Synergistic engineering of bio-inspired magnetic alginate-chitosan interpenetrating double-network beads for rapid simultaneous capture of arsenic and phosphate from complex water systems

Abstract

The increasing pollution of water aquatic systems by the oxyanionic metalloids and nutrients requires the design of multipurpose, readily arrestable adsorbents. The research problem addressed in this study was to study the synergistic production of magnetically responsive alginate-chitosan-Fe₃O₄ (ACF) nanocomposites that can operate in the simultaneous removal of pentavalent arsenic [As(v)] and orthophosphate [PO₄³⁻] in complex aqueous matrices. ACF nanocomposite was prepared using an in situ mineralized co-precipitation of ferric and ferrous salts in a biopolymeric template, which was then, cross-linked droplet wise with calcium chloride and glutaraldehyde ions. FTIR, SEM, TEM, EDX, BET, TGA, DSC, and XPS were important in characterizing the structural of prepared nanocomposites. TEM analysis showed spherical nanoparticles with a diameter of 45 nm on average and anchored in a porous carbohydrate matrix. Vibrating Sample Magnetometry showed that the saturation magnetization was 26.1 emu g⁻¹ with negligible coercivity and remanence, confirming superparamagnetic behavior at room temperature. This high magnetization enabled rapid magnetic separation of the ACF nanocomposite from aqueous suspension within 30 seconds using an external neodymium magnet (0.5 T), as determined by time-resolved turbidity and ICP-OES measurements. The experiment of batch adsorption established the maximum Langmuir adsorption capacity of As(v) and orthophosphate 114.5 mg g⁻¹ and 138.2 mg g⁻¹ at pH 6.5, respectively with initial pollutant ions concentration 50 mg L⁻¹, ACF 1.0 g L⁻¹, after 240 min. Kinetic modeling was done along pseudo-second-order mechanism implying chemisorption through inner-sphere complexation. The shifting in FTIR bands and appears new Fe–O–As/P bonds confirm key adsorption mechanisms. The nano-architectures were highly reusable with >90% of their original capacity remaining after five cycles and >85% after six cycles of regeneration. This study offers a solid guideline in the creation of bionanocomposites that may counter multi-pollutants in the design of a decentralized water treatment plant.

The increasing pollution of water aquatic systems by the oxyanionic metalloids and nutrients requires the design of multipurpose, readily arrestable adsorbents.

1. Introduction

Extensive pollution of world ground water systems by dioxin oxyanions, namely pentavalent arsenic As(v) and orthophosphate (PO₄³⁻), is an issue of great concern to ecological integrity and human health.^1,2 Exposure to arsenic continues to be a critical humanitarian catastrophe because it is estimated that there are now more than 200 million individuals in 70 countries consuming water containing in excess of the World Health Organization limit of 10 micrograms per liter of arsenic.³ At the same time, the high release of phosphates by agricultural runoff and industrial effluents has enhanced the rapid eutrophication of freshwater bodies, where the level of phosphorus is usually well above the critical value of 0.02 mg per liter to initiate harmful algal blooms.⁴ The fact that these species co-exist is of special concern because these species share structural similarities; they both appear as tetrahedral oxyanions, which compete against each other to occupy the same surface sites on conventional adsorbents, and as a result, dramatically decreases the effectiveness of traditional remediation technologies.^5,6 Conventional methods of the elimination of such contaminants are chemical precipitation,⁷ reverse osmosis,^8,9 and ion exchange.^10,11 Nevertheless, these approaches often have the shortcomings of being prohibitively expensive to operate; they produce secondary chemical sludge, and are not selective in a complex aqueous matrix.¹² Adsorption is one of the favored alternatives because it is quite economical and it is relatively simple to operate.^13,14 Nonetheless, most of the adsorbents available, including activated carbon and basic metal oxides, do not have high adsorption capacity of As(v) and orthophosphate, or have serious difficulties with regard to phase separation and post-treatment recovery.¹⁵ The challenge in the extraction of used micro-scale powders of the treated water tends to bring about secondary particulate pollution thus limiting their usefulness in the large-scale flow system.¹⁶ The importance of this technological bottleneck is the necessity of developing materials that are multifunctional and have to be both highly reactive on the surface and with intrinsic magnetic properties to quickly separate solids and liquids.^17,18 In current studies, biopolymer-based composites have been studied, but the most significant void is in the development of synchronized nano-architectures capable of co-controlling competitive interference of arsenic and phosphate. Alginate and chitosan have been investigated separately^19,20 but their combination into a single, magnetically responsive system, using in situ techniques has not been well studied. Despite the large number of literature reports on the use of sodium alginate/chitosan/Fe₃O₄ composites for the adsorption of arsenic and phosphate, several critical gaps remain:

(1) Lack of simultaneous capture studies: most studies focus on the removal of a single pollutant either arsenic or phosphate rather than their simultaneous capture from complex water systems. Given that these oxyanions co-exist in real contaminated water and compete for the same adsorption sites, understanding their competitive adsorption behavior is essential for practical applications.

(2) Lack of interpenetrating double-network (IPN) architectures: most alginate-chitosan composites are prepared by simple physical blending or single cross-linking (either ionic or covalent), resulting in materials with limited mechanical and chemical stability. The formation of a true IPN structure, where two independently cross-linked networks are physically interpenetrating, has not been systematically investigated for simultaneous arsenic and phosphate removal.

(3) Lack of in situ synthesis protocols: most magnetic biopolymer composites are prepared by ex situ mixing of pre-synthesized Fe₃O₄ nanoparticles with the polymer matrix, which often results in non-uniform distribution, leaching of nanoparticles, and reduced magnetic properties. The use of in situ co-precipitation to form Fe₃O₄ nanoparticles within the biopolymer matrix has not been well studied.

(4) Lack of mechanistic understanding: most studies focus on performance metrics (adsorption capacity, kinetics, isotherms) without detailed mechanistic understanding of competitive adsorption, inner-sphere complexation, and the role of different functional groups.

(5) Lack of rapid magnetic separation: most magnetic biopolymer composites require 60–300 seconds for magnetic separation due to lower saturation magnetization or thicker polymer coatings. Achieving rapid separation (≤30 seconds) requires optimization of both the magnetic core and the polymer coating.

These limitations are covered in this study by providing the synergistic fabrication of the alginate-chitosan-Fe₃O₄ (ACF) nano-architectures with the following innovative aspects:

(1) Simultaneous capture of As(v) and phosphate: the ACF nanocomposite achieves high adsorption capacities for both pollutants simultaneously (114.5 mg g⁻¹ for As(v) and 138.2 mg g⁻¹ for phosphate) and maintains >85% of its original capacity even in the presence of competing ions.

(2) Interpenetrating double-network (IPN) architecture: the ACF nanocomposite is prepared by sequential dual cross-linking—ionic cross-linking of alginate via Ca²⁺ bridges and covalent cross-linking of chitosan via glutaraldehyde—resulting in a true IPN structure with enhanced mechanical strength (2.4 MPa), thermal stability (T_g = 142–149 °C), and chemical stability.

(3) In situ co-precipitation of Fe₃O₄: the Fe₃O₄ nanoparticles are formed in situ within the biopolymer matrix, resulting in uniform distribution, strong chemical bonding (carboxylate-iron bridges), and high saturation magnetization (26.1 emu g⁻¹).

(4) Comprehensive mechanistic understanding: the adsorption mechanism is elucidated through kinetic modeling (PFO, PSO, IPD), isotherm modeling (Langmuir, Freundlich, Temkin), thermodynamic analysis, competitive ion studies, and FTIR/XPS analysis after adsorption. A mechanistic model is proposed involving electrostatic attraction, inner-sphere complexation, and hydrogen bonding.

(5) Rapid magnetic separation (30 seconds): the ACF nanocomposite achieves magnetic separation within 30 seconds using a 0.5 T neodymium magnet, attributed to the high saturation magnetization and thin biopolymer coating.

(6) High reusability and stability: the ACF nanocomposite maintains >90% of its original capacity after 5 cycles and >85% after 6 cycles, attributed to the IPN structure and chemical bonding between Fe₃O₄ and the polymer matrix.

Using a dual-templating format of in situ co-precipitation of magnetite and ionic cross-linking, we have designed a material, which uses amine groups of chitosan and carboxyl groups of alginate to produce a high density of positive surface charges at the micro-scale. The architecture allows the sequestration of As(v) and PO₄³⁻ to take place concurrently by an outer-sphere complexation mechanism as well as an electrostatic interaction. The main aim of this study is to measure the multi-interface sorption dynamics and thermodynamic characteristics of such hybrid structures in complicated aqueous environments. Through investigation of competitive ion displacement mechanisms, our interest is to clarify the structural organization of the ACF composite which attenuates the inhibitory properties of other competitive ions such as sulfates and nitrates. Moreover, the paramagnetic nature of the Fe₃O₄ phase facilitates rapid magnetic recovery, with the ACF nanocomposite exhibiting a saturation magnetization of approximately 26.1 emu g⁻¹ and achieving more than 99% recovery within 30 s under an external magnetic field.

2. Materials and experimental

2.1. Materials and reagents

The manufacture of the magnetically responsive nano-architectures involves the accurate choice of reagents of analytical grade so that reproducibility and high sorption rates are guaranteed. Low molecular weight chitosan (C₆H₁₁NO₄)_n with a deacetylation level of 75 to 85 percent, sourced by Sigma-Aldrich, and Sodium Alginate (NaC₆H₇O₆)_n with a high guluronic acid concentration are the two main biopolymeric scaffolds used to construct the primary component of the bio-nano-scaffold. The magnetic iron oxide core is produced with precursors of iron(iii) chloride hexahydrate (FeCl₃·6H₂O, 99.8% purity) and iron(ii) chloride tetrahydrate (FeCl₂·4H₂O, 99.0% purity) both of which are provided by Merck. In the case of in situ-co precipitation of these magnetite nanoparticles, ammonium hydroxide (NH₄OH, 28 to 30% NH₃ basis) of Fisher Scientific is the precipitating agent that is used in the process. The hybrid matrix is ionically cross-linked and strengthened by adding calcium chloride (CaCl₂, 97.0% anhydrous) which is bought at Alfa Aesar.

The aim of the sequestering studies is to use high-purity salts in order to achieve good stoichiometry of the targeted oxyanion pollutants. Sodium Arsenate Dibasic Heptahydrate (Na₂HAsO₄·7H₂O, 98.5% analytical grade) of Sigma-Aldrich is used as pentavalent arsenic in the stock solution and potassium dihydrogen phosphate (KH₂PO₄, 99.5% purity) of Merck is used as the orthophosphate species in the stock solution. Several background electrolytes and co-existing ions are used to study competitive ion displacement and influence of complex aqueous matrices, and they include sodium nitrate (NaNO₃, 99.9%), sodium sulfate (Na₂SO₄, 99.0%), and sodium bicarbonate (NaHCO₃, 99.7%), which are sold by Fisher Scientific. The hydrochloric acid (HCl, 37% fuming) and sodium hydroxide (NaOH, 98.0% pellets) of Alfa Aesar are precisely used in the adjustment of the solution pH. Glutaraldehyde (C₅H₈O₂, 25% aqueous solution) of Sigma-Aldrich is added as a optional chemical cross-linker to stabilize the structure of the chitosan chains. All solutions, reagent dilutions, and glassware washing are done in deionized water with a resistivity of 18.2 megohm-centimeters at 25 °C which has been purified by a Millipore Milli-Q purification system.

2.2. Synthesis of magnetic alginate-chitosan-Fe₃O₄

The magnetic alginate-chitosan-Fe₃O₄ (ACF) nano-architectures were prepared through a multi-step integrated protocol that involved in situ co-precipitation of iron oxide nanoparticles followed by sequential ionic and covalent cross-linking to form an interpenetrating double-network (IPN) structure. To begin with, 1.50 g of low molecular weight chitosan (85% deacetylation) was dissolved in 100 mL of 1.0% (v/v) aqueous acetic acid. The solution was stirred at 500 rpm for 4 hours until a clear and viscous solution was attained. At the same time, 2.50 g of sodium alginate was dissolved in 150 mL of deionized water at 60 °C with vigorous stirring for 2 hours. After homogenization of both biopolymer solutions, blending took place at a 1 : 1.5 volume ratio and left at room temperature to allow physical entanglement of the polymer chains.²¹

To form the magnetic nuclei in situ, 4.25 g of FeCl₃·6H₂O and 1.56 g of FeCl₂·4H₂O (representing a 2 : 1 molar ratio of Fe³⁺/Fe²⁺) were dissolved in 50 mL of deoxygenated deionized water. The iron salt solution was transferred into the biopolymer blend under a steady flow of nitrogen to avoid early oxidation. The product was mixed at 800 rpm for 45 minutes at 25 °C. The pH of the system was then raised systematically to 10.5 by adding 2.0 M NaOH dropwise at a controlled rate of 1.5 mL min⁻¹. The incumbent black suspension was then allowed to age in a 70 °C bath at a steady flow rate of 0.5 mL min⁻¹ with the aid of a syringe pump.²² The medium used to cross-link the samples was a solution of 3.0% (w/v) CaCl₂ and 0.5% (v/v) glutaraldehyde in 250 mL and brought to pH 5.0. The sequential cross-linking—ionic cross-linking of alginate via Ca²⁺ bridges and covalent cross-linking of chitosan via glutaraldehyde—in the presence of the other polymer network resulted in the formation of an interpenetrating double-network structure. The microspheres formed were left to harden in this solution over 12 hours in orbital shaking at 120 rpm to allow the formation of the interpenetrating polymer network with the use of calcium-alginate ionic bridges and chitosan-glutaraldehyde covalent bonds.

