCS use as adsorbents for removing unpleasant odors in toilets. Moreover, the production of MiCS@Alg beads opens a new path for their potential applicability in a flow system, verifying the concept of melanin isolation and encapsulation in Na-alg carriers.

3.3.2. Adsorption Kinetics

Adsorption kinetics represents one of the key practical aspects in the application of environmentally friendly adsorptive materials. To demonstrate this effect, the influence of contact time on dye adsorption onto MiCS and MiCS@Alg sorbents was investigated at an initial dye concentration of 30 mg L⁻¹, using 0.001 g of adsorbent. The experiments were performed at three different temperatures (including 25 °C), under constant magnetic stirring. Adsorption equilibrium was monitored over 24 h, with maximum adsorption achieved after 90 min. The parameters obtained from nonlinear fitting are presented in the table for 25 °C, while the results for other temperatures are shown in Figure 7. To gain a deeper understanding of the adsorption kinetics, the experimental data were analyzed using pseudo-first-order (PFO) and pseudo-second-order (PSO) kinetic models [38]. The results are presented in

Table 2. Results of the nonlinear PFO and PSO models for pollutants adsorption (Ci = 30 mg L⁻¹, m_a = 1 mg, V = 8 mL, T = 25 °C, t = 90 min).

Adsorbent	Pollutant	Model Parameters	Pseudo-First	Pseudo-Second
MiCS	BR2	qe (mg g−1)	165.7 ± 6.00	175.3 ± 5.58
		k1 (min−1)/k2 (g mg−1 min−1)	0.7401 ± 0.196	0.0031 ± 0.001
		R2	0.956	0.979
	BY28	qe (mg g−1)	125.5 ± 1.33	127.9 ± 0.88
		k1 (min−1)/k2 (g mg−1 min−1)	36.38 ± 0.00	2.4812 ± 0.006
		R2	0.995	0.999
	NH3	qe (mg g−1)	54.90 ± 1.71	60.97 ± 0.855
		k1 (min−1)/k2 (g mg−1 min−1)	0.135 ± 0.02	0.0031 ± 0.0003
		R2	0.975	0.997
MiCS@Alg	BR2	qe (mg g−1)	117.3 ± 6.61	138.9 ± 8.6
		k1 (min−1)/k2 (g mg−1 min−1)	0.1565 ± 0.043	5.7928 ± 2.077
		R2	0.921	0.962
	BY28	qe (mg g−1)	78.45 ± 1.75	80.23 ± 1.18
		k1 (min−1)/k2 (g mg−1 min−1)	10.07 ± 8.38	0.0198 ± 0.0065
		R2	0.982	0.995
	NH3	qe (mg g−1)	31.70 ± 1.24	36.35 ± 0.984
		k1 (min−1)/k2 (g mg−1 min−1)	0.0906 ± 0.015	0.0032 ± 0.0004
		R2	0.969	0.993

. The kinetic results fitting, obtained by using the Weber–Morris (W–M) diffusion model [33], are presented in Table 2.

The adsorption kinetics of BR2, BY28, and NH₃ onto MiCS and MiCS@Alg adsorbents were evaluated using pseudo-first-order (PFO) and pseudo-second-order (PSO) kinetic models (Table 2). In all cases, the equilibrium adsorption capacities (qe) calculated from the nonlinear PSO model were closer to the experimental values than those obtained from the PFO model. Moreover, the correlation coefficients (R²) were consistently higher for the PSO model (0.962–0.999) compared to the PFO model (0.921–0.995), confirming its superior fitting. Similar results were reported for chestnut shell-derived activated carbon (CNS-AC) in the removal of quinary heavy metal ions [13]. For BR2 adsorption, MiCS exhibited a higher qe value (175.3 mg g⁻¹) compared to MiCS@Alg (138.9 mg g⁻¹). Similarly, for BY28, MiCS demonstrated a slightly higher adsorption capacity (127.9 mg g⁻¹) than MiCS@Alg (80.23 mg g⁻¹). A comparable trend was observed for NH₃, where MiCS reached qe = 60.97 mg g⁻¹, while MiCS@Alg achieved a lower value of 36.35 mg g⁻¹. Interestingly, although the incorporation of alginate slightly reduced the overall adsorption capacities of the sorbent, it led to an increase in the kinetic rate constants in some cases. For instance, in BR2 removal, MiCS@Alg displayed a significantly higher k₂ value compared to MiCS, indicating that surface modification improved the accessibility and reactivity of the binding sites. For BY28 and NH₃, the differences in k₂ values between the two sorbents were less pronounced, but MiCS still generally outperformed MiCS@Alg in terms of adsorption capacity. Kinetic analysis confirmed that the adsorption of dyes (BR2 and BY28) and ammonia (NH₃) onto both sorbents is more accurately described by the pseudo-second-order model. This highlights the predominance of chemisorption, involving electron sharing or exchange between the sorbent surface and pollutant molecules, as the dominant adsorption mechanism.

