Lime-Juice-Assisted Synthesis of Magnetically Recoverable CoFe₂O₄/Chitosan Nanocomposites for Visible-Light Removal of Methylene Blue

1. Introduction

Synthetic dyes released by textile, leather, pharmaceutical, and related industries remain a persistent challenge for aquatic environments because many dye molecules are chemically stable, intensely colored, and poorly biodegradable [1,2,3]. Methylene blue (MB), a cationic phenothiazine dye, is frequently used as a model contaminant in photocatalytic studies owing to its strong visible absorption and resistance to conventional biological treatment [1,3]. Even at relatively low concentrations, dye-containing effluents can impede light penetration and oxygen transfer, disrupt aquatic photosynthesis, and create potential human-health concerns[1,2,3,4,5]. Accordingly, treatment technologies must combine high removal efficiency with low chemical demand, operational simplicity, and catalyst reusability.

Semiconductor photocatalysis is an attractive advanced oxidation process because absorbed photons generate electron–hole pairs that can initiate the formation of reactive oxygen species, including hydroxyl radicals (•OH) and superoxide species (•O₂⁻), under comparatively mild conditions [2,5,6,7,8,9]. Practical performance, however, is often limited by rapid charge recombination, insufficient visible-light utilization, nanoparticle aggregation, and difficult post-treatment separation. Material design must therefore balance optical activity, interfacial charge transfer, accessible adsorption sites, and recoverability rather than optimize any single property in isolation.

Spinel ferrites (MFe₂O₄; M = Co, Zn, Ni, Mn, Mg, or Cu) are promising photocatalytic platforms because their electronic structures support visible-light response and their magnetic properties enable separation from treated water [10,11]. CoFe₂O₄ is particularly attractive because of its chemical stability, high magnetocrystalline anisotropy, coercivity, and saturation magnetization [12,13,14,15]. Nevertheless, bare CoFe₂O₄ nanoparticles tend to aggregate through magnetic and high-surface-energy interactions, reducing the number of accessible sites and complicating reproducible photocatalytic operation.

CoFe₂O₄ has been prepared by co-precipitation, sol–gel, hydrothermal, vapor-phase, laser-ablation, and mechanochemical methods [16,17,18,19,20,21]. Several of these routes require synthetic complexing agents, hazardous solvents, strong reducing agents, or energy-intensive processing [22]. Bio-assisted synthesis partially addresses these limitations by using plant-derived metabolites as complexing, capping, and stabilizing species [14,22,23,24,25,26,27,28]. Citrus juices contain citric and ascorbic acids, reducing sugars, flavonoids, polyphenols, and other oxygen-containing compounds that can coordinate metal ions, regulate nucleation, and suppress uncontrolled particle growth [29,30,31,32]. Their successful application to ZnO, CuO, MgO, TiO₂, and related nanomaterials supports the use of lime juice as a low-cost synthesis medium [29,30,33,34].

Metal oxide–polymer composites have been widely explored to improve semiconductor photocatalysis through enhanced light harvesting, reduced charge recombination, and better catalytic performance [35]. Chitosan (CS), a renewable polysaccharide obtained by deacetylation of chitin, offers complementary functionality through abundant amino and hydroxyl groups [36,37,38]. These groups can bind cationic and molecular pollutants through electrostatic interactions, hydrogen bonding, complexation, and physical adsorption [37,39,40]. As a polymeric support, chitosan can also limit ferrite aggregation and improve nanoparticle dispersion [41,42]. The resulting composite may therefore couple adsorption-assisted concentration of MB near the catalyst surface with more effective utilization of photogenerated charge carriers. Importantly, such synergy can improve apparent removal even when polymer coverage decreases the measured BET surface area.

Despite growing interest in ferrite–biopolymer composites, the relationships among citrus-assisted synthesis, CoFe₂O₄/CS interfacial chemistry, optical response, magnetic recovery, and adsorption-coupled photocatalysis remain insufficiently resolved. A rigorous assessment should therefore integrate structural, surface, optical, textural, and magnetic characterization with kinetic, operational-parameter, recycling, and reactive-species analyses.