The nano-architectures that had been synthesized were then collected using an external neodymium magnet (0.5 T) and washed with anhydrous ethanol and deionized water several times until the supernatant had a pH of neutrality and a conductivity of less than 10 microsiemens per centimeter. The final composite was dried in a vacuum at 45 °C for 24 hours to obtain a constant mass. The resulting xerogels were kept in a desiccator with silica gel before the experiments of characterizing and sequestration. The final architecture was measured at 18.4–18.6 wt% Fe by inductively coupled plasma optical emission spectroscopy.

2.3. Characterization apparatus

Surface morphology and elemental maps were analyzed by the use of the JEOL JSM-7610FPlus system with the help of Field Emission Scanning Electron Microscopy (FESEM) attached to the Energy Dispersive X-ray Spectroscopy (EDS). The applied increasing voltage was maintained at 15 kV under high vacuum conditions and the samples were sputter-coated with a thin layer of gold lest the samples were charged. The internal morphology and distribution of Fe₃O₄ nanoparticles in the biopolymer matrix were observed using Transmission Electron Microscopy (TEM) in the FEI Tecnai G2 F20 instrument that works on the basis of accelerating voltage of 200 kV. Fourier Transform Infrared Spectroscopy (FTIR) was used in probing of functional groups identification and chemical bonding environments on a Shimadzu IRAffinity-1S spectrophotometer. The data were taken as spectra in the wavenumber of 4000 to 400 cm⁻¹ with a resolution of 4 cm⁻¹ in 64 cumulative scans using the KBr pellet method. The pore size distribution, pore volume, and specific surface area were obtained by nitrogen phreatic adsorption–desorption isotherms carried out at 77 K in a Micromeritics ASAP 2020 Brunauer–Emmett–Teller (BET) analyzer. Prior to analysis, samples were heated at 100 degrees Celsius in a period of 12 hours, to remove gases. Decomposition profiles and thermal stability of the nano-composites were measured by Thermogravimetric Analysis and Differential Scanning Calorimetry (TGA/DSC) on a TA Instruments SDT Q600. The analysis was done between 30 and 800 °C at a linear heating rate of 10 °C min⁻¹ at a constant flow of nitrogen being 100 mL min⁻¹.

Swelling and solubility tests were performed to evaluate the structural integrity and cross-linked nature of the alginate-chitosan network. Water swelling was determined after equilibrium hydration, while selective solubility tests were conducted in acetic acid and EDTA solutions. The mechanical integrity of the prepared beads was evaluated by compression testing, and the compressive strength and elastic modulus were determined. Solid-state ¹³C CP-MAS NMR spectroscopy was additionally employed to identify the characteristic carbon environments of alginate and chitosan and to assess molecular–level interactions within the composite network.

X-ray Photoelectron Spectroscopy (XPS) with thermo scientific K-Alpha system was used to investigate the surface elemental composition and oxidation status of arsenic and phosphorus after sorption. The source was monochromatic Al Ka radiation of 1486.6 eV pass energy of 50 eV at a spot size of 400 micrometers high-resolution scans. The characteristics of surface charges and the zero charge point (pH_pzc) were obtained through the measurements of zeta potential using a Malvern Zetasizer Nano ZS90. These were measured in a pH of 2.0–11.0 at a concentration of 0.01 M NaCl as background-ion. Magnetic measurements (including saturation magnetization and coercivity) were measured at room temperature with a Lake Shore 7404 Vibrating Sample Magnetometer (VSM) with a magnetic field of between −15 000 and +15 000 Oe. Lastly, the pentavalent arsenic and orthophosphate concentrations were determined by the Inductively Coupled Plasma Optical Emission Spectrophotometry (ICP-OES) on a PerkinElmer Optima 8000 with auxiliary colorimetric determination of the phosphate and arsenic species at 880 nm and 193.7 nm, respectively.

2.4. Application studies

As(v) and orthophosphate (PO₄³⁻) stock solutions were developed by dissolvingNa₂HAsO₄·7H₂O and KH₂PO₄ in deionized water to obtain concentrations of 1000 mg L⁻¹. These stock solutions were then diluted to comprise of working concentration of 5–200 mg L⁻¹ of both targets. The effect of adsorbent dosage was examined by adjusting the dosage of the magnetically responsive alginate-chitosan Fe₃O₄ nano-architecture between 0.2 and 2.5 g L⁻¹ in the 50 mL of the aliquots of the contaminated water. A 0.1 M NaOH, 0.1 M HCl solution was measured and adjusted to pH values of 2.0 to 11.0 by adding 0.1 M NaOH and 0.1 M HCl, respectively, until the pH reached 11.0. pH was adjusted and the point of zero charge and the best sorption conditions were determined.

The kinetics of batch sorption was determined by sampling the mixture at specified time intervals of between 5 to 1440 minutes. The experiments were started at the concentration of 50 mg L⁻¹ of both ions and the agitation rate was set at 200 rpm. Pseudo-first-order, pseudo-second-order, and intraparticle diffusion models were used in order to characterize the resultant kinetic data. In isothermal studies, the adsorbate concentration was varied between 10 and 500 mg L⁻¹ at pH 6.5, selected as the optimum pH for simultaneous As(v) and orthophosphate removal. The equilibrations of the suspensions took 24 h so that a thermodynamic equilibrium would be achieved.

Effects of temperature on the sequestering process were measured at 288 K, 298 K, and 318 K to help in the determination of thermodynamic parameter, such as Gibbs free energy (ΔG°), enthalpy (ΔH°), and entropy (ΔS°). Competitive ion replacement experiments were carried out in complex aqueous systems with common concurring anions of sulfate, nitrate and chloride that were present in the concentration of 50 to 500 mg L⁻¹. After the sorbing process, an external Neodymium magnet with the field strength of 0.5 T was used to collect the paramagnetic nano-architectures at the aqueous phase. The used adsorbents were subjected to regeneration by means of 0.5 M NaOH and further into a 0.1 M NaCl wash and the paramagnetic reusability was tested in six successive adsorption–desorption cycles. The experiments were conducted in triplicates to guarantee the statistical significance and the end concentrations of As(v) and orthophosphate were measured with the inductively coupled plasma optical emission spectrometry and molybdenum blue spectrophotometric techniques respectively.

2.5. Regeneration and reusability study

There was systematic evaluation of regeneration and reusability of the magnetically responsive alginate-chitosan-Fe₃O₄ (ACF) nanocomposite to ascertain the possibility of its applicability as a plausible alternative and cost effective one in case of repeated treatment of wastewater. After each adsorption experiment, the used adsorbent was washed into the aqueous medium using an external neodymium magnet which reassured both the high magnetic sensitivity as well as the simplicity of the solid liquid retrieval. This product was then washed to remove the bound arsenate and phosphate ions to reactive sites in the adsorbent. The exhausted ACF nanocomposite was gathered and disperses in 0.5 M NaOH solution and stirred gently, 60 min. The alkaline medium was able to disrupt successfully the inner-sphere Fe–O–As and Fe–O–P surface complexes, and the electrostatically bound oxyanions, and free those into the solution of desorption.²³ The sample was washed thoroughly in the deionized water after the desorption process until the neutral pH was obtained and a final wash was done in the 0.1 M NaCl in order to stabilize the surface charge sites and improve the structural integrity. The adsorbent was vacuum dried to 45 °C and reused. The operations repeated 6 times were adsorption–desorption with the following parameters: initial concentration 50 mg L⁻¹, pH 6.5, adsorbent dose 1.0 g L⁻¹, contact time 240 min, and temperature 25 °C. Efficiency and adsorption capacity after each cycle were recorded in an attempt to find out stability of operation. To further confirm the material stability, FTIR analyses of the regenerated adsorbent in the sixth cycle did not show significant changes in the characteristic functional groups or Fe–O vibrational features, which meant that the regeneration process did not cause any chemical degradation or a phase transformation. The maintenance of the magnetic properties also guaranteed uniform and fast separation efficiencies at each and every cycle.

2.6. Mathematical modeling

The evaluation of the sequestraining efficiency of the magnetically responsive alginate-chitosan Fe₃O₄ nano-architectures used was based on established quantitative measures in identifying the effectiveness of the adsorbent on As(v) and orthophosphate species. Removal efficiency (R, %) was calculated by the subsequent form: R = [(C₀ − C_e)/C₀] × 100, where C₀ and C_e are the initial and equilibrium concentrations of the target oxyanions in mg L⁻¹, respectively.²⁴ The equilibrium adsorption capacity, q_e (mg g⁻¹) which is used to estimate the amount of sorbate attached to each unit of the nano-architecture was determined to be: q_e = [(C₀ − C_e) × V]/m. V in this equilibrium represents the volume of the aqueous matrix in liters and m represents the mass of the dry adsorbent of the magnetic nature in grams. On the same note, instantaneous adsorption capacity at time t, which is denoted as q_t (mg g⁻¹), was calculated by replacing C_e with the concentration at time t (C_t).²⁵

In order to explain the rate-limiting processes and the character of the multi-interface sorption kinetics, three different models were used to analyze the experimental data. The pseudo-first-order model, which is founded on the premise that the rate of site occupancy in the adsorption process is proportional to the amount of unoccupied sites, is given as log(q_e − q_t) = log(q_e) − (k₁t/2.303). Here, k₁ (1/min) is the pseudo-first-order rate constant.²⁶ The pseudo-second-order model, usually refers to chemisorption by the valence forces based on the sharing or exchange of electrons between the adsorbent and sorbate, is presented as: t/q_t = 1/(k₂ × q_e²) + t/q_e. In addition, the mechanism of diffusion, which was also based on the Weber–Morris intraparticle diffusion, was also found by using the following: q_t = k_ip × t^0.5 + C. In this model the constant rate of diffusion inside the particle is k_ip (mg g⁻¹ min^−0.5) and C (mg g⁻¹) gives a product of the thickness of the boundary layer.²⁷

The isotherms of Langmuir, Freundlich and Temkin were used to determine the equilibrium between As(v) and orthophosphate between the solid and liquid phases. The Langmuir isotherm that assumes that monolayer coverage of a surface with finite number of identical sites would be: C_e/q_e = 1/(K_L × q_max) + C_e/q_max. Here, q_max (mg g⁻¹) will be the maximum monolayer and K_L (L mg⁻¹) will be the Langmuir constant associated with the adsorption energy. The Freundlich isotherm which is a non-ideal, reversible and multilayer adsorption on a heterogeneous surface is given as: ln(q_e) = ln(K_F) + (1/n) × ln(C_e), where K_F [(mg g⁻¹)(L mg⁻¹)^(1/n)] being the Capacity factor of Freundlich and 1/n is the intensity of adsorption. The Temkin isotherm (the indirect interactions between adsorbent and sorbate) is represented as q_e = B × ln(A) + B × ln(C_e), where B = RT/b, T is the absolute temperature (K), R is the universal gas constant (8.314 J mol⁻¹ K⁻¹), A (L mg⁻¹) is the equilibrium binding constant and b (J mol⁻¹) is the heat of adsorption.^28,29

Thermodynamic viability and the energetic character of the sequestration process were considered by determining Gibbs free energy (ΔG°), enthalpy (ΔH°), and entropy (ΔS°). The Gibbs free energy was calculated according to the equation: ΔG° = −RT × ln(K_c), K_c is the distribution coefficient (q_e/C_e). The dependence of the distribution coefficient on the temperature is also explained by Van't Hoff equation: ln(K_c) = (ΔS°/R) − (ΔH°/RT). ln(K_c) as a function of 1/T gave the values of ΔH° (kJ mol⁻¹) and ΔS° (J mol⁻¹ K⁻¹) by plot, respectively, slope and intercept, respectively. These parameters will be necessary in establishing the spontaneity of the adsorption and also identifying whether the process is exothermic or endothermic in the complex aqueous matrices. Coefficient of determination (R²) and non-linear chi-square (χ²) test were used to statistically validate the models to provide the best degree of fit of the observed experimental phenomena.³⁰

3. Result and discussion

3.1. Characterization results of alginate-chitosan-magnetite (ACF) nanocomposite

The given comprehensive characterization work is the detailed examination of the alginate-chitosan-magnetite (ACF) nanocomposite, which is created to eliminate pentavalent arsenic and orthophosphate in contaminated water simultaneously. The physicochemical, thermal stability, surface, and mechanical integrity of the material were assessed in a systematic way with the help of the state-of-the-art methods of analysis. The outcomes suggest that, the ACF nanocomposite is having the best balance in terms of high surface area, hydrophilic character, magnetic responsiveness and structural stability, which allows it to be the best candidate in the practical applications of water treatment.

3.1.1. Confirmation of interpenetrating double-network (IPN) formation

The formation of the interpenetrating double-network structure was confirmed through multiple complementary techniques. FTIR spectroscopy (Fig. 1) showed broadening of the O–H/N–H band (3400–3200 cm⁻¹) and a shift in the carboxylate peak (∼1600 cm⁻¹), indicating molecular–level interactions between the alginate and chitosan networks. The reduction in amine peak intensity (∼1590 cm⁻¹) confirmed glutaraldehyde cross-linking of chitosan, while the shifted carboxylate peak confirmed Ca²⁺ cross-linking of alginate. The persistence of characteristic peaks from both polymers confirmed the presence of both networks.

Fig. 1. FTIR chart of alginate, chitosan, magnetite (Fe₃O₄), and alginate-chitosan-magnetite (ACF) nanocomposite.

Thermal analysis (TGA/DSC, Fig. 8–10) revealed that the ACF nanocomposite exhibits a shifted decomposition profile to higher temperatures compared to individual biopolymers, with a residual weight of ∼22% at 800 °C. The glass transition temperature (T_g) appeared at 142–149 °C, significantly higher than pure alginate or chitosan, confirming restricted chain mobility due to IPN formation.

Fig. 8. Thermogravimetric analysis curve of alginate, chitosan, magnetite (Fe₃O₄), and alginate-chitosan-magnetite (ACF) nanocomposite.

Fig. 9. Differential thermogravimetry (DTG) curve of alginate-chitosan-magnetite (ACF) nanocomposite.