The results of the Weber–Morris model in

Table 3. Results of the intraparticle diffusion model (Weber–Morris) for BR2, BY28, and ammonia adsorption (Ci = 30 mg L⁻¹, m_a = 1 mg, V = 8 mL, T = 25 °C, t = 90 min).

Kinetic Model	Model Parameters	MiCS			MiCS@Alg
		BR2	BY28	NH3	BR2	BY28	NH3
Weber–Morris (Step 1)	kid1 (mg g−1 min−0.5)	6.385 ± 0.224	1.400 ± 0.050	1.078 ± 0.038	13.25 ± 0.517	0.935 ± 0.033	1.200 ± 0.042
	CBL (mg g−1)	120.2	118.3	83.41	20.64	71.10	43.78
	R2	0.992	0.997	0.879	0.909	0.865	0.982
Weber–Morris (Step 2)	kid2 (mg g−1 min−0.5)	4.015 ± 0.141	1.219 ± 0.043	0.6746 ± 0.024	12.25 ± 0.480	0.8248 ± 0.029	0.9287 ± 0.035
	CBL (mg g−1)	143.2	115.9	87.66	18.17	74.08	45.15
	R2	0.999	0.887	0.879	0.982	0.992	0.973

confirmed that the adsorption of BR2, BY28, and NH₃ on MiCS and MiCS@Alg proceeds in two distinct steps. In Step 1, the diffusion rate constants (k_id1) for BR2 were the highest among all systems, reaching 6.385 mg g⁻¹ min^−0.5 for MiCS and 13.25 mg g⁻¹ min^−0.5 for MiCS@Alg. These values indicate a rapid and intense initial uptake of BR2, consistent with a strong surface affinity. For BY28, significantly lower kid1 values were observed (1.400 for MiCS and 0.935 for MiCS@Alg), indicating slower diffusion in the early stage. Similarly, NH₃ showed modest initial diffusion rates, reflecting weaker interaction in the initial surface-controlled phase compared to BR2. In Step 2, the process slowed down, as demonstrated by reduced k_id2 values. For BR2, the diffusion rate constants decreased to 4.015 (MiCS) and 12.25 (MiCS@Alg), while for BY28, they were 1.219 (MiCS) and 0.8248 (MiCS@Alg). For NH₃, the kid2 values further decreased to 0.6746 (MiCS) and 0.9287 (MiCS@Alg), confirming that its adsorption is strongly controlled by intraparticle diffusion.

The intercept values (C_BL) were consistently high for dyes on MiCS, indicating that intraparticle diffusion is not the sole rate-limiting mechanism, but rather that film diffusion and surface interactions also play significant roles. For NH₃, the C_BL values were notably lower, suggesting that the adsorption of ammonia is less affected by surface diffusion resistance and relies more on gradual penetration into the pores. Interestingly, the incorporation of alginate significantly reduced the intercept values for dyes, which indicates that Na-alginate modification weakens the external film diffusion resistance and alters the adsorption pathway.

The correlation coefficients (R²) further supported these findings. For MiCS, the Weber–Morris model provided excellent fits, with R² values up to 0.999 (BR2, Step 2). For NH₃, the model also showed a good correlation (R² = 0.879–0.982), though slightly lower compared to dyes. In contrast, for MiCS@Alg, the model agreement was generally lower, with R² values ranging from 0.8649 to 0.9924, suggesting that alginate incorporation modifies the pore structure and sorption kinetics, thereby reducing the predictability of the intraparticle diffusion model. Overall, the Weber–Morris analysis demonstrates that BR2 exhibits the fastest sorption dynamics due to strong interactions with the sorbent surface, while BY28 shows slower kinetics but a significant contribution from intraparticle diffusion. In contrast, NH₃ adsorption is more strongly governed by diffusion within the sorbent pores. Alginate modification enhances the initial uptake rate of BR2 but reduces boundary layer effects, leading to altered adsorption kinetics compared to pristine MiCS.