In this work, CoFe₂O₄ nanoparticles were synthesized using lime juice as a natural complexing and stabilizing medium and subsequently incorporated into chitosan. The effects of calcination temperature on ferrite formation were investigated, and the resulting CoFe₂O₄/CS nanocomposite was characterized using structural, morphological, optical, surface, textural, and magnetic techniques. Its photocatalytic MB removal performance was evaluated under different irradiation times, pH values, dye concentrations, and catalyst dosages, together with reusability and scavenger tests

2. Materials and Methods

2.1. Materials

All reagents were of analytical grade and were used without further purification. Cobalt(II) nitrate hexahydrate [Co(NO₃)₂·6H₂O, 99%], iron(III) nitrate nonahydrate [Fe(NO₃)₃·9H₂O, 99%], medium-molecular-weight chitosan (degree of deacetylation, 75–85%), and methylene blue (97%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Sodium hydroxide (NaOH, 98%), acetic acid (CH₃COOH, 99%), and hydrochloric acid (HCl, 37%; density, 1.19 g mL⁻¹) were obtained from Aladdin Reagent Co., Ltd. (Shanghai, China). Deionized water was used throughout.

2.2. Preparation of Lime-Juice Extract

Fresh limes obtained in Nghe An Province, Vietnam, were washed thoroughly with deionized water and mechanically pressed. The juice was centrifuged at 13,000 rpm for 10 min and filtered, and the clear supernatant was stored at 4 °C until use.

2.3. Synthesis of CoFe₂O₄ Nanoparticles and CoFe₂O₄/CS Nanocomposite

CoFe₂O₄ nanoparticles were synthesized by dissolving 0.005 mol Co(NO₃)₂·6H₂O and 0.010 mol Fe(NO₃)₃·9H₂O in 50 mL of deionized water under stirring, thereby maintaining the stoichiometric Co:Fe ratio of 1:2. Lime-juice extract (20 mL) was added at room temperature, and the mixture was stirred at 80 °C until a gel formed. The gel was dried at 80 °C for 4 h, heated to the selected calcination temperature at 5 °C min⁻¹, held for 2 h, and cooled naturally. Samples calcined at 300, 400, 500, and 700 °C were used to evaluate phase development; the material calcined at 500 °C was used for composite preparation and subsequent photocatalytic tests.

For nanocomposite fabrication, 1.0 g of chitosan was dissolved in 100 mL of 1% (v/v) aqueous acetic acid. CoFe₂O₄ nanoparticles (0.30 g) were dispersed in the chitosan solution by magnetic stirring followed by sonication for 30 min. The pH was adjusted gradually to approximately 11 with 2 M NaOH, and the dispersion was sonicated for another 30 min and heated at 80 °C for 3 h. The solid was collected by filtration, washed repeatedly with deionized water to neutral pH, and dried at 60 °C for 5 h.

2.4. Characterization

Crystalline phases were analyzed by X-ray diffraction (XRD; D8 Advance, Bruker, Germany) using Cu Kα radiation (λ = 0.15406 nm) over 2θ = 10–70°. Crystallite size was estimated from the (311) reflection using the Scherrer equation, and lattice parameters, unit-cell volume, and X-ray density were calculated from the diffraction data.

Surface functional groups and ferrite–polymer interactions were examined by Fourier-transform infrared spectroscopy (FTIR; Thermo-Nicolet Nexus 670). Morphology and particle distribution were observed by field-emission scanning electron microscopy (FESEM; Hitachi S-4800) and transmission electron microscopy (TEM; JEOL JEM-1400). Elemental composition was assessed by energy-dispersive X-ray spectroscopy (EDX; HORIBA 7593-H) coupled to the SEM.