Fig. 10. DSC thermogram showing heat flow of alginate-chitosan-magnetite (ACF) nanocomposite.

Swelling and solubility testing (Table 1) showed that the ACF nanocomposite has dramatically reduced swelling (180% vs. 320–450% for individual polymers) and reduced solubility in both acetic acid (12% vs. 100% for chitosan) and EDTA (18% vs. 100% for alginate). This confirms that the two networks cannot be separated by selective dissolution, a defining characteristic of IPN structures. Mechanical testing (Table 1) demonstrated that the compressive strength of ACF IPN beads (2.4 MPa) is 3–5× higher than individual Ca-alginate (0.8 MPa) or chitosan (0.5 MPa) beads, demonstrating the synergistic reinforcement characteristic of IPN structures. Solid-state ¹³C CP-MAS NMR (Table 2) confirmed the presence of all characteristic peaks from both alginate and chitosan, with line broadening and chemical shift changes indicating molecular–level interactions between the two networks.

Table 1. Swelling, solubility, and mechanical properties of individual polymers and ACF IPN nanocomposite.

Sample	Water swelling (%)	Solubility in acetic acid (%)	Solubility in EDTA (%)	Compressive strength (MPa)	Elastic modulus (MPa)
Pure alginate	450 ± 25	100	100	0.8 ± 0.1	12.5 ± 1.2
Pure chitosan	320 ± 20	100	15	0.5 ± 0.1	8.3 ± 0.9
ACF IPN nanocomposite	180 ± 15	12 ± 3	18 ± 4	2.4 ± 0.2	38.7 ± 2.5

Table 2. Solid-state ¹³C CP-MAS NMR chemical shifts of alginate, chitosan, and ACF IPN nanocomposite.

Assignment	Alginate (ppm)	Chitosan (ppm)	ACF IPN (ppm)	Observation
C O (carboxyl)	175	—	175 (broadened)	H-bonding with chitosan
C1 (anomeric)	100	105	100–105 (merged)	Network interpenetration
C2–C5 (ring carbons)	65–80	55–85	55–85 (broadened)	Restricted mobility
CH3 (acetyl)	—	23	23 (reduced)	Deacetylation
C6 (CH2OH)	62	62	62 (shifted)	Cross-linking

Taken together, the FTIR, thermal, swelling/solubility, mechanical, and solid-state NMR results provide strong complementary evidence supporting the formation of an interpenetrating double-network structure, in which the alginate and chitosan networks are independently cross-linked and physically intertwined.

3.1.2. FTIR spectroscopy analysis

The FTIR spectrum (Fig. 1) indicates that magnetite, chitosan, and alginate have been successfully chemically integrated into the final ACF nanocomposite by following the changes in the characteristic vibrational modes. The wide absorption band at 3400–3200 cm⁻¹ in all the precursors and the composite is attributed to the overlapping of the O–H and N–H bands, which are significantly wider and stronger in the ACF nanocomposite because of the high level of intra- and intermolecular hydrogen bonding between the chains of chitosan and alginate. Alginate is specifically known to give a sharp peak at approximately 1600 cm⁻¹, which is an asymmetric stretching of carboxylate (–COO⁻) group and chitosan gives a characteristic peak at circa 1070 cm⁻¹, which is the C–O–C stretching of the pyranose ring.³¹ The successful insertion of magnetite into the ACF nanocomposites is demonstrated by the development of an intense absorption band at the low-wavenumber spectrum (lower than 600 cm⁻¹), which is the Fe–O stretching vibration of Fe₃O₄.³² Moreover, the changing as well as the relative intensity variation of the carboxylate peaks (alginate) and amino peaks (chitosan) on the composite indicate that magnetite particles are not simply dispersed physically, but rather it is probable that they are complexed with these functional groups to form a stabilized, multifunctional network. It is this strong chemical synergy that allows the ACF composite to possess good structural integrity, in addition to high surface reactivity in a wide range of environmental and biomedical applications.³³

3.1.3. TEM spectroscopy analysis

The transmission electron microscopy (TEM) analysis is direct observation of the morphological development of raw biopolymers to the composite of alginate-chitosan-magnetite (ACF) nanocomposite. Alginate (Fig. 2A) and chitosan (Fig. 2B) have a relatively amorphous porous structure as shown in the micrographs and irregular globular aggregates respectively. The magnetic image introduced by the magnetic alginate-chitosan-Fe₃O₄ (ACF) (Fig. 2C and D) image demonstrates that it has been successful in the incorporation of the dark and crystalline magnetite nanoparticles within the polymeric matrix. Such nanoparticles show a typical cubic or near-spherical geometry, which is a clear indication of a crystalline nature. A consistent hybrid structure can be observed in the finalized alginate-chitosan-magnetite (ACF) nanocomposite, where the biopolymer chains are seen to enclose the magnetic cores to give a stable and homogenous dispersion. This encapsulation is very important in ensuring that the magnetite particles are not oxidized and that the surface to volume ratio is kept high and the functional amine and carboxyl groups of the chitosan and alginate are easily accessible to the active site in specific applications like the removal of heavy metals.³⁴

Fig. 2. TEM images ((A): alginate, (B): chitosan, (C): Fe₃O₄, (D): ACF).

3.1.4. SEM spectroscopy analysis

The scanning electron microscopy (SEM) images (Fig. 3) offer essential data on the morphological change and the topography of the surface of the precursors and the obtained Alginate-chitosan-magnetite (ACF) nanocomposite. Alginate (Fig. 3A) is characterized by a highly porous interlaced fibrillar network in the form of a three dimensional scaffold that gives it a high internal surface area in which other components can be immobilized. On the other hand, chitosan (Fig. 3B) has a smaller, more compact, and rough flake-like topography.³⁵ The Magnetic Fe₃O₄ (Fig. 3C) sample shows a high concentration of small, rounded nanoparticles leading to a cluster, which offers the properties of magnetism needed by the composite. The overall characteristics of the final alginate-chitosan-magnetite (ACF) nanocomposite (Fig. 3D) are a much differentiated heterogeneous texture, with magnetite nanoparticles being closely incorporated in the system of alginate-chitosan. This morphological cooperation leads to a hybrid morphology containing a larger effective surface area and many protruding nodes, which is fundamental towards the optimization of the availability of active functional sites.³⁶ The strong bond between the magnetic core and the biopolymer shells does not only ensnare the oxidation of the magnetite but also develops a complex surface topography, which promotes the quick diffusion and mass transfer of pollutants during the adsorption mechanisms.

Fig. 3. SEM images (A): alginate, (B): chitosan, (C): Fe₃O₄, (D): ACF).

3.1.5. EDX spectroscopy

Energy-dispersive X-ray (EDX) spectroscopy is used to give both quantitative and qualitative information of the evolution of the elemental composition of the organic compounds to the completed (ACF) nanocomposite. According to the Fig. 4, raw chitosan shows mostly peaks related to organic carbon (C) and oxygen (O), which are characteristics of a carbohydrate-based polymeric structure. Meanwhile, peaks of high-intensity of iron (Fe) at a go of 0.7 keV (L-line) and 6.4 keV (K-alpha line), and a strong oxygen (O) signal, prove the presence of Fe₃O₄ phase. The (ACF) Composite spectrum is a certain hybridization of the chemical elements and still has the carbon (C) signals of the biopolymers in it but the typical iron (Fe) and calcium (Ca) signal is also present. The content of sodium (Na) and calcium (Ca) in the composite indicates that the ionic cross-linkage between the alginate and chitosan chains took place successfully, probably due to the synthesis reagents. Such a profound level of elemental hybridization confirms the presence of the magnetic nanoparticles as not being surface-adsorbed but chemically entrapped within the hydrogel scaffold to form a stable multimodal platform with the active sites necessary to provide high-performance pollutant remediation.³⁷

Fig. 4. EDX spectroscopy of chitosan, magnetite (Fe₃O₄), and alginate-chitosan-magnetite (ACF) nanocomposite.

3.1.6. BET surface area evaluation

Measurements of the adsorption–desorption isotherms of nitrogen were taken at 77 K and are given in Fig. 5. This indicates that the isotherm has typical Type IV behavior based on the IUPAC categorization, which is typical of mesoporous materials whose pores are in the range of 2–50 nm in diameter. A gradual increase followed by a sharp uptake with increasing relative pressures at low relative pressure (P/P₀ < 0.4), with monolayer-multilayer adsorption on the pore walls, and with intermediate relative pressure (0.4 < P/P₀ < 0.8) with capillary condensation in the mesopores. The desorption branch is not coincident with the adsorption branch, which forms a strong hysteresis loop, which remains until P/P₀ ≈ 0.42, typical in H3-type hysteresis. This hysteresis behavior is generally related to the aggregates of particles that create slit-like pores with non-rigid structures as is the case with the composite nature of the ACF material.³⁸

Fig. 5. N₂ adsorption–desorption isotherm of alginate, chitosan, magnetite (Fe₃O₄), and alginate-chitosan-magnetite (ACF) nanocomposite at 77 K.

The surface area of BET (determined by the linear part of BET plot) was obtained and was found to be 182.4 m² g⁻¹. It is significantly higher than the values obtained with pristine alginate (20–40 m² g⁻¹) or chitosan (15–35 m² g⁻¹) films, which confirms that the effects of the addition of Fe₃O₄ nanoparticles and the inotropic cross-linking process are significant in terms of porosity of the biopolymer matrix. The Langmuir surface area of alginate-chitosan-magnetite (ACF) nanocomposite that presupposes the monolayer coverage of all the accessible surfaces was 245.7 m² g⁻¹. At P/P₀ = 0.99 the total pore volume was 0.328 cm³ g⁻¹, of which 90.9% (0.298 cm³ g⁻¹) was of mesopores, 3.7% (0.012 cm³ g⁻¹) of micropores and 5.5% (0.018 cm³ g⁻¹) of macropores. These mesopores are predominant and this is very beneficial in adsorption applications, since the mesopores are large enough to permit rapid diffusion of the hydrated oxyanion species but still has a large surface area, so that the density of sites is large. Table 3 summarizes the porosity parameters. Fig. 6 shows the Barrett–Joyner–Halenda (BJH) pore size distribution, which is obtained by means of desorption branch of the isotherm. The distribution is bimodal with two peaks characterized by a high level of intensity around 6.5 nm and 9.5 nm, which portrays the existence of two distinct groups of mesopores. The mean diameter of the pore was evaluated by using the relationship 4V/A (V being the total pore volume and A being the BET surface area), which was found to be 8.2 nm. The pore size is suitable to oxyanion adsorption processes due to the fact that it is large enough to absorb hydrated ionic species of arsenate (effective diameter of the first hydration shell is 5–6 A) and phosphate (effective diameter of the first hydration shell is 4–5 A) and yet has high density of adsorption sites. A bimodal distribution can be attributed to two factors smaller pores (6.5 nm) in the cross-linked biopolymer network and larger pores (9.5 nm) between Fe₃O₄ nanoparticles and the polymer matrix.³⁹

Table 3. BET surface area and porosity parameters of alginate-chitosan-magnetite (ACF) nanocomposite.

Parameter	Value	Unit
BET surface area (SBET)	182.4	m2 g−1
Langmuir surface area	245.7	m2 g−1
Total pore volume	0.328	cm3 g−1
Micropore volume	0.012	cm3 g−1
Mesopore volume	0.298	cm3 g−1
Macropore volume	0.018	cm3 g−1
Average pore diameter	8.2	Nm

Fig. 6. BJH pore size distribution derived from the desorption branch of alginate, chitosan, magnetite (Fe₃O₄), and alginate-chitosan-magnetite (ACF) nanocomposite.

3.1.7. Particle size distribution

Dynamic light scattering (DLS) analysis is used to give a quantitative assessment of the hydrodynamic diameter and colloidal stability of the precursors as well as the synthesized ACF nanocomposite (in aqueous). All components are shown to have a relatively narrow, unimodal distribution, as shown by all curves in Fig. 7, suggesting that the synthesis process was controlled and there was limited large-scale aggregation. The raw alginate and chitosan polymers exhibit peak intensities that are clustered around 50–60 nm which corresponds to the size of the solvated chains of the polymers in a dilute medium. Interestingly, the magnetite (Fe₃O₄) nanoparticles are slightly smaller in size in terms of their hydrodynamic profile with a maximum around 40–45 nm. To the end ACF nanocomposite, the particle size distribution is shifted where the maximum distribution is at around 45 nm with the relative intensity of the maximum being greater than the single precursors. This particular sizing indicates that the alginate and chitosan layers coating the magnetite core are effective in encasing it so that it is a compact hybrid layer and not a loose aggregate.⁴⁰ The total distribution of the nanocomposite is more or less 25–100 nm range which is optimal in environmental application as it provides the best surface area of contact with the pollutant and high enough of the magnetic responsive range to recover fast.

Fig. 7. Dynamic light scattering particle size distribution of alginate, chitosan, magnetite (Fe₃O₄), and alginate-chitosan-magnetite (ACF) nanocomposite.