The 25, 35, and 45 °C confirm that the PSO model provides a consistently better fit to the experimental data compared to the PFO model for all studied systems. This indicates that the adsorption rate is not solely governed by physical diffusion, but also involves chemisorptive interactions between the active sites of the sorbent and the adsorbate molecules. The adsorption process in all cases (Figure 7) exhibited a very fast initial phase, with most of the equilibrium uptake reached within the first 20–30 min, followed by a gradual plateau due to surface saturation. Temperature was found to influence adsorption differently depending on the pollutant and sorbent. For BR2 dye, increasing the temperature enhanced adsorption, confirming the endothermic nature of the process, whereas for BY28 dye and NH₃ pollutant, the adsorption capacity remained nearly unchanged with temperature. When comparing the two sorbents, MiCS@Alg generally showed a slower approach to equilibrium than MiCS, which can be attributed to the presence of alginate that modifies diffusion pathways and the accessibility of active sites. The compatibility between the experimental data and the kinetic models was evaluated based on the determination coefficient (R²) and the amount of adsorbed pollutants (q_e). Since the calculated q_e value for the pseudo-first-order model was significantly lower than the experimental value, it can be concluded that this model does not provide adequate results for either sorbent. The pseudo-first-order equation fails to accurately describe the experimental data for both sorbents in the removal of the investigated pollutants. Similar results to those presented in the table were reported for biochar-alginate in MB removal [6].

3.3.3. Thermodynamic Study and Adsorption Mechanism

Gibbs free energy (ΔG^Θ), enthalpy (ΔH^Θ), and entropy (ΔS^Θ) were calculated using Van’t Hoff equations to analyze the thermodynamic aspect of the adsorption process [38]. The calculated parameters are presented in

Table 4. Thermodynamic parameters for BR2, BY28 and ammonia removal using MiCS and MiCS@Alg (Ci =30 mg L⁻¹, m_a = 1–10 mg, V = 8 mL, T = 25 °C, 35 °C and 45 °C, t = 90 min.

Adsorbent	Pollutants	ΔGΘ (kJ mol−1)			ΔHΘ (kJ mol−1)	ΔSΘ (J mol−1 K−1)	R2
		25 °C	35 °C	45 °C
MiCS	BR2	−39.33	−40.78	−42.43	6.78	154.5	0.941
	BY28	−38.58	−40.01	−41.50	4.91	145.8	0.988
	NH3	−32.59	−33.74	−34.95	2.63	118.1	0.959
MiCS@Alg	BR2	−39.37	−40.82	−42.54	7.77	157.9	0.915
	BY28	−40.14	−41.52	−43.05	3.23	145.4	0.869
	NH3	−33.83	−35.05	−36.29	2.87	123.1	0.997

. A better agreement with the experimental values (R²) in the calculation of thermodynamic parameters was achieved by using the Langmuir constant obtained from the linear form of the Langmuir model. Larger deviations were obtained when applying the nonlinear form, as shown in Table 4.

The thermodynamic parameters obtained for the adsorption of BR2 and BY28 dyes and NH₃ on MiCS and MiCS@Alg provide clear insights into the nature of the process. The negative ΔG^Θ values, ranging from −32.59 to −43.05 kJ mol⁻¹, confirm that adsorption is spontaneous within the investigated temperature range, with spontaneity generally increasing as the temperature rises. The positive ΔH^Θ values (2.63–7.77 kJ mol⁻¹) indicate that the adsorption process is slightly endothermic, while the positive ΔS^Θ values (118.1–157.9 J mol⁻¹ K⁻¹) suggest an increase in randomness at the solid–solution interface, likely associated with the release of solvent molecules and structural rearrangements during adsorption. The most favorable adsorption was observed for BR2 on MiCS, with ΔG^Θ values between −39.33 and −42.43 kJ mol⁻¹ and ΔS^Θ = 154.5 J mol⁻¹ K⁻¹, reflecting the most potent driving force and a high entropic contribution. By contrast, the lowest spontaneity was recorded for NH₃ on MiCS, where ΔG^Θ values ranged from −32.59 to −34.95 kJ mol⁻¹ and ΔS^Θ was 118.1 J mol⁻¹ K⁻¹. Interestingly, although adsorption of BY28 on MiCS also showed relatively high spontaneity (ΔG^Θ = −38.58 to −41.50 kJ mol⁻¹; ΔS^Θ = 145.8 J mol⁻¹ K⁻¹), this system exhibited the best agreement with the Van’t Hoff model (R² = 0.9877), indicating strong consistency with the linear thermodynamic approach. Similar values of the thermodynamic parameters ΔH^Θ and ΔS^Θ were obtained for the removal of TC, as presented in the study [39].