Surface elemental composition and oxidation states were investigated by X-ray photoelectron spectroscopy (XPS; ESCALab 250, Thermo VG, UK). Nitrogen adsorption–desorption measurements were performed on a Tristar-3000 analyzer, and the specific surface area and pore-size distribution were determined using the BET and BJH methods, respectively.

Diffuse-reflectance spectra were recorded with a Cary 5000 UV–Vis–NIR spectrometer. Reflectance data were transformed using the Kubelka–Munk function, and optical band gaps were estimated from Tauc plots assuming a direct allowed transition. Photoluminescence spectra were measured with a Fluorolog-3 spectrofluorometer (FL3C-22, HORIBA Scientific) to compare the relative radiative recombination behavior of CoFe₂O₄ and CoFe₂O₄/CS.

Magnetic hysteresis loops were recorded at room temperature using a Lake Shore 7404 vibrating sample magnetometer (VSM). Saturation magnetization (Mₛ), remanent magnetization (Mᵣ), and coercivity (Hc) were extracted from the hysteresis curves. Magnetic separation from water was assessed qualitatively using an external permanent magnet.

2.5. Photocatalytic Experiments

Photocatalytic tests were performed in a glass reactor containing 50 mL of MB solution and the required amount of catalyst. A 30 W visible-light LED was positioned 20 cm above the liquid surface. Before irradiation, each suspension was stirred in the dark for 60 min to establish adsorption–desorption equilibrium. Aliquots were withdrawn at 30 min intervals, the catalyst was separated, and the residual MB concentration was determined by UV–Vis spectrophotometry. Control experiments included direct photolysis and dark adsorption by CS, CoFe₂O₄, and CoFe₂O₄/CS. The MB removal efficiency (RE) was calculated as follows:

$$\mathrm{R}\mathrm{E}\%={\displaystyle \frac{C_o-C_t}{C_o}}\mathrm{x}100\mathrm{\%}$$

(1)

where C₀ and Cₜ (mg L⁻¹) are the MB concentrations at the beginning of the test and at irradiation time t, respectively.

The apparent photocatalytic kinetics were evaluated using the pseudo-first-order expression:

$$\mathrm{L}\mathrm{n}{\displaystyle \frac{C_o}{C_t}}=kt$$

(2)

where k is the pseudo-first-order rate constant (min⁻¹), and t is the irradiation time (min). The value of k was obtained from the slope of the linear plot of ln(C₀/Cₜ) versus t.

Operational parameters were examined over the following ranges: initial pH, 3–11; initial MB concentration, 10–40 mg L⁻¹; and CoFe₂O₄/CS dosage, 0.2–1.0 g L⁻¹. Unless a parameter was being varied, the remaining conditions were held constant. The point of zero charge (pH_PZC) of the nanocomposite was determined using the pH-drift method.

Reactive-species tests were performed by adding p-benzoquinone (BQ), isopropanol (IPA), ethylenediaminetetraacetic acid disodium salt dihydrate (Na₂EDTA·2H₂O, abbreviated as EDTA), or AgNO₃ as scavengers for •O₂⁻, •OH, h⁺, and e⁻, respectively. The resulting removal efficiencies were compared with a scavenger-free control under otherwise identical irradiation conditions.

Reusability was evaluated for five consecutive cycles under the optimized conditions (pH 9, 20 mg L⁻¹ MB, 0.7 g L⁻¹ catalyst, and 120 min irradiation). After each cycle, the nanocomposite was recovered with an external magnet, washed with ethanol and deionized water, vacuum-dried at 80 °C for 5 h, and reused in a freshly prepared MB solution. Phase stability was additionally assessed by comparing the XRD patterns of fresh material and a separately recovered specimen after five irradiation–recovery cycles.

All photocatalytic experiments were conducted in triplicate. Results are reported as mean ± standard deviation (SD), and error bars represent three independent measurements (n = 3).

