Skip to main content
NCBI home page
RSC Advances logoLink to RSC Advances
. 2026 Oct 19. Online ahead of print. doi: 10.1039/d6ra05504d

Adsorption-assisted removal of chlorinated pollutants using a magnetically recoverable Fe3O4/CeO2 nanocomposite: kinetics, machine learning and first-principle insights

Monika Dubey a, Sunita Saharan a, Monika Joshi b, Piyush Kuchhal a, Ranjit Kumar c,✉
PMCID: PMC13679149  PMID: 42899129

Abstract

The development of efficient and recyclable catalysts for persistent organic pollutant removal remains a major challenge in environmental nanotechnology. Herein, a mesoporous Fe3O4/CeO2 nanocomposite was synthesized and systematically characterized, confirming a particle size of (20 ± 10 nm) with a high surface area (44.6 m2 g−1), magnetic recoverability, and abundant oxygen vacancies. The catalyst exhibited excellent adsorption-assisted removal of chlorobenzene (CB) and 2,4-dichlorophenol (2,4-DCP), achieving ∼100% removal within 70 min at near-neutral pH, with ∼98% stability after five cycles. Scavenger experiments revealed that hydroxyl radicals (˙OH) are the dominant reactive species, with additional contribution from superoxide radicals (O2˙−), confirming adsorption-assisted removal pathway. Kinetic analysis followed pseudo-first order (kpfo = 0.072 min−1 for CB and 0.0399 min−1 for 2,4-DCP) behaviour and showed strong agreement with Langmuir–Hinshelwood and Runge–Kutta models (R2 ≈ 0.99). An ANFIS model, trained on 380 datasets, accurately predicted adsorption-assisted removal under varying conditions. Density functional theory (DFT) calculations demonstrated that the Fe3O4/CeO2 (1 : 1) system possesses the most negative binding energy, indicating superior interfacial stability. PDOS analysis revealed enhanced overlap of Fe 3d, Ce 4f, and O 2p orbitals. This study highlights a synergistic experimental–theoretical–AI approach for advanced environmental remediation.

The development of efficient and recyclable catalysts for persistent organic pollutant removal (image was enhanced using Microsoft Copilot).graphic file with name d6ra05504d-ga.webp

1. Introduction

Chlorinated organic pollutants (COPs) such as chlorobenzene (CB) and 2,4-dichlorophenol (2,4-DCP) are widely utilized in chemical manufacturing, pesticides, pharmaceuticals and dye industries.1,2 Because of their high chemical stability, toxicity and resistance to natural biodegradation, these compounds persist in aquatic environments.3–10 These compounds pose severe risks to ecosystems, marine life and human health.11 Their long-time accumulation in the environment can be responsible for bioaccumulation, endocrine disruption and carcinogenic effects. Therefore, their removal from industrial wastewater is an urgent environmental priority.12

Conventional treatment approaches such as biological degradation, chemical oxidation and physical adsorption have been widely explored for COP removal.10,13–15 However, these methods often suffer from several limitations, such as slow removal kinetics, incomplete mineralization, high operational cost, a complex separation process and the generation of secondary pollutants.16,17 Consequently, there is a need to develop advanced multifunctional nanomaterials capable of simultaneously facilitating adsorption and removal to enhance pollutant removal efficiency.18