For the alginate-modified material, MiCS@Alg, BR2 adsorption remained highly favorable, with ΔG^Θ values between −39.37 and −42.54 kJ mol⁻¹, ΔH^Θ = 7.77 kJ mol⁻¹, and ΔS^Θ = 157.9 J mol⁻¹ K⁻¹. This system showed a slightly higher enthalpic contribution compared to pristine MiCS (ΔH^Θ = 6.78 kJ mol⁻¹), while maintaining a strong entropic effect. Adsorption of NH₃ on MiCS@Alg was characterized by ΔG^Θ values of −33.83 to −36.29 kJ mol⁻¹, ΔH° = 2.87 kJ mol⁻¹, and ΔS^Θ = 123.1 J mol⁻¹ K⁻¹, showing improved spontaneity relative to NH₃ on MiCS, with excellent model agreement (R² = 0.997). On the other hand, BY28 on MiCS@Alg demonstrated the lowest fit to the Van’t Hoff model (R² = 0.869), with ΔG^Θ values ranging from −40.14 to −43.05 kJ mol⁻¹, ΔH^Θ = 3.23 kJ mol⁻¹, and ΔS^Θ = 145.4 J mol⁻¹ K⁻¹, suggesting that alginate modification alters the thermodynamic profile by slightly reducing the entropic contribution compared to MiCS.

Overall, the results from Table 4 indicate that alginate modification modulates the balance between enthalpic and entropic contributions, strengthening spontaneity for NH3 adsorption while slightly decreasing the model’s applicability for BY28.

Adsorption and thermodynamic results indicate that the ion-exchange is an operative mechanism with participation of electrostatic and π–π stacking interactions. Additional proof of the adsorption mechanism was accessed through supplementary experiments. In the course of the adsorption experiments, using BR2 and BY28 dyes, the initial and final pH, as well as the time-dependent changes, were followed. It was found that pHi was decreased by 0.58 and 0.74 units for MiCS, respectively. Even lower change was recorded for MiCS@Alg adsorbent. Thus, the most plausible mechanism can be described by the sodium exchange, due to pH adjustment with NaOH, with cationic dyes and nucleophilic ammonia (Figure S3). The selection of operative pH (Section 2.5. Adsorption experiments) was based on the availability of the most natural and polluted water and deprotonation of MiCS surface functionalities (mainly carboxylic group: pKa of carboxylic acid varies in the range 3.6–4.5). In that manner, effective deprotonation of carboxylic acid groups creates an anionic structure able to participate in an ion-exchange process with cation dyes. Otherwise, the use of lower pH in the course of ammonia adsorption is in favor of a simple neutralization reaction producing ammonium salt. Also, the study of ionic strength (IS) (0.01, 0.05, and 0.1 mol dm⁻³ KNO₃) showed a low effect on the change in adsorption efficiency (less than 10% adsorbent capacity decrease). It primarily affects adsorbent/adsorbate electrostatic interactions at the adsorbent surface, ion diffusivity, and hydration shell stability.