3. Results and Discussion

3.1. Calcination-Dependent Formation of CoFe₂O₄

Figure 1a shows the XRD patterns of the ferrite precursors calcined at 300–700 °C. The sample treated at 300 °C exhibited only weak, broad features, indicating limited crystallization. At 400 °C, reflections appeared at 2θ = 30.3°, 35.6°, 43.3°, 53.7°, 57.3°, and 62.9°, corresponding to the (220), (311), (400), (422), (511), and (440) planes of cubic spinel CoFe₂O₄ (JCPDS 22-1086). Increasing the calcination temperature sharpened and intensified these reflections, consistent with progressive crystallite growth and improved long-range order.

The mean crystallite size D was estimated from the (311) reflection using the Scherrer equation [43]:

$$D={\displaystyle \frac{K\lambda }{\beta \cos \theta }}$$

(3)

where K = 0.9 is the shape factor, λ = 1.5406 Å is the Cu Kα wavelength, β is the full width at half maximum of the (311) peak (in radians), and θ is the corresponding Bragg angle.

$$a=d_{hkl}\sqrt{\left(h^2+k^2+{\mathcal{l}}^2\right)}$$

(5)

$$\rho ={\displaystyle \frac{8M}{NV}}$$

(6)

where dₕₖₗ is the interplanar spacing, h, k, and l are the Miller indices, M is the molar mass of CoFe₂O₄, Nₐ is Avogadro’s constant (6.022 × 10²³ mol⁻¹), and the factor 8 is the number of formula units per spinel unit cell.

The calculated crystallographic parameters are summarized in Table 1.

The crystallite size increased from 8.1 nm at 400 °C to 10.0 nm at 500 °C and 18.9 nm at 700 °C, reflecting thermally activated domain growth and coalescence [14,15]. Concurrently, a increased from 8.3304 to 8.3530 Å and V from 578.09 to 582.81 ų, while ρₓ decreased slightly from 5.39 to 5.35 g cm⁻³. The shift of the (220) and (311) reflections toward lower 2θ values (Figure 1b) is consistent with this small lattice expansion [44,45,46].

A calcination temperature of 500 °C was selected for subsequent work because it produced well-defined spinel CoFe₂O₄ while retaining a comparatively small crystallite size (10.0 nm). The 700 °C treatment increased crystallinity but nearly doubled the crystallite size, a change expected to reduce external surface area and potentially diminish photocatalytic accessibility. The selected temperature therefore represents a compromise between phase purity and nanoscale dimensions.

3.2. Crystal Structure of the CoFe₂O₄/CS Nanocomposite

The XRD patterns of CS, CoFe₂O₄, and CoFe₂O₄/CS are compared in Figure 2. Chitosan displayed a broad maximum near 2θ = 20°, characteristic of its semi-crystalline polysaccharide structure [47,48].

The nanocomposite retained the characteristic spinel reflections of CoFe₂O₄ together with the broad chitosan contribution. The reduced ferrite peak intensities primarily reflect dilution by the polymer matrix and partial attenuation of the diffracted signal; interfacial disorder may provide an additional contribution. No new crystalline impurity phase or substantial peak displacement was observed, demonstrating that composite formation preserved the CoFe₂O₄ spinel framework [49,50].

3.3. FTIR Analysis

Figure 3 compares the FTIR spectra of CS, CoFe₂O₄, and CoFe₂O₄/CS.

Chitosan exhibited a broad band at 3462.6 cm⁻¹ assigned to overlapping O–H and N–H stretching vibrations. The bands at 2933.5 and 2877.6 cm⁻¹ arose from asymmetric and symmetric C–H stretching, respectively [51,52,53,54].

The band at 1660.5 cm⁻¹ was associated with residual amide C=O stretching, whereas the features at 1592.1 and 1394.1 cm⁻¹ were assigned to amino-group bending (amide II region) and C–H deformation, respectively [55,56,57,58,59]. Bands at 1171.9 and 1073.4 cm⁻¹ corresponded to C–O–C stretching and coupled C–O/C–N vibrations of the glucosamine framework [55,56,57].