3.1.8. Thermogravimetric analysis (TGA)

Thermogravimetric analysis (TGA) of the ACF nanocomposite and its individual constituents was conducted under a nitrogen atmosphere at a heating rate of 10 °C min⁻¹ over the temperature range of 25–800 °C. The TGA profiles provide information on the thermal stability and decomposition characteristics of the biopolymer precursors and the resulting ACF nanocomposite, as shown in Fig. 8. All biopolymer-based materials exhibited a multi-stage degradation process. The first stage of weight loss occurred below 150 °C, which can be attributed to the evaporation of physically adsorbed and bound water, as well as the loss of loosely held water molecules. Pure alginate and chitosan showed pronounced weight losses between 200 and 500 °C, which are associated with thermal depolymerization and the degradation of functional groups, such as carboxyl and amine groups, within the polysaccharide backbones. In contrast, pure magnetite (Fe₃O₄) exhibited high thermal stability over the investigated temperature range, with only a minor weight loss.⁴¹

For the ACF nanocomposite, the thermal profile showed an increased residual mass of approximately 22% compared with the starting biopolymers. The total mass loss up to 800 °C amounted to approximately 78% of the initial mass, leaving a residual mass of approximately 22%. This residual mass represents the thermally stable fraction remaining after decomposition of the organic matrix and may include iron-containing inorganic species together with thermally generated carbonaceous residues. Moreover, the decomposition profile of the ACF nanocomposite shifted toward higher temperatures compared with its constituents, suggesting that ionic cross-linking between chitosan and alginate, together with the presence of the mineral phase, enhanced the overall thermal stability of the hybrid structure. This enhanced thermal stability is relevant to the structural durability of the composite during water-treatment applications and potential regeneration processes.⁴²

Detailed thermal degradation kinetics of the differential thermogravimetric (DTG) analysis Fig. 9 gives high-resolution information of the thermal degradation of the alginate-chitosan-magnetite (ACF) nanocomposite, indicating the exact temperatures of the maximum mass loss. According to the profile, the highest pathway of decomposition is in the form of a multi-stage decomposition reaction with the highest peak at 312.5 °C and the maximum rate of 0.2847% per min as tabulated in Table 4 that can be directly attributed to the rapid thermal scission of the C–O and C–N bonds in the biopolymer backbone. This pronounced degradation phase follows a preliminary small peak around 100 °C, which is associated with the loss of moisture and volatile matter enclosed within the hydrogel structure. Practically, the given thermal stability profile highlights the appropriateness of the (ACF) nanocomposite as a water treatment agent since the material can be both structurally and chemically stable even at operating temperatures that are below 200 °C. Moreover, the capacity of the composite to maintain the temperatures without experiencing any substantial structural degradation indicates that it could survive the regeneration cycles with moderate thermal conditions to guarantee the long term re-use of the adsorbent in removing wastewater in industry.⁴³ This thermal strength, together with effective incorporation of the magnetic Fe₃O₄ phase, makes the material a candidate that is durable and a high-performance material in the removal of organic and inorganic pollutants.

Table 4. Thermal decomposition stages in TGA analysis.

Stage	Temp. Range (°C)	Weight loss (%)	Assignment
I	25–100	3.0	Moisture evaporation
II	100–250	18.0	Bound water, volatiles
III	250–350	35.0	Biopolymer decomposition
IV	350–500	18.0	Carbonization
V	500–800	4.0	Residue formation

The thermal transitions and energy change that occurs during heating of the ACF nanocomposite were also studied using Differential scanning calorimetry (DSC) and essential information on the structural integrity of the material and the phase behavior was observed as shown in Fig. 10. The thermogram of DSC shows that there are various endothermic and exothermic processes: the initial endothermic peak of the thermogram at 84.7 °C (enthalpy: −142.5 J g⁻¹) is associated with the dehydration of the hydrogel matrix and an extensive endothermic region at 142–149 °C is the glass transition temperature (T_g) of the amorphous alginate-chitosan network. The occurrence of a major endothermic reaction at 295.2 °C with a high enthalpy of −285.7 J g⁻¹ affirms the primary thermal degradation reaction of the biopolymer backbone, and thereafter an exothermic peak at 408.3 °C (+62.3 J g⁻¹) which can be ascribed to oxidative degradation of carbon-based molecules.⁴⁴ The results are summarized in Table 5. Functional wise, these findings demonstrate that the material is very appropriate in water treatment, because the stability of the T_g and the energy demand to disintegrate it makes the ACF nanocomposite to retain its mechanical, as well as chemical properties in normal operating conditions of aqueous related procedures. The distinct division of the dehydration stage and the peak of the structural decomposition points to the solid hybrid structure that is able to withstand the thermal strains of industrial filtration and the repeated regeneration process.⁴⁵

Table 5. DSC thermal transitions of ACF nanocomposite.

Peak	Type	T onset (°C)	T peak (°C)	Enthalpy (J g−1)
1	Endothermic	72.3	84.7	−142.5
2	Endothermic	135.8	145.1	−38.6
3	Endothermic	278.5	295.2	−285.7
4	Exothermic	395.2	408.3	+62.3

3.1.9. XPS analysis

The elemental composition and chemical states of the magnetite and the final alginate-chitosan-magnetite (ACF) nanocomposite were investigated by use of X-ray photoelectron spectroscopy (XPS) Fig. 11 which gives conclusive evidence of the successful hybridization of the biopolymer-mineral network. The O 1s and Fe 2p characteristic peaks were identified on the spectrum of the survey of pure magnetite (Fe₃O₄), which validates the existence of iron oxide. Precisely, the Fe 2p region represents two different signals: Fe 2p_3/2 with a binding energy of about 711 eV and Fe 2p_1/2 at 725 eV, which is the mixed-valence state of Fe²⁺ and Fe³⁺ in a spinel structure.^46,47 The incorporation of the organic matrix into the ACF nanocomposite spectrum is observed in the appearance of two prominent new signals: the C 1s peak at 284.8 ev, indicating the backbone of the carbon chain of the organic structure, i.e., the alginate and chitosan, and the N 1s peak at 399.2 ev, which is a distinct signature of the amino groups (–NH₂) and protonated amines (–NH³⁺) as a result of the presence of the chitosan.⁴⁸ The O 1s peak at 531.5 eV in the composite is significantly stronger and wider than pure magnetite, which is a combination of the oxygen contributions of the iron oxide core and the large number of carboxyl and hydroxyl functional groups of the biopolymers.⁴⁹ The continued presence of the Fe 2p signals in the composite at even a lower relative intensity because of the biopolymer coating proves the successful sequestriction of the magnetite nanoparticles to the hybrid matrix. It has been established that this surface chemical analysis is very detailed showing that the ACF nanocomposite is abundant in both magnetic properties and various sites of functionality, which are vital in the effective adsorption of multi-class pollutants in the water treatment process.

Fig. 11. XPS spectroscopy of magnetite (Fe₃O₄), and alginate-chitosan-magnetite (ACF) nanocomposite.

3.1.10. Magnetic properties and magnetic separation performance

The magnetic properties of the ACF nanocomposite were investigated using a Lake Shore 7404 Vibrating Sample Magnetometer (VSM) at room temperature (298 K) over an applied magnetic field range of −15 000 to +15 000 Oe. The resulting complete magnetization–magnetic field (M–H) hysteresis loop is presented in Fig. 12, while the enlarged low-field region enables a more accurate assessment of the coercive field and remanent magnetization.

Fig. 12. M–H hysteresis loop of the ACF nanocomposite measured at 298 K under an applied magnetic field of −15 000 to +15 000 Oe. The inset shows the low-field region, highlighting the negligible coercivity (H_c) and remanent magnetization (M_r).

The ACF nanocomposite exhibited a narrow and nearly reversible M–H loop, The coercive field (H_c) was approximately 0 Oe, while the remanent magnetization (M_r) was approximately 0 emu g⁻¹. The negligible coercivity and remanence indicate that the material retains very little magnetic memory after removal of the external magnetic field and exhibits superparamagnetic-like behavior at room temperature. Such magnetic characteristics are particularly advantageous for adsorbent recovery because the composite can be rapidly collected under an external magnetic field while minimizing magnetic agglomeration after the field is removed.

The saturation magnetization (M_s) of the ACF nanocomposite was approximately 26.1 emu g⁻¹, as determined from the VSM curve. This value is lower than that typically reported for bulk magnetite (approximately 80–92 emu g⁻¹), which is reasonable for a hybrid polymeric composite in which the magnetic iron-containing phase is incorporated within a non-magnetic alginate/chitosan-based matrix. Importantly, the measured (M_s) value is consistent with the magnetic response observed during the separation experiments described below.

The magnetic separation capability of the ACF nanocomposite was quantitatively evaluated using an aqueous suspension containing 1.0 g L⁻¹ ACF in 50 mL of water at pH 6.5, employing a NdFeB permanent magnet (0.5 T, N42). The residual concentration of suspended magnetic material was monitored at different separation times, and the recovery ratio was calculated according to:where (C₀) and (C_t) represent the initial and residual concentrations of suspended ACF at time zero and separation time (t), respectively. The time-dependent magnetic separation results are summarized in Table 6.

Table 6. Magnetic separation efficiency of ACF nanocomposite vs. time.

Time (s)	Residual particle concentration (%)	Recovery ratio (%)	Supernatant appearance
0	100	0	Turbid (black)
5	45	55	Turbid
10	22	78	Slightly turbid
15	10	90	Slightly turbid
20	4	96	Nearly clear
30	<1	>99	Clear
45	<0.5	>99.5	Clear
60	<0.1	>99.9	Clear

At the beginning of the experiment, the suspension was highly turbid owing to the dispersed ACF particles. Magnetic separation proceeded rapidly during the first few seconds, with recovery increasing from 55% after 5 s to 78% after 10 s and 90% after 15 s. After 20 s, the recovery reached approximately 96%, while the suspension became nearly clear. At 30 s, more than 99% of the ACF was recovered, and the aqueous phase appeared visually clear. Further extension of the separation time to 45 and 60 s resulted in recovery values exceeding 99.5% and 99.9%, respectively.

These results clarify the apparent difference between the previously stated separation times. Thirty seconds represents the practical magnetic separation time required to achieve more than 99% recovery, whereas 60 s corresponds to near-complete recovery (>99.9%). Therefore, the two values describe different recovery criteria rather than contradictory experimental results. The rapid separation is attributed to the magnetic response of the Fe-containing phase embedded within the ACF network and demonstrates the practical advantage of magnetic recovery over conventional solid–liquid separation methods.

From a mechanistic perspective, the magnetic force acting on a magnetically responsive particle can be expressed as:where (χ) is the magnetic susceptibility, (V) is the particle volume, (B) is the magnetic flux density, (∇B) is the magnetic-field gradient, and (µ₀) is the permeability of free space. The magnetic force is opposed by the hydrodynamic drag force:F_d = 6π_rηνwhere (η) is the viscosity of the surrounding liquid, (r) is the effective particle radius, and (v) is the particle velocity. Under the condition (F_m=F_d), the particle velocity can be approximated by:

These relationships indicate that increasing magnetic susceptibility and magnetic-field gradient enhances particle migration toward the magnet, thereby facilitating rapid solid–liquid separation.

To further assess the consistency between the magnetic response and the iron content of the composite, the Fe content determined by ICP-OES was considered. The measured Fe concentration was 18.4–18.6 wt%. Assuming, as a simplifying approximation, that all of the detected Fe is present as Fe₃O₄, the corresponding Fe₃O₄-equivalent content can be estimated using the molecular-weight ratio:Fe₃O₄/Fe = 231.53/167.55 = 1.382

Thus, an Fe content of approximately 18.5 wt% corresponds to about 25.6 wt% Fe₃O₄-equivalent. Using the saturation magnetization of bulk magnetite (approximately 92 emu g⁻¹) as an upper reference gives an estimated composite magnetization of:M_s sestimated = 0.256 × 92 ≅ 23.6 emu g⁻¹

The experimentally measured value of approximately 26.1 emu g⁻¹ is reasonably close to this estimate. This agreement supports the consistency between the VSM and ICP-OES results and does not require assuming the presence of an additional magnetic phase. Nevertheless, the ICP-OES measurement quantifies total elemental Fe and cannot distinguish between individual iron-containing phases. Therefore, the calculated 25.6 wt% should be regarded as an Fe₃O₄-equivalent content rather than a direct phase-specific quantification of magnetite. Definitive identification and quantification of individual iron oxide phases would require complementary phase-specific techniques such as Mössbauer spectroscopy or X-ray absorption spectroscopy.

The thermal analysis provides additional information concerning the inorganic fraction of the ACF nanocomposite. The TGA curve showed a residual mass of approximately 22% at 800 °C under nitrogen. This residual mass should not be directly equated with the Fe₃O₄ content estimated from ICP-OES because TGA residue represents the total thermally stable fraction remaining after decomposition of the organic components and may contain iron-containing inorganic species as well as thermally generated carbonaceous residues. In addition, the chemical state of iron-containing phases may change under prolonged thermal treatment. Accordingly, the TGA result is considered supportive evidence for the presence of a substantial thermally stable inorganic fraction, whereas ICP-OES provides the basis for estimating the Fe₃O₄ equivalent content.

Overall, the VSM results demonstrate a low-coercivity, low-remanence magnetic response with a saturation magnetization of approximately 26.1 emu g⁻¹, while the time-resolved separation experiment demonstrates >99% magnetic recovery within 30 s and >99.9% recovery within 60 s. The combination of negligible magnetic hysteresis, measurable saturation magnetization, and rapid magnetic recovery confirms the suitability of the ACF nanocomposite for efficient post-treatment solid separation and potential repeated use in aqueous adsorption applications.

3.2. Systematic evaluation of alginate-chitosan-Fe₃O₄ nanocomposite for simultaneous As(v) and phosphate removal from aqueous solutions

The adsorption capacity of the magnetically responsive alginate-chitosan-Fe₃O₄ (ACF) nanocomposite was comprehensively quantified by a set of batch adsorption studies aimed at explaining the effects of the most important working parameters on the removal efficiency of pentavalent arsenic [As(v)] and orthophosphate [PO₄³⁻] in water. The experiment design involved the solution pH, dosage of adsorbent, contact time, concentration of pollutant in the beginning, temperature, and the existences of competing ion; all these played a significant role in the adsorption process and evaluation on whether the material would be practical in the treatment of water. All parameters were explored in triplicate in order to make sure that it was statistically reliable, and the mean values and standard deviations were reported everywhere. These findings were also measured by using the known models of isotherm, kinetic, and thermodynamic to give quantitative data in the adsorption capacity, rate-limiting processes, and energetic viability of the process.