These phenomena were reflected through adsorption efficiency due to the same/opposite charge sign repulsion/attraction, causing higher deterioration in the adsorption efficiency of weakly bonded pollutants. Thus, MiCS could be effectively used at increased concentrations of anionic competing ions if we consider selective removal, considering organic pollutant sodium benzoate or Cl⁻ and SO₄² (less than 5% of adsorption capacities decrease, given in Table 1, was found). Otherwise, use of cationic pollutants, e.g., calcium or magnesia, and especially Al³⁺, causes a significant decrease in dyes adsorption capacity. A large decrease in adsorption efficiency (higher than 70%) indicates the predominance of an electrostatic interaction between carboxylate anion and cation (e.g., Ca²⁺ and Mg²⁺), i.e., higher affinity to ion-exchange with calcium cation, forming preferably calcium–carboxylate salt. Ca²⁺ and Mg²⁺ ions show a high affinity with respect to carboxylate ions, interacting through multiple binding modes by forming monodentate and bidentate metal complexes. The extent and type of metal/carboxylate ion binding are influenced by the number and orientation of neighboring carboxylate groups. The strength of these bindings can be affected by the presence of other carboxylic groups in close proximity, which help in the formation of bidentate complexes (cooperative mechanism). At a higher presence of negatively charged functionalities with diverse structural arrangements, they can participate in forming different binding structures on the adsorbent surface. Strong affinity of hard metal cation Al³⁺ to carboxylate ion, i.e., oxygen as electron donor, provides effective formation of stable complexes, even higher stability than those obtained with Ca²⁺ and Mg²⁺ cations (adsorption efficiency drop for more than 87% with respect to BY28). In general, MiCS and MiCS@Alg can effectively remove cationic species in the presence of anionic species, as they have a higher affinity for metal cations compared to textile dyes.

These results indicate the potential of the produced MiCS adsorbent to be applicable in the process of natural water softening. High influences on dye removal arise from their structural complexity, hydration, and steric hindrance to the attainment of multiple adsorbent/adsorbate interactions. The moderate effect of methylene blue on BY28 dye removal indicates that similar competitive potential (somewhat higher for methylene blue, 62% adsorption capacity decrease was found for BY28). It indicates the significance of structural limitation and charge concentration/availability for interaction with the carboxylate ion. On the contrary, the low competitive effect of the cationic pollutants (Ca²⁺, Mg²⁺, and Al³⁺) with ammonia removal at pH 4 was observed. In this case, a neutralization reaction takes place freely with no competition from the side processes.

3.4. Desorption Study

Desorption studies play a key role in understanding the nature of the adsorption process and evaluating the regeneration of adsorbents [40] and were therefore conducted to investigate their reusability. The sorbent recycling study was performed over five cycles using 0.1 M HCl as the desorbing agent. The obtained results are presented in Figure 8 and summarized in Table S4.

By comparing the performance of the MiCS and MiCS@Alg sorbents, it is evident that MiCS exhibits higher adsorption capacities and slightly better regenerability than the alginate-modified sorbent, which is consistent with the trend obtained from the Langmuir adsorption model. The maximum adsorption capacity for BR2 on MiCS reached 291.1 mg g⁻¹, whereas for MiCS@Alg it was significantly lower (180.5 mg g⁻¹). A similar trend was observed for BY28 and NH₃, where MiCS demonstrated higher adsorption capacities across all five cycles. The desorption efficiency of both materials was high in the first cycle (>90%) (Figure 8). However, MiCS@Alg showed a more pronounced decline in subsequent cycles, particularly for NH₃ (92.2% → 70.8% for MiCS vs. 88.2% → 68.8% for MiCS@Alg), corresponding to an overall capacity loss of 30–40% after five cycles, which indicates limited long-term regenerability, especially for MiCS@Alg. These results suggest that, although alginate contributes to forming a more mechanically stable composite, it may restrict the accessibility of active sites and reduce the overall efficiency of the adsorption–desorption process. MiCS, in contrast, retained a more stable capacity in the initial cycles, highlighting the advantage of the unmodified sorbent in maintaining accessible active sites. Comparable desorption values for BY28 dye from a natural-origin sorbent were reported in the study by [23], where desorption efficiencies of about 98–99% were achieved after the first and second cycles.

3.5. Bed Column Study

To evaluate the adsorption performance of MiCS@Alg in a flow system, it is essential to conduct adsorption experiments beyond batch tests. The maximum adsorption capacity (q_exp, mg g⁻¹) for each pollutant was calculated based on the number of bed volumes processed until the concentration in the effluent exceeded the maximum permissible concentration (MPC), which is controlled by COD determination (Section 2.4. Characterization and analysis). Data analysis was simplified by applying the Bohart–Adams and Yoon–Nelson models [33], which assume that adsorption kinetics control the rate-limiting step and are valid only for single-component systems. The test solution was prepared by passing either BR2 and BY28 dyes, or ammonia at a concentration of 5 mg L⁻¹ to model water, adjusted to pH 7.5 for the dyes and pH 4 for ammonia. This solution was then passed through the column at controlled flow rates of 0.5, 1.0, and 1.5 mL min⁻¹. Effluent samples were collected at set intervals to measure the concentration of dyes and NH₃. The obtained results are presented in

Table 5. B-A and Y–N fitting for BR2, BY28, and NH₃ adsorption onto MiCS@Alg (Ci_[BR2] = Ci_[BY28] = Ci_[NH3] = 5.0 mg dm⁻³, m_ads = 570 mg, T = 25 °C).