CoFe₂O₄ displayed broad O–H stretching and adsorbed-water bending bands at 3451.2 and 1635.3 cm⁻¹, respectively [15,41,60]. The feature at 1383.7 cm⁻¹ was attributed to residual surface-bound organic species from the bio-assisted synthesis [61]. The bands at 586.2 and 416.7 cm⁻¹ are characteristic of metal–oxygen vibrations at tetrahedral and octahedral sites of the spinel lattice, respectively [62].

Both polymer and ferrite vibrations were present in CoFe₂O₄/CS. The metal–oxygen bands shifted from 586.2 and 416.7 cm⁻¹ to 575.6 and 408.8 cm⁻¹, and the broad O–H/N–H band shifted from 3462.6 to 3441.7 cm⁻¹. These coordinated shifts, together with intensity changes, support hydrogen-bonding and/or coordination interactions between surface metal/oxygen sites and the –NH₂/–OH groups of chitosan [63,64].

3.4. Morphology and Elemental Composition

Pristine CoFe₂O₄ consisted of predominantly spherical-to-quasi-spherical particles with an approximate size range of 10–20 nm (Figure 4a and inset). The particles formed aggregates, as expected from their high surface energy and magnetic interactions [65]. After incorporation into chitosan, ferrite domains were distributed across the polymer matrix and appeared more spatially separated (Figure 4c). TEM likewise indicated a lower degree of compact agglomeration in the composite. These observations support the role of chitosan as a steric and interfacial stabilizer that restricts direct ferrite–ferrite contact [63,66].

EDX detected Co, Fe, and O in the ferrite sample and additionally C and N in CoFe₂O₄/CS, confirming the presence of the chitosan phase (Figure 4b,d). The lower relative Co and Fe signals in the composite are consistent with polymer dilution and surface coverage of the ferrite particles.

3.5. Optical Absorption and Photoluminescence

Figure 5a shows the diffuse-reflectance response of CS, CoFe₂O₄, and CoFe₂O₄/CS. Chitosan absorbed mainly in the ultraviolet region, whereas CoFe₂O₄ exhibited broad visible-light absorption. Tauc analysis (Figure 5b) yielded optical band gaps of 2.17 eV for CoFe₂O₄ and 2.61 eV for CoFe₂O₄/CS. The blue shift after polymer incorporation may arise from interfacial electronic perturbation, altered surface states, and the lower optical contribution of ferrite in the composite. Thus, the enhanced photocatalytic removal of CoFe₂O₄/CS cannot be attributed to band-gap narrowing; it must instead involve interfacial, adsorption, and charge-utilization effects.

PL emission provides a comparative indication of radiative electron–hole recombination [6,67,68,69]. CoFe₂O₄ showed a broad emission centered near 430 nm, whereas CoFe₂O₄/CS exhibited substantially lower intensity without a pronounced shift in the emission maximum (Figure 6). The quenching is consistent with reduced radiative recombination and/or more efficient trapping and interfacial transfer of photoexcited carriers. Because the composite also contains a lower fraction of emissive ferrite, dilution may contribute to the absolute intensity decrease; nevertheless, the PL trend, together with the higher rate constant, supports improved charge utilization in CoFe₂O₄/CS.

3.6. XPS Analysis

The XPS survey spectrum of CoFe₂O₄/CS confirmed the presence of C, N, O, Co, and Fe (Figure 7a). The comparatively weak Co and Fe intensities are consistent with partial coverage of the ferrite domains by the chitosan-rich surface layer rather than an absence of the inorganic phase.