3.2.1. Synergy evaluation: comparison with individual components, physical mixtures, and single-cross-linked networks

To justify the synergy claim, the ACF IPN nanocomposite was compared with its individual components (alginate, chitosan, Fe₃O₄), a physical mixture of alginate and chitosan, and single-crosslinked networks (Alg-Ca and Chi-GA) under identical conditions (C₀ = 50 mg L⁻¹ each of As(v) and PO₄³⁻, pH 6.5, dose = 1.0 g L⁻¹, T = 25 °C, contact time = 240 min). The results are summarized in Table 7.

Table 7. Comparative adsorption capacities of individual components, physical mixtures, single-crosslinked networks, and ACF IPN nanocomposite.

Sample	As(v) qe (mg g−1)	PO43−qe (mg g−1)	Total qe (mg g−1)	Synergy factor (SF)a
Alginate (Alg)	28.5 ± 2.1	35.2 ± 2.5	63.7	—
Chitosan (chi)	42.3 ± 2.8	48.6 ± 3.1	90.9	—
Fe3O4	18.2 ± 1.5	22.4 ± 1.8	40.6	—
Alg + Chi physical mixture	52.6 ± 3.2	61.8 ± 3.5	114.4	0.74
Alg-Ca (single-crosslinked)	58.4 ± 3.5	68.2 ± 3.8	126.6	0.82
Chi-GA (single-crosslinked)	65.2 ± 3.8	75.4 ± 4.2	140.6	0.91
ACF (this work)	114.5 ± 2.1	138.2 ± 2.5	252.7	1.29

^aSynergy factor (SF) = q_e(ACF)/[q_e(Alg) + q_e(Chi) + q_e(Fe₃O₄)] = 252.7/195.2 = 1.29 (vs. sum of individuals), SF vs. physical mixture = 252.7/114.4 = 2.21, SF vs. best single-crosslinked = 252.7/140.6 = 1.80.

The individual components showed low adsorption capacities: alginate (63.7 mg g⁻¹ total), chitosan (90.9 mg g⁻¹ total), and Fe₃O₄ (40.6 mg g⁻¹ total). The physical mixture of alginate and chitosan showed improved capacity (114.4 mg g⁻¹ total) due to the combined functional groups, but this was still far below the ACF nanocomposite. The single-crosslinked networks (Alg-Ca and Chi-GA) showed better capacity (126.6 and 140.6 mg g⁻¹ total, respectively) due to the stabilization of the polymer network, but again much lower than ACF.

The ACF IPN nanocomposite exhibited a total adsorption capacity of 252.7 mg g⁻¹ (114.5 mg g⁻¹ for As(v) and 138.2 mg g⁻¹ for PO₄³⁻), which is:

1.29× the sum of individual components (195.2 mg g⁻¹).

2.21× the physical mixture (114.4 mg g⁻¹).

1.80× the best single-crosslinked network (140.6 mg g⁻¹).

The synergy factor (SF > 1) confirms that the ACF IPN nanocomposite exhibits true synergistic enhancement, not merely additive behavior. The synergy arises from the interpenetrating double-network structure, the in situ Fe₃O₄ incorporation, the dual cross-linking, the complementary functional groups, and the enhanced mass transport through the mesoporous structure.

3.2.2. Effect of solution pH on adsorption performance

The pH of the solution is generally considered to be one of the most important parameters in oxyanion adsorption reactions, since it not only controls the surface charge properties of the adsorbent, but also it controls the speciation of the adsorbate species in solution. The influence of the pH of initial solutions on the adsorption of As(v) and phosphate on the ACF nanocomposite was examined at pH of 2.0 to 10.0 and the results are shown in Fig. 13. Each of the experiments was performed at an initial pollutant concentration of 50 mg L⁻¹, adsorbent dose of 1.0 g L⁻¹, at 25 °C, and 240 minutes of contact time to ascertain equilibrium achievement.

Fig. 13. Effect of solution pH on the adsorption capacity of ACF nanocomposite for As(v) and PO₄³⁻. Experimental conditions: C₀ = 50 mg L⁻¹, dose = 1.0 g L⁻¹, T = 25 °C, contact time = 240 min. Error bars represent standard deviation (n = 3).

Both phosphate and As(v) are shown to be pH dependent adsorbents with different optimal regions as shown in Fig. 13. In the case of As(v), the adsorption capacity rises with pH between 50.3 ± 2.8 mg g⁻¹ at pH 2.0 to a peak of 114.5 ± 2.1 mg g⁻¹ at pH 6.5 after which it decreases gradually to 75.8 ± 4.2 mg g⁻¹ at pH 10.0. The phosphate adsorption shows the same tendency with a maximum capacity of 138.2 ± 2.5 mg g⁻¹ at pH 6.0 with a lower value of 60.2 ± 3.5 mg g⁻¹ at pH 2.0 and 82.5 ± 5.8 mg g⁻¹ at pH 10.0. Although the maximum individual adsorption capacities were obtained at pH 6.5 for As(v) (114.5 ± 2.1 mg g⁻¹) and pH 6.0 for orthophosphate (138.2 ± 2.5 mg g⁻¹), pH 6.5 was selected as the optimum operating condition for simultaneous removal of both oxyanions because it provided the best overall adsorption performance. The vertical dashed line at pH 6.72 is the point of zero charge (pHpzc) of the solution as indicated by measurement of the zeta potential which is used as a critical reference to interpret the electrostatic interactions that drive the adsorption process.

The measurement of zeta potential was done at a pH of between 2 and 10 in 0.01 M NaCl as the electrolyte solution to observe the nature of surface charges of ACF nanocomposite as shown in Fig. 14. The zeta potential profile shows a systematic decline in potential values of positive to negative with a pH change to zero and onward to pH 6.72. This value is the point of zero charge (pHpzc), at which the surface is neutral in terms of electrical charge. The surface is positively charged in a pH value less than 6.72 since the amino groups (–NH₂ → –NH₃⁺) of chitosan are protonated and the highest zeta potential value is +12.2 mV at pH 4.0. At higher pH pHpzc, the surface is made negatively charged with carboxyl groups of alginate, which deprotonate (–COOH → –COO⁻), and attain a maximum −28.6 mV pH 9.0.

Fig. 14. Zeta potential of ACF nanocomposite as a function of pH.

The adsorption behavior of the observed pH-dependency may be explained with references to the surface charge and adsorbate speciation. At pH below the pHpzc (pH < 6.72), the protons of the magnetite surface and the amino functionalities of chitosan (–NH₂ + H⁺ → –NH₃⁺) protonate to provide the ACF nanocomposite surface with a net positive charge. This has been discovered to provide good electrostatic environment to the negatively charged species of arsenate that adsorb on the surface owing to this positive charge on the surface. In the pH 4.0–7.0, arsenic exists mostly in a form of H₂AsO₄⁻ (pK_a1 = 2.2, pK_a2 = 6.9, pK_a3 = 11.5), and of a strongly electrostatically bound entity with the protonated surface locations. On the same note, in this pH range, phosphate is mostly in the form of H₂PO₄⁻ and HPO₄²⁻ (pK_a1 = 2.1, pK_a2 = 7.2, pK_a3 = 12.3) and that is why the adsorption rate was high at neutral pH.⁵⁰

The decreasing adsorption capacity even when the positive surface charge is strong (pH < 3.0) may be explained by a number of factors. To begin with, over protonation causes the repulsion of the positively charged surface sites of the arsenious acid by electrostatic repulsion, and the creation of neutral arsenious acid (H₃AsO₃) species. Second, pH modifiers like Cl- and SO₄²⁻ which are in high concentrations take up binding sites. Third, the high acidity level can lead to the partial dissolution of the surface functional groups, and the density of the effective adsorption sites will decrease. When the pH becomes alkaline (pH 8.0 or above), deprotonation of functional groups (–NH₃⁺ → –NH₂, –OH → –O⁻), which causes a more negative charge on the surface, results in repulsion of anionic arsenate (HAsO₄²⁻, AsO₄³⁻) and phosphate (HPO₄²⁻, PO₄³⁻) Also, competitive adsorption by hydroxyl ions (OH) would take place when the pH is high and this also decreases the number of sites to which target pollutants can be adsorbed.⁵¹

The reason why the adsorption capacity with phosphate was slightly higher than that with As(v) in most pH scales is that the affinity of the ACF surface to phosphate species was greater. This selectivity is probably due to the creation of more stable inner-sphere complexes between inner-sphere phosphate and iron oxide positions using bidentate coordination.⁵² The error bars in Fig. 13 show that the most precise measurements are observed around the range of optimal pH (pH 5.0–7.0), and that the variability increases at the ends of the pH range because the surface charge and speciation equilibrium changes rapidly. Besides, the zeta potential profile is an important consequence of the adsorption of oxyanions. The composite surface maintains a net positive charge at the working pH regime of 4.0–6.5, and this is favorable to the electrostatic attraction and binding of the negatively charged arsenic species (H₂AsO₄⁻, HAsO₄²⁻) and phosphate ions (H₂PO₄⁻, HPO₄²⁻). The assessment of colloidal stability shows excellent dispersion stability (|ζ| > 30 mV) at pH < 3.5 and pH > 7.5, good stability (20 < |ζ| < 30 mV) at pH 3.5–5.0 and pH 7.0–7.5, and incipient instability (|ζ| < 20 mV) These results confirm the fact that the ACF nanocomposite retains surface charge properties that are optimal in oxyanion sequestration of slightly acidic to neutral pH. As a result, pH 6.5 was chosen to be the best value in the course of conducting further adsorption experiments since it offers a good balance between choosing an optimal removal capacity and minimizing the operation costs of changing pH of large-scale wastewater treatment.

3.2.3. Effect of adsorbent dose on removal efficiency

Adsorbent dosage is an important economic parameter that determines the practical water treatment application since it directly influences the operational cost, as well as, the treatment efficiency. Systematic study of the dosage of ACF nanocomposites on the removal of As(v) was conducted by changing the concentration of the adsorbent between 0.2 and 2.5 g L⁻¹ but keeping the initial concentration of the pollutant (50 mg L⁻¹), pH (6.5), temperature (25 °C), and contact time (240 minutes) constant. The adsorption capacity (q_e) vs. adsorbent dose dual-panel presentation in Fig. 15 shows that the competing effects of adsorbent dose on the adsorption capacity (q_e) and removal efficiency are competing.

Fig. 15. Effect of adsorbent dose on adsorption capacity and removal efficiency for As(v) (A and B), and PO₄³⁻(C and D) experimental conditions: C₀ = 50 mg L⁻¹, pH = 6.5, T = 25 °C, contact time = 240 min. Error bars represent standard deviation (n = 3).

This gives an inverse relationship between adsorbent dosage and adsorption capacity as shown by Fig. 15(A). The maximum adsorption capacity of 228.5 ± 8.2 mg g⁻¹ is attained at the lowest dose of 0.2 g L⁻¹ but gradually declines to 114.5 ± 2.8 mg g⁻¹ at 1.0 g L⁻¹ and the lowest adsorption capacity is 45.8 ± 3.5 mg g⁻¹ at 2.5 g L⁻¹. On the other hand, Fig. 15(C) shows that PO₄³⁻ adsorption capacity also has a similar decreasing pattern with start and end points being approximately 180 mg g⁻¹ and reduction with increase of dose respectively. This is a range of monotonic decrease in q_e as amount of dose increases, and is documented in the scientific literature of adsorption and also is a result of the principle of mass distribution. Low concentrations of adsorbent pollutant are spread evenly at low concentration of adsorbent (mass of adsorbent) (50 mg L⁻¹ × volume) over a comparatively small mass of adsorbent, causing a high concentration of pollutant per unit mass. However, in high doses of adsorbent, the same fixed quantity of pollutant is spread on a very large mass of adsorbent, inevitably decreasing the per-mass capacity although more total active sites are accessible.⁵³

The growing standard deviation of the higher adsorbent doses (large error bars in Fig. 15(A and C) is indicative of experimental difficulties with concentrated particle suspension. When the concentration exceeds 1.5 g L⁻¹, there is significant particle–particle interactions that result into the effects of aggregation that causes the effective surface area to become smaller thereby lowering the adsorption capacity. The aggregates form diffusion barriers which tend to hamper access of pollutants to interior binding sites leading to low utilization efficiency and high variability in measurement. Although the magnetic character of ACF nanocomposite is advantageous in separation process, it also can cause aggregation of particle at extremely high density due to magnetic attraction which makes this phenomenon to be even worse.⁵⁴

In sharp contrast, Fig. 15(B) demonstrates that the removal efficiency is higher with adsorbent dose, which starts with 45.7 ± 2.8% at 0.2 g L⁻¹, then gradually reaches its highest point of 96.2 ± 1.2% at 1.5 g L⁻¹ and then becomes constant. This is attributed to the availability of more and more adsorption sites as dose increases.⁵⁵ At 0.2 g L⁻¹, the few binding sites become saturated and all of the molecules of the pollutant are not eliminated and a large amount of concentration is left in the solution. With a continued rise in the dose up to 1.0 g L⁻¹, one has enough sites to bind most of the pollutant, with a 91.5 ± 2.1% removals. Adding 0.5–2.5 g L⁻¹ gives the system excess capacity, and the removal efficiency levels off at around 96–98% as the adsorbent drains out of the solution the decreasing returns above 1.5 g L⁻¹ shows that adding further adsorbents does not offer much value to the system, but the adsorbents cost is not justified. Although the ratio of removal efficiency of phosphate in Fig. 15(D) shows a clear positive relationship with the mass of adsorbent, increasing between 38% at 0.2 g L⁻¹ to more than 97% at 2.5 g L⁻¹. In a practical sense, the choice of the best dose of adsorbent involves a tradeoff between the effectiveness of the treatment and economic factors. Although the 2.5 g L⁻¹ has slightly better removal (97.8 ± 1.5%), the difference between the 2.5 g L⁻¹ and the 1.0 g L⁻¹ (91.5 ± 2.1) is not worth the 150 percent increase in the amount of material used. Thus, the adsorbent dose of 1.0 g L⁻¹ was found as optimal operating point to be used in the further experiments, which offers good removal efficiency (>90%), acceptable adsorption capacity (114.5 mg g⁻¹) and cost effective consumption of material used. The suspension viscosity of this dose is also acceptable to handle and reduces the complexity of aggregation.⁵⁶

3.2.4. Adsorption kinetics and contact time optimization

The time-dependent analysis of adsorption is important in the design of continuous-flow treatment systems. The time dependence of capacity (q_t) during the initial contact and up to the equilibrium was studied and plotted in Fig. 16. Contact time was assessed to determine the impact of contact time on the pollutant sequestration of As(v) and PO₄³⁻ onto the ACF nanocomposite and it was found that there was a rapid initial uptake phase (0–30 min) with As(v) reaching 68.5 ± 5.2 mg g⁻¹ (59.8% of equilibrium) and phosphate reaching 85.3 ± 6.8 mg g⁻¹ (61.7%). The high concentration gradients and large number of vacant sites lead to this rapid stage which is succeeded by an approach to equilibrium that is asymptotic. The kinetics of both oxyanions were surprisingly rapid, with more than 80% of the overall capacity being reached in the first 100 minutes, presumably because the highly porous alginate-chitosan scaffold and the Fe₃O₄ nanoparticles cited in the previously mentioned morphological studies provided a large number of active sites on the surfaces.⁵⁷

Fig. 16. Adsorption kinetics showing contact time effect on As(v) and PO₄³⁻ uptake onto ACF nanocomposite. Conditions: C₀ = 50 mg L⁻¹, pH = 6.5, dose = 1.0 g L⁻¹, T = 25 °C. Symbols: experimental data (mean ± SD, n = 3); dashed lines: pseudo-second-order model fit. Error bars represent standard deviation (n = 3).