Among various nanomaterials, magnetic metal-oxide nanocomposites have attracted considerable attention due to their unique physicochemical attributes like high surface area, tunable redox activity and ease of magnetic separation from aqueous systems.19 Magnetite (Fe3O4) nanoparticles are promising materials on account of their excellent magnetic responsiveness, chemical stability and capability to participate in electron transfer reactions. Fe3O4 possesses a mixed-valence iron structure (Fe2+/Fe3+) which can decompose various organic contaminants. These radicals can effectively disintegrate complex organic molecules, together with chlorinated compounds, dyes, pharmaceuticals and pesticides into less harmful or mineralized products like CO2 and H2O. Another important advantage of Fe3O4 nanoparticles is their superparamagnetic behaviour.20 Because of that, the rapid magnetic separation of the catalyst from treated water using an external magnetic field significantly simplifies catalyst recovery and reuse.19 So, the remediation process can be more economical and environmentally sustainable. However, pure Fe3O4 nanoparticles tend to aggregate because of strong magnetic interactions, which can minimize their active surface area and catalytic efficiency. Cerium oxide (CeO2), a rare earth metal oxide with a fluorite crystal structure, has emerged as a highly attractive catalytic material owing to its remarkable oxygen storage capacity and reversible Ce3+/Ce4+ redox cycling.21–24 The presence of a mesoporous structure further enhances the catalytic performance of CeO2 by providing a large specific surface area, uniform pore distribution and more abundant active sites.25,26 When combined with magnetic Fe3O4, CeO2 can form synergistic heterostructures that enhance interfacial electron transfer, improve catalytic activity and provide magnetic recoverability.24 The integration of Fe3O4 with mesoporous CeO2 can also improve surface area and dispersion of active sites, which are critical for adsorption-assisted removal processes. Although Fe3O4/CeO2 composites have previously been investigated for Fenton-like degradation of chlorophenols, our study focuses on adsorption-assisted removal and oxidative transformation of chlorobenzene and 2,4-dichlorophenol under near-neutral aqueous conditions and integrates experimental characterization with kinetic, numerical, DFT and ANFIS analyses. The objective is therefore not to claim novelty of the Fe3O4/CeO2 composition itself, but to provide an integrated assessment of pollutant removal, transformation behaviour and structure property relationships. Despite recent progress in magnetic oxide nanocomposites, the integration of adsorption-assisted catalysis with predictive modelling and artificial intelligence approaches remains relatively unexplored.27 The combination of advanced nanomaterials with first principle theory (Density Functional Theory-DFT: a quantum mechanical modelling method)28 and machine learning tools can significantly improve the understanding and optimization of the integration of nanomaterials and pollutant removal systems.

Herein, we report the successful synthesis of a magnetically recoverable mesoporous Fe3O4/CeO2 nanocomposite with mesoporous architecture for the efficient removal of chlorinated pollutants from aqueous media. DFT analysis has been performed to study the stability of synthesized nanocomposite. The structural, morphological and magnetic properties of the synthesized nanocomposite were systematically characterized and analyzed. The adsorption-assisted removal performance toward CB and 2,4-DCP was investigated under various experimental conditions. Furthermore, the adsorption kinetics were evaluated using pseudo-first order and Langmuir–Hinshelwood models. The adsorption behaviour was numerically simulated using the Runge–Kutta (RK4) method. To further enhance predictive capability, an Adaptive Neuro-Fuzzy Inference System (ANFIS) was implemented to model the influence of operational parameters on pollutant removal efficiency. The integration of nanomaterial engineering, kinetic modelling, machine learning prediction, and first-principles calculations provides a comprehensive framework for understanding and optimizing catalytic wastewater treatment systems. The results demonstrate that the Fe3O4/CeO2 nanocomposite offers a highly efficient, recyclable and scalable platform for environmental remediation of persistent organic pollutants.

2. Materials and methods

2.1. Materials

All chemicals of analytical grade were purchased from Merck Chemicals (India). Chlorobenzene (CB), 2,4-dichlorophenol (2,4-DCP), cerium ammonium nitrate-CAN (NH4)2Ce(NO3)6, NaOH, KOH, acrylamide, (CH2OH)2, FeCl2, HNO3, FeCl3, isopropanol (IPA), benzoquinone (BQ) and ethylene di amine tetra acetic acid (EDTA) were used without any purification.

2.2. Synthesis of Fe3O4/CeO2 nanocomposite

Mesoporous CeO2 was synthesized via a hydrothermal approach using cerium ammonium nitrate (CAN) as the cerium precursor.29 In this procedure, 1.25 g of CAN was dissolved in 30 mL of deionized water, followed by the addition of 10 mL of ethylene glycol drop by drop under continuous stirring. Subsequently, 500 mg of acrylamide was introduced to the mixture and stirring was maintained for 2 h. The pH was adjusted to 11 by the dropwise addition of 1 M NaOH solution. The precipitate was transferred into an autoclave and subjected to hydrothermal treatment at 150 °C for 48 h. The product was washed, dried at 100 °C for 6 h and calcined at 300 °C for 3 h. The resulting white powder corresponded to mesoporous cerium(iv) oxide.