The high-resolution Fe 2p spectrum contained satellite features at 718.1 and 732.4 eV and components assigned to Fe³⁺ in octahedral and tetrahedral coordination (Figure 7b) [70,71,72]. The Co 2p spectrum (Figure 7c) showed Co²⁺ spin–orbit components at 779.1/794.5 eV and 781.3/795.8 eV, together with satellites at 785.1 and 801.1 eV [70,71,73]. No distinct Co³⁺ contribution was resolved [74]. These assignments are consistent with the expected mixed-site cation distribution of spinel CoFe₂O₄.

The O 1s envelope was resolved into components at 530.0, 531.4, and 532.9 eV, attributable to lattice oxygen, carbonyl/defect-associated oxygen, and hydroxyl/ether oxygen from chitosan, respectively (Figure 7d) [40,71]. The C 1s spectrum (Figure 7e) contained contributions from C–C/C–H (284.5 eV), C–N/C–O (286.1 eV), and O–C–O/C=O (287.6 eV), whereas the N 1s spectrum (Figure 7f) showed amino and protonated-amino environments at 399.5 and 400.3 eV [75,76]. The coexistence of ferrite lattice oxygen and chitosan-derived C/N/O environments corroborates the composite structure indicated by XRD and FTIR.

3.7. BET Analysis and Magnetic Properties

Both materials exhibited type-IV nitrogen adsorption–desorption isotherms with H3 hysteresis, indicating mesoporous structures (Figure 8). CoFe₂O₄ had a BET surface area of 81.9 m² g⁻¹ and pore volume of 0.149 cm³ g⁻¹, whereas CoFe₂O₄/CS showed corresponding values of 51.3 m² g⁻¹ and 0.122 cm³ g⁻¹ (Table 2). The decrease is attributable to polymer coverage of ferrite surfaces and partial occupation or shielding of pore space. The mean pore diameter changed only slightly, from 6.8 to 7.4 nm, indicating that mesoporosity was broadly retained [41].

Room-temperature magnetic hysteresis loops are shown in Figure 9. CoFe₂O₄ exhibited Mₛ = 51.5 emu g⁻¹, Mᵣ = 3.8 emu g⁻¹, and Hc = 70.3 Oe. For CoFe₂O₄/CS, Mₛ decreased to 31.4 emu g⁻¹ because of dilution by non-magnetic chitosan, while Mᵣ and Hc increased to 7.46 emu g⁻¹ and 679.8 Oe, respectively. The change in coercivity may reflect altered interparticle interactions, anisotropy, and magnetic-domain reversal after immobilization in the polymer matrix.

Despite the lower saturation magnetization, CoFe₂O₄/CS was rapidly collected from water with an external magnet (Figure 9 inset). This recoverability is operationally important because it enables catalyst removal without filtration and supports repeated use. The textural and magnetic results therefore reveal a deliberate trade-off: chitosan decreases accessible ferrite surface area and magnetization but provides a dispersing, adsorptive, and recoverable composite architecture.

3.8. Photocatalytic Performance and Mechanistic Analysis

3.8.1. Time-Dependent MB Removal and Kinetics

The rate constant increased from 0.016 min⁻¹ for CoFe₂O₄ to 0.035 min⁻¹ for CoFe₂O₄/CS (Figure 10b), corresponding to an approximately 2.2-fold enhancement. This improvement occurred despite the lower BET surface area and wider apparent optical band gap of the composite. This improvement may be associated with a cooperative mechanism in which chitosan promotes MB enrichment near the interface, helps reduce ferrite aggregation, and creates an interfacial environment consistent with lower radiative recombination and more effective generation or utilization of oxidizing specie [76].

3.8.2. Effect of Initial pH

MB removal increased from 72.6% at pH 3 to 84.7%, 90.9%, and 98.5% at pH 5, 7, and 9, respectively, before decreasing slightly to 96.7% at pH 11 (Figure 11a). The corresponding rate constants were 0.011, 0.016, 0.020, 0.035, and 0.028 min⁻¹ (Figure 11b). Thus, pH 9 provided the highest efficiency and kinetic rate under the investigated conditions.