Under the prescribed conditions (C₀ = 50 mg L⁻¹, pH = 6.5, dose = 1.0 g L⁻¹, T = 25 °C), the maximum adsorption capacities were about 115 ± 2.1 mg g⁻¹ in As(v) and 138.2 ± 2.5 mg g⁻¹ in PO₄³⁻ at equilibrium (240 minutes). This could be explained by the fact that phosphate has a higher charge density at pH 6.5; when attempting to have stronger electrostatic interactions with the protonated amine moieties of the chitosan matrix and the hydroxylated surface of the magnetite core. This high-capacity and fast performance implies that the (ACF) nanocomposite follows chemisorption, including inner-sphere complexation, and would make it an outstanding product when it comes to treating real-world water in which short residence times may be necessary.⁵⁸

3.2.5. Adsorption isotherms and maximum capacity

Adsorption isotherms can give data regarding the distribution of adsorbates in both the solid and liquid phases at the equilibrium. Isotherms were used to plot the different initial concentration of dose 1.0 g L⁻¹, time 240 min, and pH 6.5 at temperature 25 °C with a dose of 10–500 mg L⁻¹. The adsorption isotherm of As(v) and PO₄³⁻ on (ACF) nanocomposite were studied in an attempt to determine the maximum adsorption capacity and the mode of interaction between the oxyanions and the adsorbent surface Fig. 17. Since the experimental results show that the adsorption capacity (q_e) rises exponentially on the low equilibrium concentrations (C_e), then it curves off a curve, such as is typical of the high-affinity adsorption processes. The experimental data were found to be within close range of the Langmuir model (dashed lines) which was indicating that the sequestration process takes place in the monolayer adsorption style at a constant number of energetically comparable active sites.⁵⁹

Fig. 17. Adsorption isotherms with Langmuir model fits. Symbols: experimental data (mean ± SD, n = 3); dashed lines: Langmuir model. Conditions: pH = 6.5, dose = 1.0 g L⁻¹, T = 25 °C, time = 240 min.

These strengths are higher than most of the reported adsorbents, which can be explained by the synergistic effect of high surface area (182.4 m² g⁻¹), mesoporous structure, and the synergistic effect of the alginate-chitosan matrix, which provides high concentrations of amines and carboxyl groups, and the magnetite (Fe₃O₄) nanoparticles, which offer selective coordination sites where the oxyanions can bind. The affinity slightly increased towards phosphate could be because of its lesser ionic radius or more intense hydrogen bonding of the protonated amine groups on the chitosan at pH 6.5. The correlation with the Langmuir model, along with the large values of q_max shows that the (ACF) nanocomposite is an effective high-capacity adsorbent that can be used to remediate the arsenic and phosphate contamination of the water source.⁶⁰

3.2.6. Temperature effect and thermodynamic analysis

Adsorption is affected by temperature in terms of diffusion kinetics, surface activity and thermodynamics. The temperature effect was investigated at 15–55 °C, at C₀ = 50 mg L⁻¹, pH 6.5, dose 1.0 g L⁻¹, and time 240 min.

The temperature effect of the encroachment of As(v) and PO₄³⁻ was assessed and found to exhibit a unique endothermic character of the encroachment process on the (ACF) nanocomposite. To ascertain the effect of temperature on the equilibrium adsorption capacity (q_e), the temperature of the mixture was raised, as shown in Fig. 18(A and B), where a sharp rise in As(v) uptake, 110.2 ± 2.8 mg g⁻¹, and PO₄³⁻ uptake, 83 mg g⁻¹, was observed when the temperature was increased to 55 °C. A calculated enthalpy change of ΔH° = (+) endothermic indicated that this is a chemically controlled process (chemisorption) whereby higher temperatures increase the diffusion rate of oxyanions into the internal pores of the alginate-chitosan matrix and decreases the activation energy barrier to bond formation. It can be assumed that the identified trend may be attributed to a higher mobility of the arsenate and phosphate ions at higher temperatures as well as to the possible expansion of the biopolymer network that reveals more active amine and hydroxyl sites to use. Also, the endothermic nature could be attributed to the dehydration of the oxyanions prior to their coordination with the iron oxide surface which involves a lot of energy input.⁶¹ This thermodynamic profile is used to establish the fact that the (ACF) nanocomposite is quite effective in changing the thermal conditions of the environment and is more efficient in treating warm industrial effluents, which is why it can be applied in water treatment practice.

Fig. 18. Temperature effect on (A) adsorption capacity of As(v), (B) adsorption capacity of PO₄³⁻ showing endothermic behavior. Conditions: C₀ = 50 mg L⁻¹, pH = 6.5, dose = 1.0 g L⁻¹, time = 240 min. Error bars: SD (n = 3).

3.3. Selectivity, competing-ion effects, and quantitative competitive adsorption

The selectivity of the ACF nanocomposite toward As(v) and PO₄³⁻ was evaluated in the presence of representative coexisting ions to assess its performance under chemically complex aqueous conditions. For As(v), the influence of common anions and cations was investigated at pH 6.5 using an initial As(v) concentration of 50 mg L⁻¹. As shown in Fig. 19, the control system, containing As(v) in the absence of competing ions, exhibited an adsorption capacity of approximately 117 mg g⁻¹. The presence of monovalent anions, including Cl⁻ and NO₃⁻ at concentrations of 50–200 mg L⁻¹, and divalent cations such as Ca²⁺ and Mg²⁺ at 100 mg L⁻¹ resulted in only minor changes in the adsorption capacity. This limited response to common background ions suggests that the adsorption of As(v) is not governed predominantly by nonspecific electrostatic ion exchange.

Fig. 19. Effect of co-existing ions on As(v) adsorption capacity. Colors indicate severity: green (control), light blue (minimal), gold (moderate), and coral (significant). Conditions: C₀ (As) = 50 mg L⁻¹, pH = 6.5, dose = 1.0 g L⁻¹. Error bars: SD (n = 3).

A moderate decrease in As(v) uptake was observed in the presence of SO₄²⁻ (50 mg L⁻¹) and HCO₃⁻ (100 mg L⁻¹), for which the adsorption capacities decreased to approximately 105–110 mg g⁻¹. More pronounced inhibition occurred at higher SO₄²⁻ concentration (200 mg L⁻¹) and in the presence of PO₄³⁻ (50 mg L⁻¹), with phosphate producing the greatest reduction, lowering the capacity to approximately 100 mg g⁻¹. The pronounced competitive effect of PO₄³⁻ is consistent with its chemical and structural similarity to arsenate and its ability to interact with surface Fe–OH groups and protonated functional groups within the chitosan-containing matrix. Nevertheless, the ACF nanocomposite retained more than 85% of its initial As(v) adsorption capacity under the most challenging conditions investigated, demonstrating relatively high tolerance toward competing species commonly encountered in aqueous systems.⁶²

The competitive behavior of PO₄³⁻ was further examined using a set of coexisting anions at 100 mg L⁻¹ to assess the selectivity of the ACF nanocomposite in multicomponent aqueous systems Fig. 20. Under the control conditions (C₀ = 100 mg L⁻¹, pH = 6.5, adsorbent dose = 1.0 g L⁻¹), the maximum adsorption capacity for PO₄³⁻ was 96.1 mg g⁻¹. The addition of competing anions resulted in a progressive decrease in PO₄³⁻ uptake, although the magnitude of inhibition varied considerably among the tested species. Monovalent anions, particularly Cl⁻ and NO₃⁻, caused only minor reductions, with adsorption capacities of 94.3 and 92.6 mg g⁻¹, respectively. In contrast, SO₄²⁻ and CO₃²⁻ reduced the capacity to 87.2 and 81.4 mg g⁻¹, respectively, while SiO₃²⁻ produced the strongest inhibition, decreasing the capacity to 76.3 mg g⁻¹.

Fig. 20. Influence of various co-existing anions (100 mg L⁻¹) on the adsorption capacity (q_e) for PO₄³⁻. The “control” represents the system with only PO₄³⁻ present. The decreasing trend highlights the competitive inhibition of active sites by interfering ions, particularly SiO₃²⁻. Conditions: C₀ (PO₄³⁻) = 100 mg L⁻¹, pH = 6.5, dose = 1.0 g L⁻¹. Error bars: SD (n = 3).

The stronger competitive effects of multivalent oxyanions may arise from a combination of their charge characteristics, surface affinity, and ability to interact specifically with reactive sites associated with Fe–OH groups and protonated functional groups in the alginate–chitosan matrix. The particularly pronounced effect of SiO₃²⁻ suggests that specific surface interactions may contribute to the competition in addition to nonspecific electrostatic effects. Importantly, the ACF nanocomposite retained approximately 80% of its PO₄³⁻ adsorption capacity even in the presence of the strongest interfering ion investigated, indicating substantial resistance to competitive inhibition in multicomponent aqueous environments.⁶³

To provide a quantitative assessment of the competitive effects, the adsorption behavior of As(v) and PO₄³⁻ was further evaluated using distribution coefficients (K_d), equal-mass versus equal-molar comparisons, and controlled ionic-strength experiments. The calculated K_d values are summarized in Table 8. In the absence of competing ions, the K_d values were 2.29 and 2.76 L g⁻¹ for As(v) and PO₄³⁻, respectively, corresponding to a K_d(P)/K_d(As) ratio of 1.21. The presence of Cl⁻ and NO₃⁻ at 100 mg L⁻¹ caused only small changes in the distribution coefficients, with Kd values of 2.25 and 2.22 L g⁻¹ for As(v), and 2.71 and 2.68 L g⁻¹ for PO₄³⁻, respectively. Greater decreases were observed in the presence of SO₄²⁻ and HCO₃⁻, while SiO₃²⁻ caused a more substantial reduction in both coefficients. At 500 mg L⁻¹, SO₄²⁻ decreased K_d to 1.25 L g⁻¹ for As(v) and 1.52 L g⁻¹ for PO₄³⁻, whereas SiO₃²⁻ reduced the corresponding values to 0.98 and 1.18 L g⁻¹. Despite these decreases, the K_d(P)/K_d(As) ratio remained close to 1.21 across the investigated conditions, ranging from 1.20 to 1.22. This consistency indicates that the relative affinity of the ACF nanocomposite toward the two oxyanions was comparatively stable despite changes in the competing-ion environment.

Table 8. Selectivity coefficients (K_d) for As(v) and PO₄³⁻ in the presence of coexisting ions.

Coexisting ion	Concentration (mg L−1)	K d (As) (L g−1)	K d (P) (L g−1)	Selectivity ratio (Kd(P)/Kd(As))
None (control)	0	2.29	2.76	1.21
Cl−	100	2.25	2.71	1.20
NO3−	100	2.22	2.68	1.21
SO42−	100	1.85	2.24	1.21
HCO3−	100	1.72	2.08	1.21
SiO32−	100	1.42	1.72	1.21
SO42−	500	1.25	1.52	1.22
SiO32−	500	0.98	1.18	1.20

The influence of the concentration basis used to express competing-ion levels was examined by comparing equal-mass and equal-molar conditions Table 9. At 100 mg L⁻¹, Cl⁻ and NO₃⁻ produced adsorption capacities of 94.3 and 92.6 mg g⁻¹, respectively, under the equal-mass condition, compared with 93.8 and 92.1 mg g⁻¹ under the equal-molar condition. The corresponding differences were only 0.5%. For SO₄²⁻ and HCO₃⁻, the differences increased modestly to 2.1% and 2.0%, respectively, while SiO₃²⁻ exhibited the largest difference of 3.7%. Overall, the differences remained below 4%, indicating that the observed competitive effects were relatively insensitive to whether the competing-ion concentration was expressed on an equal-mass or equal-molar basis within the investigated conditions.

Table 9. Effect of equal-molar vs. equal-mass coexisting ions on adsorption capacity.