Magnetic mesoporous Fe3O4/CeO2 nanocomposites were fabricated using an ultrasonically assisted co-precipitation method. Initially, a 0.2 M FeCl3 solution was added to 0.1 M urea, and the mixture was stirred at 90 °C for 2 h. Following this, 0.1 M FeCl2 was introduced and stirring continued for 25 min. Co-precipitation was initiated by the dropwise addition of 0.3 M NaOH until the pH reached 11. Subsequently, 500 mg of as-prepared CeO2 was incorporated in a 1 : 1 mass ratio with Fe3O4. The resulting dispersion was ultrasonicated (30 °C, 150 W, 40 kHz) for 1 h and allowed to age undisturbed for 15 h. The product was repeatedly washed and magnetically separated, followed by drying at 100 °C for 6 h. The pH played a critical role in regulating the morphology and interfacial interaction within the Fe3O4/CeO2 nanocomposite. The precipitation of Fe(OH)3 and Fe(OH)2 commenced around pH 9, whereas incorporation of CeO2 at pH 11 ensured that the mixed valence states of Ce3+/Ce4+ remained unaffected by Fe2+ ions, as represented in eqn (1):

2.2. 1

Ultrasonic irradiation facilitated stable magnetic Fe3O4/CeO2 hybrid on account of intense microturbulence, and shear forces generated at the ceria nanoparticles surface, promoting nucleation and subsequent growth of Fe3O4 nanoparticles.30 The collapse of cavitation bubbles produced abundant reactive H+ and OH− species, which can enhance nucleation rates and ensure high product yield.31 However, direct chemical coupling at the Fe3O4/CeO2 interface may not be conclusively established solely from the synthesis procedure. A schematic representation (Fig. 1a) illustrates the proposed formation mechanism, highlighting the uniform distribution of Fe3O4 nanoparticles within the mesoporous CeO2 framework, leading to improved structural integrity and synergistic catalytic functionality. Fe3O4 nanoparticles were synthesized under identical conditions as the composite, but without the addition of CeO2, to enable comparative analysis.

Fig. 1. (a) Schematic of Fe3O4/CeO2 nanocomposites synthesis (b) XRD pattern of synthesized (black) Fe3O4, (red) CeO2 and (blue) mesoporous magnetic Fe3O4/CeO2 (* represents magnetite peak and Inline graphic represents CeO2 peak).

Fig. 1

Open in a new tab

2.3. Simulation and characterization

The unit cells of CeO2, Fe3O4 and nanocomposites were constructed in QuantumATK. Geometry optimization of the unit cells was realized via the Limited-Memory Broyden–Fletcher–Goldfarb–Shanno (L-BFGS) algorithm. A maximum force tolerance of 0.02 eV Å−1 and a stress error tolerance of 0.1 GPa were applied for all structural parameters. A gamma-centred 5 × 5 × 1 k-point mesh of the Brillouin zone with a density mesh cut off 2000 eV and a temperature of 300 kelvin (K) was employed to perform geometry optimization.32 Generalized gradient approximation (GGA) with predefined functional Perdew–Burke–Ernzerhof (PBE) integration with Grimme DFT-D3 (van der Waals corrections) and energy exchange–correlation and DFT calculations were evaluated for different ratios of CeO2 and Fe3O4 in nanocomposites such as (1 : 1, 1 : 2, 2 : 1). The projected density of states (PDOS) of the 1 : 1 Fe3O4/CeO2 heterostructure was calculated using a spin-polarized DFT+U approach. The on-site Coulomb interaction parameters were set to U = 5 eV for Ce 4f states and U = 4 eV for Fe 3d states. A density mesh cut-off of 250 Ry was employed to ensure accurate representation of the charge density.

The adsorption process was modelled using an Adaptive Neuro-Fuzzy Inference (ANFI) System based on a Sugeno-type fuzzy logic framework and implemented in MATLAB (R2025a). The ANFIS architecture comprises a five-layer hybrid neuro-fuzzy network, in which each layer performs a specific computational function essential for fuzzy inference.33 In this study, the system inputs consisted of key operational parameters influencing adsorption pH, initial COP concentration, contact time, Fe3O4/CeO2 dosage and stirring rate. These variables were transformed into fuzzy sets through first-order Sugeno membership functions in the fuzzification layer.34 The rule-generation layer (third layer) established the logical relationships between the input membership functions to form the rule base governing system behaviour. The resulting fuzzy outputs were subjected to defuzzification in the subsequent layer, where output membership functions were used to convert the fuzzy rule consequences into numerical estimates of the adsorption efficiency. In the final layer, all normalized firing strengths and corresponding outputs were aggregated into a single node that represented the predicted adsorption percentage. The overall ANFIS configuration used in this work is illustrated schematically in Fig. S1, demonstrating the interaction between input variables, fuzzy rule sets, and the model's predictive output.