The pH dependence is consistent with the measured pH_PZC of 6.7 (Figure S1) and with protonation/deprotonation of chitosan amino groups [77,78,79]. Below pH_PZC, the composite surface and protonated –NH₃⁺ groups are predominantly positive, producing electrostatic repulsion toward cationic MB and limiting adsorption near photocatalytic sites. Above pH_PZC, a more negative surface favors MB accumulation and increases the local probability of oxidation. The modest decline at pH 11 may arise from excessive alkalinity, increased suspension screening/competition at surface sites, and altered radical chemistry [80].

The close correspondence between removal efficiency and k confirms that the pH effect was kinetic rather than solely an endpoint adsorption phenomenon. The optimum near pH 9 reflects a balance between favorable electrostatic uptake of MB and efficient photochemical oxidation at the composite–solution interface.

3.8.3. Effect of MB Concentration

At initial MB concentrations of 10, 20, 30, and 40 mg L⁻¹, removal efficiencies after 120 min were 99.3%, 98.5%, 90.5%, and 84.1%, respectively (Figure 12a). The decline at higher concentration reflects saturation of a finite number of adsorption/photocatalytic sites and stronger attenuation of incident light by the intensely colored solution [1,2,4]. These effects reduce the ratio of reactive species to dye molecules and slow removal.

The rate constants decreased from 0.044 to 0.035, 0.019, and 0.015 min⁻¹ as the MB concentration increased from 10 to 40 mg L⁻¹ (Figure 12b). An initial concentration of 20 mg L⁻¹ was selected for subsequent optimization because it retained near-complete removal and a relatively high rate constant while representing a more demanding dye loading than 10 mg L⁻¹.

3.8.4. Effect of Catalyst Dosage

Catalyst dosage controls the number of accessible reaction sites and the fraction of incident light that reaches the suspended particles [81]. CoFe₂O₄/CS loadings from 0.2 to 1.0 g L⁻¹ were therefore examined (Figure 13).

Increasing the dosage from 0.2 to 0.7 g L⁻¹ raised MB removal from 65.3% to 84.5% and then to 98.5%. The rate constant increased in parallel from 0.009 to 0.015 and 0.035 min⁻¹. The improvement is attributed to the larger number of adsorption and photocatalytic sites and the resulting increase in productive photon absorption [1,4].

At 1.0 g L⁻¹, removal decreased slightly to 96.7% and k to 0.027 min⁻¹. Excess catalyst can increase turbidity, light scattering, and particle shielding and may promote aggregation, thereby lowering the effectively illuminated surface area [2,80]. A dosage of 0.7 g L⁻¹ was therefore selected as the optimum.

3.8.5. Reusability, Reactive Species, and Proposed Mechanism

Under the optimized conditions, MB removal decreased from 98.5% in the first cycle to 96.2%, 95.7%, 92.6%, and 90.3% in cycles 2–5, respectively (Figure 14a). The loss after five cycles was 8.2 percentage points, indicating that most of the initial performance was retained. The gradual decline may result from incomplete desorption of MB/intermediates, partial blockage of pores and active sites, small handling losses during recovery, or progressive surface modification [1,5,81]. Magnetic collection simplified catalyst recovery and is a practical advantage of the composite architecture.

Scavenger tests provided qualitative evidence for the dominant reactive pathways (Figure 14b). In the absence of scavengers, removal was 98.5%. BQ and AgNO₃ reduced removal to 83.6% and 87.4%, respectively, suggesting secondary contributions from •O₂⁻- and electron-mediated reactions. EDTA and IPA caused much larger decreases, to 54.8% and 60.7%, demonstrating that photogenerated holes and •OH radicals were the principal oxidizing species [81].

The XRD pattern of a separately recovered sample after five irradiation–recovery cycles retained the characteristic spinel reflections with no evident new crystalline phase (Figure 15). This result indicates good phase-level stability, although it does not exclude gradual surface fouling, amorphous changes, or metal-ion leaching.