Ion	Equal Mass (100 mg L−1)	Equal molar (1.0 mM)	Difference (%)
Cl−	94.3 mg g−1	93.8 mg g−1	0.5%
NO3−	92.6 mg g−1	92.1 mg g−1	0.5%
SO42−	87.2 mg g−1	85.4 mg g−1	2.1%
HCO3−	81.4 mg g−1	79.8 mg g−1	2.0%
SiO32−	76.3 mg g−1	73.5 mg g−1	3.7%

The contribution of ionic strength to the adsorption behavior was subsequently evaluated by increasing the NaCl concentration from 0.01 to 0.50 M Table 10. The As(v) adsorption capacity decreased from 114.5 to 105.6 mg g⁻¹, corresponding to a decrease of approximately 7.8%, whereas the PO₄³⁻ capacity decreased from 138.2 to 128.4 mg g⁻¹, corresponding to a decrease of approximately 7.1%. The relatively moderate influence of ionic strength suggests that nonspecific electrostatic interactions are not the sole contributors to adsorption and that specific surface interactions remain important under elevated electrolyte concentrations. The observed decrease, however, also indicates that ionic-strength-dependent effects cannot be completely excluded.

Table 10. Effect of ionic strength on adsorption capacity.

Ionic strength (M NaCl)	As(v) qe (mg g−1)	PO43−qe (mg g−1)
0.01	114.5	138.2
0.05	112.8	136.5
0.10	110.2	133.8
0.50	105.6	128.4

Collectively, the competitive-ion experiments and quantitative analyses demonstrate that the adsorption performance of the ACF nanocomposite is relatively resistant to common background electrolytes, whereas multivalent oxyanions exert more pronounced competitive effects. The limited response to Cl⁻ and NO₃⁻, together with the measurable but moderate influence of ionic strength, supports a contribution from specific surface interactions in the adsorption process. At the same time, the stronger inhibition produced by SO₄²⁻ and SiO₃²⁻ highlights the importance of considering specific oxyanion–surface interactions when evaluating the material under realistic multicomponent water conditions. The relatively stable K_d(P)/K_d(As) ratio further indicates that the comparative affinity of the ACF nanocomposite toward PO₄³⁻ and As(v) is maintained across the investigated competitive environments.

3.4. Regeneration and reusability assessment

Performance of the ACF nanocomposite in regeneration and reusability was tested through six consecutive adsorption–desorption cycles as shown in Fig. 21, and the results revealed that the ACF nanocomposite had great regeneration and reusability performance in terms of wastewater treatment. The material was then used to retain capacity using a desorption eluent of 0.5 M NaOH and 0.1 M NaCl as the desorption eluent, and retention capacity was well above 90% of both As(v) and PO₄³⁻, with arsenic retention ending at about 95%. This recyclability is due to the synergistic stability of the alginate-chitosan framework and the chemically immobilized magnetite core which prevents active site leaching and allows easy magnetic recovery with a mass recovery of >99.5 of the entire mass as a result of its high saturation magnetization of 26.1 emu g⁻¹. The alkaline desorption reaction is efficient in order to eliminate bound oxyanions by electrostatic repulsion and competition with the hydroxyl ion, proving that interactions with the amino and hydroxyl sites observed in FTIR and XPS are reversible. Although smaller capacity reductions can be explained by incomplete desorption of strongly interacting species and mechanical erosion, the regular nature of the performance with more than 90% capacity after five cycles proves the validity of the ACF nanocomposite as a strong and sustainable adsorbent that can be used repeatedly in the industry.

Fig. 21. Regeneration and reusability performance over six cycles for As(v) and PO₄³⁻. Red dashed line indicates 90% retention threshold. Desorption: 0.5 M NaOH + 0.1 M NaCl wash. Error bars: SD (n = 3).

3.5. Mechanistic modeling studies

3.5.1. Kinetic modeling

In order to elucidate the rate-controlling characteristics of oxyanion adsorption onto the ACF nanocomposite, three commonly used kinetic models were employed to interpret the experimental data: the pseudo-first-order (PFO) model Fig. 22A, the pseudo-second-order (PSO) model Fig. 22B, and the intraparticle diffusion (IPD) model Fig. 22C. The models were fitted to the replicate-averaged experimental data using nonlinear regression, and the fitted parameters were reported as mean ± standard deviation based on uncertainty propagation analysis.

Fig. 22. Comparative adsorption kinetics modeling over a 240 minute interval; (A) the PFO model shows poor linearity and underestimates q_e, indicating that physisorption is not the rate-limiting step; (B) the PSO model shows excellent linearity with R² > 0.998 and q_e,calc matching experimental values, confirming chemisorption as the rate-limiting step; (C) the IPD model shows non-zero intercepts and multi-linearity, indicating that intraparticle diffusion is not the sole rate-limiting step; film diffusion and chemisorption also contribute.

The kinetic results demonstrated that the pseudo-second-order (PSO) model provided a substantially better fit to the experimental adsorption data than the pseudo-first-order (PFO) model for both As(v) and PO₄³⁻. The PFO model yielded (R²) values of 0.951 and 0.941 for As(v) and PO₄³⁻, respectively, whereas the PSO model exhibited considerably higher (R²) values of 0.999 and 0.998, respectively Table 11. Moreover, the equilibrium adsorption capacities calculated using the PSO model (q_e, calc) were 115.2 and 133.2 mg g⁻¹ for As(v) and PO₄³⁻, respectively, and were in close agreement with the corresponding experimental values. These results indicate that the PSO model provides a more appropriate description of the adsorption kinetics of the ACF nanocomposite under the investigated conditions.

Table 11. Kinetic model parameters for As(v) and PO₄³⁻ adsorption onto the ACF nanocomposite.

Model	Parameter	As(v)	PO43−
Pseudo-first-order	k 1 (min−1)	0.038	0.042
	q e calc (mg g−1)	98.3	88.5
	R 2	0.951	0.941
Pseudo-second-order	k 2 (g mg−1 min−1)	0.0021	0.0018
	q e (mg g−1)	115.2	133.2
	R 2	0.999	0.998
Intraparticle diffusion	k id (mg g−1 min−1/2)	3.78	4.12
	C	45.8	31.6
	R 2	0.964	0.962

The comparatively lower (R²) values obtained from the PFO model, together with its underestimation of the equilibrium adsorption capacities (q_e, calc) = 98.3 and 88.5 mg g⁻¹ for As(v) and PO₄³⁻, respectively), indicate that the PFO model does not adequately describe the overall adsorption kinetics. In contrast, the close agreement between the PSO-calculated and experimental equilibrium capacities, together with the high (R²) values, supports the involvement of surface chemical interactions in the adsorption process. However, the kinetic fitting results alone should not be considered conclusive evidence of chemisorption; rather, they should be interpreted in conjunction with the FTIR, XPS, and thermodynamic analyses discussed subsequently.

The intraparticle diffusion (IPD) model provided moderate fits, with (R²) values of 0.964 for As(v) and 0.962 for PO₄³⁻Table 11. The corresponding non-zero intercepts (C) of 45.8 and 31.6 mg g⁻¹, respectively, indicate that intraparticle diffusion was not the sole rate-controlling step. Instead, the adsorption process appears to involve multiple mass-transfer stages, including external film diffusion followed by diffusion within the porous structure and subsequent interaction with the available adsorption sites. The intraparticle diffusion rate constants (k_id) were 3.78 and 4.12 mg g⁻¹ min^−0.5 for As(v) and PO₄³⁻, respectively. The positive intercepts further suggest a significant contribution of boundary-layer or film diffusion, particularly during the initial rapid uptake stage.⁵⁸

Both As(v) and PO₄³⁻ exhibited a rapid initial adsorption stage followed by a gradual approach to equilibrium, which was reached within approximately 240 min. Overall, the kinetic results indicate that adsorption proceeds through a multistep process involving external mass transfer, intraparticle diffusion, and interactions with surface-active sites, with the PSO model providing the best overall representation of the experimental kinetic behavior.⁶⁴

3.5.2. Isotherm modeling

Adsorption isotherms can give data regarding the distribution of adsorbates in both the solid and liquid phases at equilibrium. Isotherms were plotted at different initial concentrations (10–500 mg L⁻¹), dose 1.0 g L⁻¹, time 240 min, pH 6.5, and temperature 25 °C. The adsorption isotherms of As(v) and PO₄³⁻ on ACF nanocomposite were studied to determine the maximum adsorption capacity and the mode of interaction between the oxyanions and the adsorbent surface (Fig. 22). The experimental results show that the adsorption capacity (q_e) rises exponentially at low equilibrium concentrations (C_e) and then curves off, which is typical of high-affinity adsorption processes.

The experimental data were found to be within close range of the Langmuir model Fig. 23(A), which indicates that the sequestration process takes place in the monolayer adsorption style at a constant number of energetically comparable active sites. The Langmuir model gave the highest R² values (0.998 for As(v) and 0.999 for PO₄³⁻) and the lowest RMSE, confirming that:

Fig. 23. Isotherm modeling of As(v) and PO₄³⁻ adsorption onto ACF nanocomposite, (A): the Langmuir model shows excellent linearity (R² > 0.998) with q_max values of 114.5 mg g⁻¹ (As) and 138.2 mg g⁻¹ (P), confirming monolayer adsorption on homogeneous sites, (B) the Freundlich model shows moderate fit (R² ∼0.98) with 1/n < 1, indicating favorable adsorption with some surface heterogeneity; (C) the Temkin model shows moderate fit (R² ∼0.94) with b values in the chemisorption range, confirming chemical bonding rather than physisorption.

1. Adsorption occurs via monolayer coverage.

2. The adsorption binding sites are identical energetically.

3. The adsorbates do not interact laterally.

4. Monolayer coverage occupies each site with a single adsorbate molecule.

The maximum Langmuir adsorption capacities were 114.5 mg g⁻¹ for As(v) and 138.2 mg g⁻¹ for PO₄³⁻. These strengths are higher than most reported adsorbents, which can be explained by the synergistic effect of high surface area (182.4 m² g⁻¹), mesoporous structure, and the synergistic effect of the alginate-chitosan matrix, which provides high concentrations of amines and carboxyl groups, and the magnetite (Fe₃O₄) nanoparticles, which offer selective coordination sites where the oxyanions can bind. The affinity slightly increased towards phosphate, which could be because of its lesser ionic radius or more intense hydrogen bonding of the protonated amine groups on the chitosan at pH 6.5.⁶⁵

The Freundlich model Fig. 23(B) showed moderate fit (R² = 0.983 for As(v) and 0.980 for PO₄³⁻) with 1/n values of 0.42 and 0.38, respectively. The 1/n values between 0 and 1 indicate favorable adsorption, but the lower R² values compared to Langmuir indicate that the surface is not purely heterogeneous.⁶⁶

The Temkin model Fig. 23(C) including indirect adsorbent–adsorbate interactions demonstrated moderate fit (R² = 0.942 for As(v) and 0.938 for PO₄³⁻) having binding constants (A) of 2.15 L mg⁻¹ and 1.98 L mg⁻¹ and heats of adsorption (b) of 542.8 J mol⁻¹ and 518.3 J mol⁻¹. The b values are in the chemisorption range (20–80 kJ mol⁻¹), confirming that the adsorption involves chemical bonding rather than physical forces.⁶⁷

Apparently, this means that the Langmuir model is the most appropriate to be employed to describe As(v) and PO₄³⁻ adsorption onto ACF nanocomposite, with the qmax related to PO₄³⁻ being larger (138.2 mg g⁻¹) compared to As(v) (114.5 mg g⁻¹), which implies a stronger electrostatic attraction and inner-sphere complexation of the phosphate anions with the surface Fe–OH and protonated amine groups. The largest R², the lowest RMSE, and the theoretical Langmuir fit is excellent, indicating that: (1) adsorption occurs via monolayer coverage, (2) the adsorption binding sites are identical energetically, (3) the adsorbates do not interact laterally, and (4) monolayer coverage occupies each site with a single adsorbate molecule. Such assumptions are consistent with the inner-sphere complexation mechanism because arsenate is directly coordinated with particular surface sites (Fe–OH, –NH₃⁺, –COO⁻) and is not accreted in more than one layer.⁶⁸

3.5.3. Thermodynamic interpretation

To measure the nature and feasibility of the sequestration process in terms of energy, the thermodynamic analysis of As(v) and PO₄³⁻adsorption onto the ACF nanocomposite was performed to investigate the thermodynamic feasibility of the process. The enthalpy (ΔH°) and entropy (ΔS°) can be determined precisely by a linear relationship between (ln K_dversus 1/T) (R² = 0.998 with arsenic and R² = 0.8873 with PO₄³⁻, as shown in Fig. 24A and B). The positive value of ΔH° of adsorption process of AS(v) of + 42.85 ± 1.9 kJ mol⁻¹, + 48.92 ± 1.1 of PO₄³⁻ as in Table 12 confirms the fact that the adsorption process of AS(v) is endothermic. The positive positive enthalpy shows that it is an endothermic reaction that needs the input of energy as is typical of chemisorption where covalent or coordination bonds are formed between adsorbate and adsorbent. The size (20–40 kJ mol⁻¹ range) falls between the range of physisorption (<20 kJ mol⁻¹) and the range of strong chemisorption (>80 kJ mol⁻¹), showing that this process is not necessarily dominated by physical attraction, but the formation of special chemical bonds namely inner-sphere coordination and ligand exchange. This endotherm character suggests that increased temperatures are the fine catalysts to dehydrate the oxyanions and allow them to interact with the amine and hydroxyl functional groups of the chitosan-magnetite matrix.⁶⁹ Moreover, the positive ΔS° (+206.4 J mol⁻¹ K⁻¹) for AS(v), and + 236.0 J mol⁻¹ K⁻¹ for PO₄³⁻ indicates an augmentation in the randomness at the solid–liquid interface, presumably through the eruption of displaced water molecules into the bulk solution on the binding of oxyanions. The positive entropy indicates greater randomness of the system, which is mainly due to the displacement of various water molecules of both the hydrated arsenate ion and solvated surface sites into bulk solution, by passing the entropy loss to limiting the mobility of the arsenate. More importantly, the Gibbs free energy diagrams Fig. 24C and D indicate that at all the tested temperatures, in the range of −18.74 to −22.89 kJ mol⁻¹ in the case of As(v) and even more negative values −21.36 to −26.08 with PO₄³⁻,that ΔG° is negative, that is, the adsorption in question is spontaneously negative. The decreasing pattern of ΔG° with rise in temperature confirms that the reaction order and intensity of the reaction are stronger at elevated thermal conditions which confirms the strong thermodynamic capability of the ACF nanocomposite in purifying affected water.⁷⁰

Fig. 24. Thermodynamic analysis: (A and B) Van't Hoff plot for enthalpy and entropy determination, (C and D) temperature-dependent Gibbs free energy showing spontaneous adsorption of As(v) and PO₄³⁻across all temperatures. Linear fit in (a): R² = 0.998.