Model and Parameters		Pollutant	Q (cm3 min−1)
			0.5	1.0	1.5
B-A	KBA (dm3 mg−1 min−1)	BR2	0.0360 ± 0.0002	0.0703 ± 0.0006	0.109 ± 0.0012
	qo (mg g−1)		147.6 ± 0.19	122.8 ± 0.25	100.1 ± 0.30
	R2		0.999	0.999	0.993
	KBA (dm3 mg−1 min−1)	BY28	0.0409 ± 0.0005	0.0949 ± 0.0023	0.138 ± 0.0017
	qo (mg g−1)		94.71 ± 0.32	77.02 ± 0.44	61.58 ± 0.26
	R2		0.998	0.998	0.998
	KBA (dm3 mg−1 min−1)	NH3	0.0665 ± 0.0008	0.130 ± 0.0024	0.216 ± 0.0047
	qo (mg g−1)		47.28 ± 0.18	43.19 ± 0.28	37.51 ± 0.29
	R2		0.999	0.999	0.998
Y-N	kYN (min−1)	BR2	0.360 ± 0.0025	0.351 ± 0.0032	0.344 ± 0.0040
	θ (min)		16.83 ± 0.022	13.99 ± 0.050	11.42 ± 0.034
	R2		0.999	0.999	0.999
	kYN (min−1)	BY28	0.409 ± 0.0054	0.475 ± 0.0099	0.480 ± 0.0001
	θ (min)		10.80 ± 0.036	8.78 ± 0.069	7.02 ± 0.0003
	R2		0.999	0.999	0.999
	kYN (min−1)	NH3	0.665 ± 0.00001	0.668 ± 0.012	0.719 ± 0.0056
	θ (min)		5.39 ± 0.00001	4.92 ± 0.031	4.27 ± 0.032
	R2		0.999	0.998	0.998

and Figure S4.

The results obtained from the fixed-bed column tests (Table 5) indicate that MiCS@Alg performs effectively across all the pollutants studied. In similar tests carried out with a flow rate of 3 mL min⁻¹ and an initial contaminant concentration of 5 mg L⁻¹ (yielding an EBCT of 1.66 min), the observed adsorption capacities were 88.3 mg g⁻¹ for BR2, 49.4 mg g⁻¹ for BY28, and 33.2 mg g⁻¹ for ammonia. These findings demonstrate that shorter residence times within the column result in lower adsorption capacities for each of the tested substances. The concentration of pollutants in the influent also plays a critical role in adsorption efficiency; lower feed concentrations were associated with increased adsorption capacities, improving by roughly 7–16%. This behavior underscores the potential of MiCS@Alg in treating drinking water, particularly at lower contaminant levels where the system operates on the initial, more efficient portion of the adsorption isotherm. The material’s high porosity and favorable surface properties contribute to reduced internal mass transfer resistance, making adsorption sites more readily accessible.