To evaluate the thermodynamic feasibility of the proposed reactive pathways, the conduction-band (E_CB) and valence-band (E_VB) edge potentials were estimated using Mulliken electronegativity theory [82]:

$$E_{VB}=\chi -E_e+0.5E_g$$

(7)

$$E_{CB}=E_{VB}-E_g$$

(8)

Here, χ is the absolute electronegativity of the semiconductor (5.8 eV for CoFe₂O₄), Eₑ is the free-electron energy on the hydrogen scale (4.5 eV vs. NHE), and E_g is the measured optical band gap. Using E_g = 2.61 eV for CoFe₂O₄/CS gave E_CB ≈ −0.005 eV and E_VB ≈ +2.605 eV versus NHE. The strongly positive valence-band potential is sufficient for direct hole oxidation and for oxidation of H₂O/OH⁻ to •OH, consistent with the strong inhibition caused by EDTA and IPA [83]. By contrast, E_CB is less negative than the standard O₂/•O₂⁻ potential, indicating that direct superoxide generation is not the dominant thermodynamic pathway. The moderate BQ effect may therefore arise from secondary oxygen-reduction routes associated with defects/interfacial states or from non-ideal scavenger selectivity.

The proposed mechanism is summarized in Figure 16. Visible-light excitation generates conduction-band electrons and valence-band holes. Interfacial interactions with chitosan and reduced radiative recombination increase the probability that holes reach the surface, where they can oxidize adsorbed MB directly or react with H₂O/OH⁻ to form •OH. Chitosan simultaneously concentrates cationic MB near the reactive interface, creating an adsorption-coupled oxidation pathway. Electrons and oxygen-derived species make a smaller contribution, in agreement with the scavenger results.

The dominant reactions can be represented schematically as follows:

CoFe₂O₄/CS + hν → CoFe₂O₄/CS (e⁻(CB) + h⁺(VB))

(9)

H₂O + h⁺(VB) → •OH + H⁺

(10)

OH⁻ + h⁺(VB) → •OH

(11)

MB + •OH/h⁺(VB) → CO₂ + H₂O + other products

(12)

These reactions describe a proposed degradation pathway rather than demonstrated complete mineralization. Because total organic carbon/chemical oxygen demand, degradation intermediates, cobalt/iron leaching, and real-wastewater matrix effects were not measured, the present results should be interpreted primarily as evidence of efficient MB removal and decolorization under controlled laboratory conditions. These measurements are important priorities for future scale-up and environmental-safety assessment.

4. Conclusions

A magnetically recoverable CoFe₂O₄/CS nanocomposite was prepared by combining lime-juice-assisted CoFe₂O₄ synthesis with chitosan immobilization. XRD confirmed preservation of the cubic spinel phase, whereas FTIR and XPS supported interfacial interactions between ferrite surface sites and chitosan amino/hydroxyl groups. Chitosan reduced ferrite aggregation and improved dark MB uptake, although it decreased the BET surface area from 81.9 to 51.3 m² g⁻¹, increased the apparent optical band gap from 2.17 to 2.61 eV, and lowered Mₛ from 51.5 to 31.4 emu g⁻¹. The composite nevertheless remained magnetically separable and exhibited a lower PL intensity and superior light-assisted removal. Under the optimized conditions (pH 9, 20 mg L⁻¹ MB, 0.7 g L⁻¹ catalyst, 30 W visible-light LED), 98.5% removal was achieved within 120 min with k = 0.035 min⁻¹, approximately 2.2 times the value for pristine CoFe₂O₄. Removal remained 90.3% after five cycles, and h⁺, •OH were identified as the dominant reactive species. This improvement may be partly associated with reduced radiative recombination of photogenerated charge carriers. Overall, these findings support the potential practical application of Lime-Juice-Assisted CoFe₂O₄/CS nanocomposites as magnetically recoverable photocatalysts for MB-containing wastewater.