Table 12. Thermodynamic parameters for As(v) and PO₄³⁻adsorption onto ACF composite.

Ion	T (K)	ΔG° (kJ mol−1)	ΔH° (kJ mol−1)	ΔS° (J mol−1 K−1)	R 2 (Van't Hoff)
As(v)	298	−18.74	42.85	206.4	0.9931
	308	−20.81	42.85	206.4	0.9931
	318	−22.89	42.85	206.4	0.9931
PO43−	298	−21.36	48.92	236.0	0.9954
	308	−23.72	48.92	236.0	0.9954
	318	−26.08	48.92	236.0	0.9954

3.5.4. Direct mechanistic evidence from XPS, and FTIR difference spectra

To provide direct evidence for the proposed adsorption mechanism, XPS, FTIR difference spectra, and surface complexation modeling (SCM) were performed. The corresponding XPS, FTIR difference spectra, and surface complexation modeling results are presented in Fig. 25, with panels R1, R2, and R3, respectively, providing complementary evidence for the proposed adsorption mechanism.

Fig. 25. Direct mechanistic evidence for As(v) and PO₄³⁻ adsorption onto the ACF nanocomposite based on XPS analysis of As 3d spectrum, and P 2p spectrum (R1); Fe 2p and O 1s XPS deconvolution spectra of ACF before and after As(v) and PO₄³⁻ adsorption (R2); FTIR difference spectra (R3).

XPS analysis after adsorption:

The As 3d spectrum showed a doublet at 45.2/46.1 eV (As 3d_5/2/3/2) with a shoulder at 44.5 eV, indicating the formation of inner-sphere Fe–O–As complexes.

The P 2p spectrum showed a doublet at 133.2/134.1 eV with a shoulder at 132.5 eV, indicating the formation of inner-sphere Fe–O–P complexes.

The Fe 2p peaks shifted to higher binding energies (+0.3–0.4 eV) after adsorption, confirming coordination with arsenate/phosphate.

The O 1s deconvolution showed a decrease in the Fe–OH peak and a shift in the Fe–O peak, confirming ligand exchange and Fe–O–As/P bond formation.

FTIR difference spectra:

New peaks appeared at 820–850 cm⁻¹ (As–O stretch) and 1150 cm⁻¹ (P–O stretch), confirming As(v) and PO₄³⁻ adsorption.

The Fe–O peak shifted from 580 to 570 cm⁻¹, confirming Fe–O–As/P bond formation.

The carboxylate peaks shifted, confirming complexation with Fe³⁺.

These direct evidence collectively confirm that the adsorption of As(v) and PO₄³⁻ onto the ACF nanocomposite occurs via inner-sphere complexation (Fe–O–As/P bonds), electrostatic attraction (protonated amines), and hydrogen bonding (carboxylate and hydroxyl groups).

3.6. Proposed mechanism

The adsorption of As(v) and PO₄³⁻ onto the alginate-chitosan-Fe₃O₄ (ACF) nanocomposites is proposed to follow a multifaceted synergy of physical and chemical interactions because of the various functional groups of the hybrid matrix as illustrated in Fig. 26. One of the leading forces is the electrostatic attraction in which the positively charged protonated amine end groups (–NH₃⁺) of the chitosan backbone and the hydroxylated surfaces of the iron are attracted by negatively charged oxyanions (AsO₄³⁻) and (PO₄³⁻) in slightly acidic operating conditions. In addition to plain attraction, the mechanism changes into chemisorption due to the appearance of inner-sphere complexes and coordination bonds (Fe–O–As and Fe–O–P) the oxyanions directly coordinate with the iron centers of the magnetite nanoparticles through ligand exchange with the surface hydroxyl groups.³⁴ Moreover, hydrogen bonding is an important process in preserving the adsorbed species because the oxygen atoms of the arsenate and phosphate ions are coupled to the owing –OH^- and COOH groups that are abundant in the alginate-chitosan net. These interactions are further supported by the complex of the 3D porous structure of the bio-polymer matrix because it offers a high surface area and protection of the magnetic core that, in turn, enables effective magnetic separation of the treated water. The combination of electrostatic forces, covalent coordination, and a wide range of networks of hydrogen bonds guarantees the high capacity and high stability of the ACF nanocomposite to act as an effective multi-pollutants remediation.¹⁸ The co-precipitation process is synergistic in nature to make certain that the Fe₃O₄ nanoparticles are not just encapsulated but are chemically bonded to the polymer chains via carboxylate-iron bridges. Such an arrangement eliminates the leaching of the magnetic core in the competitive ion displacement process. The high charge density and the steric capacity of As(v) and phosphate to fit with the interstitial voids of the ionic cross-linked network in complex aqueous matrices explain the preference of these tetrahedral species. The dipolarity of the Fe₃O₄ core permits the imposition of a magnetic gradient that is capable of focusing these contaminants within the porous skeleton, and redistributing the concentration of the magnetic flux locally in regions around the reactive sites and boosting the overall kinetics of the sequestering reaction.⁷¹

Fig. 26. Schematic illustration of the proposed adsorption mechanism of for As(v) and PO₄³⁻adsorption onto ACF composite.

At the same time, secondary stabilization is achieved by the organic structure due to hydrogen bonding and physical chelation. The hydroxyl groups that are positioned on both ends of the alginate backbone create strong hydrogen bonds with the oxygen atom of the As(v) and orthophosphate ions. In addition, Fourier Transform Infrared (FTIR) spectroscopy of the ACF nanocomposite after adsorption of As(v) and PO₄³⁻Fig. 27 give convincing data of the chemical interactions and use of the functional groups in the sequestration process. It can be seen that there are great spectral shifts and intensity variations in the characteristic bands of the biopolymer matrix especially the broad peak of 3300–3400 cm⁻¹, which is comprised by overlapping –OH and –NH₂ frequencies. The observable widening and decrease in intensity of this band following loading with the two oxyanions indicate that the formation of large numbers of hydrogen-bonds and possibly coordination complex of the pollutants and the amino/hydroxyl groups of the chitosan and alginate strands occurred. Moreover, the typical carboxylate (–COO⁻)-stretches around 1600 cm⁻¹ and 1415 cm⁻¹ display different shifts, which demonstrates that the alginate structure actively involves in the ion binding process via electrostatic attraction and inner-sphere complexation.⁷² The fingerprint area lower than 1000 cm⁻¹ was the most crucial area where the evidence of successful adsorption can be found. In the case of As(v)-loaded sample a new or enhanced shoulder at 820–850 cm⁻¹ can be related to the As–O stretch vibration and the arsenate species are present on the adsorbent. On the same note, the PO₄³⁻ loaded spectrum will show a strong intensified band in the 1150 cm⁻¹ region, which is the characteristic P–O stretch in groups of phosphates. The stability of the Fe–O vibration at around 580 cm⁻¹, is an indication that the magnetite core is structurally sound, whereas the middle-level shift indicates that the surface of iron oxide is one of the main loci where a ligand exchange and development of Fe–O–As or Fe–O–P coordination bonds may occur.⁷³ All these spectral changes confirm that the ACF nanocomposite employs a synergistic chemisorption mechanism, and it makes use of its multiple functional groups to attain high-capacity water remediation in polluted water.

Fig. 27. FTIR of alginate-chitosan-magnetite (ACF) nanocomposite after adsorption proceass of AS(v), and PO₄³⁻.

3.7. Comparison with previous studies

The comparison of the produced magnetically responsive alginate-chitosan-Fe3O4 nanocomposite with the current literature shows that the hybrid system has a high level of performance, especially in its dual-functional capture efficiency and structural stability. Although most of the conventional adsorbents are either cation or anionic in nature, the extensive synergistic combination of amine and carboxyl species on chitosan and alginate respectively, with an in situ formed magnetite core, enables simultaneous removal of pentavalent arsenic and orthophosphate. Table 13 offers a detailed comparison of the adsorption capacities and working conditions of different engineered materials of the recent studies. In this paper the alginate-chitosan-Fe₃O₄ composite exhibited the highest adsorption capacity of 114.5 mg g⁻¹ and 138.2 mg g⁻¹ in As(v) and orthophosphate respectively at optimum pH of 6.5 and at 298 K respectively. This performance is better than what has been reported with bimetallic Zr/Fe0-modified chitosan hydrogels composites or Magnetic composite microparticles based on Fe₃O₄, TiO₂ and chitosan/quaternary ammonium salt chitosan derivatives.

Table 13. Comparison of alginate-chitosan-Fe₃O₄ performance with other reported materials.

Adsorbent	As(v) adsorption capacity (mg g−1)	PO43− adsorption capacity (mg g−1)	Contact time (min)	pH@maximum adsorption	Ref.
Magnesium-alginate/chitosan modified biochar microspheres	—	46.56	30	7	74
Polyvinyl alcohol/metal ions sodium alginate hydrogel beads	—	34.2	240	4	75
Bimetallic Zr/Fe-modified chitosan hydrogels	—	126.50	240	7	14
TiO2@chitosan/alginate nanocomposite	202.27	—	100	5	70
Core–shell/bead-like ethylenediamine-functionalized Al-pillared montmorillonite/calcium alginate	61.94	—	150	4	76
Magnetic composite microparticles based on Fe3O4, TiO2 and chitosan/quaternary ammonium salt chitosan derivatives	34.61	—	120	7	23
Alginate-chitosan-Fe3O4 nanocomposite	114.5	138.2	240	6.5	This work

The ability to develop a competitive advantage of the existing nano-architecture is explained by a number of decisive factors. First, in situ co-precipitation the method allows the more uniform distribution of Fe₃O₄ nanoparticles in the biopolymeric matrix than simple physical blending and avoids the leaching of nanoparticles and retains a large portion of surface area. Secondly, the ionic cross-linking of the alginate and chitosan chains form a strong three dimensional network resistant to osmotic swelling and hence maintaining the structural integrity of the sites through several regeneration cycles. Lastly, the paramagnetic property of the composite allows quick isolation of the compound under complex aqueous matrices with an external magnetic field with an efficiency of more than 98%. This removes the requirement of any energy demanding filtration or centrifugation which is a major setback in most of the non-magnetic high capacity adsorbents found in Table 13 The concurrent absorption of arsenic and phosphate species in the absence of strong inhibitory effects of competitive ions also helps to emphasize the usefulness of the material in the context of multi-pollutant groundwater and agricultural runoff treatment.

4. Conclusion

Magnetically responsive nano-architectures of alginate-chitosan-Fe₃O₄ were successfully synthesized using a simplified in situ co-precipitation approach and sequential ionic/covalent cross-linking to form an interpenetrating double-network structure. The IPN formation was confirmed through FTIR, TGA/DSC, swelling/solubility testing, mechanical testing, and solid-state NMR. Thorough characterization established consistency of distribution of Fe₃O₄ nanoparticles. ACF nanocomposite was found to have a high specific surface area of 182.4 m² g⁻¹ and a mesoporous structure with a mean pore size of 8.2 nm through Brunauer–Emmett–Teller analysis. The saturation magnetization of Vibrating Sample Magnetometer was 26.1 emu g⁻¹ with negligible coercivity and remanence, confirming superparamagnetic behavior. This high magnetization enabled rapid magnetic separation of the ACF nanocomposite from aqueous suspension within 30 seconds using an external neodymium magnet (0.5 T), as determined by time-resolved turbidity and ICP-OES measurements. The rapid separation is attributed to the well-dispersed magnetite nanoparticles and the thin biopolymer coating that does not excessively shield the magnetic core. The highest rate of sequestration was observed at pH 6.5, where the composite had the highest adsorption capacities of 114.5 mg g⁻¹ towards As(v) and 138.2 mg g⁻¹ towards orthophosphate. The pseudo-second order model was closely followed by kinetics, which indicated that chemisorption happened through surface complexation and electrostatic attraction of the protonated amino groups of chitosan and the oxoanions. The material was found to be very selective in aqueous matrices, and its efficiency was more than 85% even with the presence of other competitive chloride and sulfate ions. In addition to this the adsorbent was paramagnetic and this aspect enabled rapid phase separation and the regeneration studies revealed that the adsorbent maintained >90% of its original capacity after five successive cycles and >85% after six cycles. These results highlight the promise of such nano-architectures as high-performance and sustainable candidates to large-scale water cleanup. The next directions of future research include the determination of these sorbents in continuous-flow column mode and also the stability of the biopolymer shell over time under different hydrodynamic shear forces.

Author contributions

Eman Fayad: resources, formal analysis, data curation, conceptualization; Maha Alsunbul: project administration, funding acquisition, software, validation; Mohammad Albayari and Abeer Adaileh; resources, methodology, investigation, formal analysis, data curation, conceptualization; Hazem I. Bendary: investigation, software; Eslam A. Mohamed and Mahmoud F. Mubarak: resources, methodology, investigation, formal analysis, data curation, conceptualization, software, validation, visualization, writing – original draft, writing – review & editing.

Conflicts of interest

The 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.

Funding

The research was funded by Princess Nourah bint Abdulrahman University Researchers Supporting Project number (PNURSP2026R736), Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia.

Data availability

The authors declare that the data supporting the findings of this study are available within the paper. Should any raw data files be needed in another format, they are available from the corresponding author upon reasonable request.