XRD confirmed crystalline phases of both Fe3O4 and CeO2 and validated the formation of the Fe3O4 and CeO2 for the 2θ range 20–70°. A scanning electron microscope (SEM; ZEISS EVO-18) with energy dispersive X-ray spectroscopy (EDS) analyser was used to evaluate the structure and morphology of Fe3O4, and CeO2 and a transmission electron microscope ThermoFisher Talos F200X (S) TEM, was used for Fe3O4/CeO2 high resolution micrograph. BET (Brunauer, Emmett and Teller) surface analyser, and BJH (Barrett–Joyner–Halenda) model were applied for surface area analysis and porosity. VSM magnetization curves demonstrated strong magnetic behaviour suitable for rapid separation. The magnetic characteristics of synthesized iron oxides and nanocomposites have been studied by measuring room temperature magnetization (emu g−1) versus magnetic field (Oe) variation using VSM (Model number-PAR 155). Fourier-transform infrared spectroscopy (FTIR) (wavenumber range 4000–400 cm−1) was done using Nicolet iS5, ThermoScientific, India. FTIR spectroscopy established surface functional groups and pollutant catalyst interactions. To determine the chemical composition of Fe3O4/CeO2 X-ray Photoelectron Spectroscopy (XPS) Prevac XPS was used. The UV absorbance spectra of adsorption in an aqueous solution were measured using UV-vis spectrophotometer UV-1800 (Shimadzu's Co., UV-Probe 2.42). High-performance liquid chromatography (HPLC) analysis was performed to monitor the residual concentration of chlorinated organic pollutants during the treatment process. A C18 reversed-phase column (250 × 4.6 mm, 5 µm) was used for chromatographic separation. For chlorobenzene (CB), the mobile phase consisted of acetonitrile/water/acetic acid (50 : 48 : 2, v/v/v), with an isocratic flow rate of 1.0 mL min−1 and UV detection at 280 nm. For 2,4-dichlorophenol (2,4-DCP), methanol/water (70 : 30, v/v) was employed as the mobile phase at a flow rate of 1.0 mL min−1, and detection was carried out at 265 nm.

2.4. Adsorption-assisted removal experiments and kinetics

The pH point of zero charge-pHpzc was investigated using the pH drift method. Briefly, 50 mg of catalyst was added to 50 mL of 29 mg NaCl solution with an initial pH adjusted between 2 and 12. The suspensions were stirred for 24 h at room temperature (RT), and the final pH was noted. The pHpzc was evaluated at the point where ΔpH (pH_final − pH_initial) equals zero.

To elucidate the dominant reactive species responsible for the adsorption-assisted removal of CB and 2,4-DCP, radical scavenger experiments were performed under optimized conditions (100 ppm pollutant, 100 mg per L catalyst, pH ≈ 7). Different scavengers were introduced into the reaction system, IPA as a hydroxyl radical (˙OH) scavenger, BQ as a superoxide radical (O2˙−) scavenger and ethylene diamine tetraacetic acid (EDTA) as a hole (h+) scavenger. 1 mM amount of each scavenger was added prior to the Fe3O4/CeO2 nanocomposite addition.

For an aqueous solution of 100 ppm CB and 2,4-DCP, 100 mg of COPs were added to 1 L of DI water. For pH adjustment, 10% HNO3 and 1 M NaOH were added. The percentage removal efficiency (R%) of COPs was investigated by:

2.4. 2

The experimental kinetic data were evaluated using the pseudo-first-order model (eqn (3)) and second-order model (eqn (4)) in MATLAB 2025:

2.4. 3
2.4. 4

where, qe = equilibrium adsorbed amount, qt = adsorbed amount at time t, and kpfo = pseudo-first order rate constant and kpso = pseudo-second order rate constant. Where C0 = starting concentration of COPs solution, Ct = concentration at reaction time t (min), q (mg g−1) = the quantity of adsorbed COPs onto Fe3O4/CeO2, v (L) = volume of COP solution and m (g) = mass of catalysts.35 The UV absorbance spectrum of COPs in an aqueous solution was measured using UV-vis spectrophotometer UV-1800 (Shimadzu's Co., UV-Probe 2.42).