3.6. Comparative Overview of Adsorption Data

Based on the data in Table S5 and previously reported studies, the adsorbents investigated in this work can be classified among materials with relatively high adsorption capacities. Notably, isolated melanin and sodium–alginate beads containing melanin (MiCS and MiCS@Alg) have not, to the best of our knowledge, been reported in the available literature for the removal of such cationic pollutants. These materials are particularly attractive because of their low cost, ready availability, and inherent biodegradability. MiCS exhibits a higher adsorption capacity than conventional activated carbon derived from horse chestnut (≈169 mg g⁻¹; [41]) and citric-acid-modified corn stalk (~200 mg g⁻¹; [42]) but lower than the highest values reported for ZnCl₂-activated chestnut shell [43] (≈1.1–1.4 g g⁻¹) and for the chestnut–snail shell composite MCS3-1 (up to 1.6 g g⁻¹; [44]). For ammonia (NH₃), MiCS shows a markedly lower capacity (≈73 mg g⁻¹), as expected from the different binding mechanism compared with cationic dyes. The encapsulation of melanin in sodium–alginate beads further decreases the adsorption capacity (e.g., from 300 to 189 mg g⁻¹ for BR2) but produces a striking enhancement in adsorption kinetics. The pseudo-second-order rate constant (k₂) for BR2 increases from 0.0031 g mg⁻¹ min⁻¹ for MiCS to as high as 5.79 g mg⁻¹ min⁻¹ for MiCS@Alg-several orders of magnitude faster than most reported sorbents, including citric-acid-modified corn stalk (maximum 8.7 g mg⁻¹ min⁻¹ at elevated temperatures). Typical k₂ values for comparable sorbents fall in the 10⁻⁴–10⁻³ g mg⁻¹ min⁻¹ range (e.g., chestnut–snail composite 1.25 × 10⁻⁴ [40]; almond–walnut 9.16 × 10⁻⁴, [45]; Elaeagnus 1.26 × 10⁻⁴, [46]). Within Table S5, the highest adsorption capacity is again observed for the chestnut–snail composite (MCS3-1), reaching ~1.6 g g⁻¹ for cationic dye removal, followed by ZnCl₂-activated chestnut (1.1–1.4 g g⁻¹). The lowest value is recorded for MiCS toward ammonia, at ~73 mg g⁻¹. This range highlights both the exceptional performance of the chestnut snail composite and the expected limitation of MiCS when targeting non-aromatic, anionic pollutants, such as ammonia. Regarding selectivity, MiCS exhibits excellent affinity for cationic dyes but lower affinity for ammonia, indicating that adsorption is primarily governed by electrostatic interactions and π–π stacking between the aromatic dye rings and the melanin structure. A slight increase in qmax with temperature (25–45 °C) follows the trend reported for ZnCl₂-activated chestnut and other carbonized materials.

Overall, the results summarized in Table S5 demonstrate that MiCS@Alg, particularly in the form of alginate beads, offers a competitive, rapid, and environmentally friendly approach for treating dye-contaminated and ammonium-rich wastewaters, even though its maximum capacity is lower than commercial activated carbons having high surface areas.

4. Conclusions

In this study, produced MiCS and MiCS@Alg, derived from chestnut shells, were successfully used as efficient adsorbents for the cationic dyes BR2 and BY28 and NH₃. Utilizing waste chestnut shells treated with environmentally friendly extraction and modification procedures, the melanin-based materials exhibited functional groups capable of removing cationic pollutants from aqueous solutions. Characterization results revealed significant changes in the structure and morphology of MiCS during the purification and encapsulation process in MiCS@Alg. The isolated melanin particles demonstrated improved adsorption efficiency compared to raw chestnut shell particles, which can be attributed to their smaller size and higher negative zeta potential (−36.6 ± 2.9 mV at pH 10.5). DLS measurements showed that the average particle size of iCS was 786 nm at pH 5.8, whereas MiCS measured 612 nm at pH 10.5. The process was found to be low endothermic and spontaneous, with maximum adsorption capacities of 300.2 mg g⁻¹ (BR2), 201.5 mg g⁻¹ (BY28) and 73.08 mg g⁻¹ (NH₃) on MiCS, and 189.3 mg g⁻¹ (BR2), 117.1 mg g⁻¹ (BY28) and 50.06 mg g⁻¹ (NH₃) on MiCS@Alg at 45 °C, as determined using the Langmuir model. Thermodynamic, kinetic, and mechanistic studies indicated a complex adsorption process, primarily governed by electrostatic interactions between the negatively charged functional groups of melanin and the cationic dyes. Positive enthalpy values (ΔH^Θ = 2.50–7.77 kJ mol⁻¹) suggest that adsorption is mildly endothermic, while negative Gibbs free energy values (ΔG^Θ from −33 to −43 kJ mol⁻¹) confirm that the process is spontaneous and feasible, with spontaneity generally increasing with temperature. Additionally, positive entropy values (ΔS^Θ = 119.3–157.9 J mol⁻¹ K⁻¹) indicate an increase in randomness at the solid–solution interface, likely due to structural rearrangements and solvent molecule release during adsorption. The results obtained from the fixed-bed column tests indicate that MiCS@Alg performs effectively across all the pollutants studied. The materials also demonstrated good reusability over multiple adsorption–desorption cycles, maintaining significant adsorption efficiency, highlighting their potential for sustainable water treatment applications.