The Langmuir–Hinshelwood model36 was employed to describe the surface-mediated adsorption mechanism, as shown in eqn (5), and was implemented in MATLAB 2025.

2.4. 5

where, r = adsorption-assisted removal rate (mg L−1 min−1), k = surface reaction rate constant (min−1), K = adsorption equilibrium constant (L mg−1) and C = COPs concentration in solution (mg L−1). The concentration-dependent adsorption behaviour and reduction in efficiency at higher pollutant concentrations confirm that the reaction proceeds via adsorption of COP molecules on the catalyst surface, followed by surface reaction.

The adsorption dynamics were further simulated using the fourth-order Runge–Kutta (RK4)37 numerical method, which is widely recognized for its accuracy and stability in solving differential equations without the need for an analytical solution. This approach was employed to model the time-dependent removal of chlorinated organic pollutants (COPs) by calculating the adsorbed amount at successive time intervals following the rate equation (eqn (6)) as modelled in MATLAB 2025.

2.4. 6

RK4 can simulate this change accurately and stably across time steps as eqn (7)

2.4. 7

where, s = time step, l1 = f(tn,qtn), Inline graphic, Inline graphic and l4 = f(tn + h,qtn + hl3).

This RK4 algorithm offers a highly precise estimate of the next value qtn+1, given the current value qtn and adsorption-assisted removal rate.

3. Results and analysis

3.1. Structural and magnetic characterization

XRD patterns confirmed the coexistence of spinel Fe3O4 and fluorite CeO2 phases. The powder X-ray diffraction (XRD) profiles of calcined CeO2, Fe3O4, and the Fe3O4/CeO2 nanocomposite are presented in Fig. 1b. The diffraction pattern of Fe3O4 displays well-defined reflections at (220), (311), (400), (422), (511) and (440), characteristic of the inverse cubic spinel structure and consistent with the standard JCPDS File No. 85-1436.38 CeO2 exhibits distinct and narrow diffraction peaks indexed to the (111), (200), (220) and (311) planes of the fluorite-type face-centered cubic phase (JCPDS No. 34-0394),39,40 affirming the high crystallinity and phase purity of the material within limit of XRD. No impurity phases were detected. The diffraction pattern of the Fe3O4/CeO2 nanocomposite contains the full set of characteristic peaks from both constituent phases. The slight shift in the CeO2 peaks in the composite can be attributed to interfacial strain from the lattice mismatch between the Fe3O4 and CeO2 domains, and/or limited substitution of Fe3+ (0.645 Å) for Ce4+ (0.97 Å) at the interface, which shrinks the CeO2 cell. The sharper peaks in the composite is possibly a result of ultrasonic treatment of the reaction mixture, as acoustic cavitation promotes interparticle collision and oriented attachment of the nanocrystalline building blocks, and the alkaline aging promotes partial dissolution–reprecipitation of the smallest, most defective crystallites. The longer aging time could have promoted larger crystallite size. In addition, repeated washing and magnetic separation retain only the CeO2 associated with Fe3O4, which can bias the retained fraction toward larger crystallites.

The SEM (Fig. 2a and b) and EDS (Fig. 2c) analyses confirm the CeO2 nanoparticles with an average particle size of approximately 40 nm (Fig. 2d). Energy dispersive X-ray spectroscopy (EDS) analysis (Fig. 2c) was performed to verify the elemental composition of the synthesized Ce and O in CeO2 nanocomposite. The SEM micrograph (Fig. 2e and f) illustrates a cluster of Fe3O4/CeO2 nanocomposite quasi-spherical nanoparticles with a characteristic size distribution of approximately 20 ± 10 nm (Fig. 2h). The EDS spectrum (Fig. 2g) clearly shows characteristic peaks corresponding to oxygen (O), iron (Fe) and cerium (Ce), confirming the successful integration of both oxide phases. The strong peak at approximately 0.52 keV corresponds to the O Kα transition, indicating the presence of lattice oxygen within the oxide structure. The signals observed at around 0.71 keV and 6.40 keV are attributed to Fe Lα and Fe Kα transitions, respectively, confirming the presence of magnetite nanoparticles. Additionally, distinct peaks at approximately 4.84 keV and 5.26 keV correspond to Ce Lα and Ce Lβ emissions, which verify the incorporation of cerium oxide within the composite structure. Quantitative analysis indicates that Fe, Ce, and O are the dominant elements with no detectable impurity peaks, suggesting the high purity of the synthesized material. The presence of these characteristic peaks collectively confirms the successful formation of the Fe3O4/CeO2 nanocomposite, which is expected to enhance catalytic activity due to the synergistic interaction between magnetite and ceria phases. The porous interconnected morphology, nanoscale dimensions and uniform particle size distribution are expected to provide a high surface area and abundant active sites, making the material highly suitable for applications in catalysis and environmental remediation.

Fig. 2. (a and b) SEM images (c) EDS (inset: elemental analysis) (d) particle size distribution of CeO2 (e and f) SEM images (g) EDS (inset: elemental analysis) (h) particle size distribution of Fe3O4/CeO2 nanocomposites.

Fig. 2

Open in a new tab

The TEM images (Fig. 3a–c) also show particles (20 ± 10 nm) displaying well-defined but slightly diffuse boundaries, which indicate the presence of polycrystalline domains with moderate crystallinity typical for mixed-oxide nanocomposites. A notable feature of the micrograph (SEM and TEM of nanocomposite) is the difference in relative contrast between particles, which is characteristic of a composite consisting of elements with distinct electron-scattering cross-sections. The nanoparticles appear to form intimate interparticle contacts, which suggest the formation of a hetero-assembled architecture rather than simple physical mixing. This close spatial arrangement is essential for efficient interfacial charge transfer and magnetically coupled behavior. The particle shape and significant aggregation indicate good colloidal stability during synthesis. The microscopic graphs support the successful preparation of Fe3O4/CeO2 nanocomposites with nanoscale integration. That is a prerequisite to achieve synergistic functional properties arising from the Ce3+/Ce4+ redox cycling of ceria. Fig. 3d demonstrates selected-area electron diffraction (SAED) pattern of Fe3O4/CeO2 nanocomposites. The diffraction pattern exhibits concentric rings consistent with polycrystalline CeO2 (fluorite, Fm3̄m) and Fe3O4 (inverse spinel, Fd3̄m). Rings were indexed by azimuthal integration of the pattern and matching of observed ring radii to theoretical reciprocal spacings: the three strongest inner rings were assigned to Fe3O4 (400) at d = 0.210 nm, CeO2 (220) at d = 0.191 nm, and Fe3O4 (422) at d = 0.171 nm (see Table S1 in SI). Indexing used lattice parameters a(CeO2) = 5.411 Å and a(Fe3O4) = 8.396 Å; calculated d-spacings are reported in Table S1.

Fig. 3. TEM of Fe3O4/CeO2 (a) 500 nm (b) 100 nm (c) 10 nm scale (d) SAED pattern of Fe3O4/CeO2.

Fig. 3

Open in a new tab

Nitrogen adsorption–desorption isotherms (Fig. 4a) for Fe3O4/CeO2 are characteristic of type-IV profiles, confirming mesoporosity. The steep uptake in the low relative pressure region (p/p0 = 0.1–0.4) further supports the presence of well-developed mesopores. The specific surface area is calculated as 44.6 m2 g−1. In our previous work BET analysis yielded a specific surface area of 270.2 m2 g−1 for synthesized CeO2.29 Fe3O4 often has a lower surface area than mesoporous CeO2. In the nanocomposite, CeO2 may contribute to the high surface area, but Fe3O4 may partially block pores or increase particle growth. Thus, the composite surface area is lower than that of pure mesoporous CeO2.

Fig. 4. Synthesized Fe3O4/CeO2 (a) BET analysis, inset: BJH analysis (b) VSM graph (inset: (A) aqueous suspension (B) magnetic separation).

Fig. 4

Open in a new tab

The magnetic characteristics of Fe3O4 and Fe3O4/CeO2 were evaluated via room-temperature magnetization (M–H) hysteresis (Fig. 4b). Fe3O4 exhibits clear ferromagnetic behavior with a saturation magnetization (Ms) of 40.2 emu g−1 (1.5µB), consistent with exchange interactions between Fe ions occupying tetrahedral and octahedral sites. The Fe3O4/CeO2 composite retains ferromagnetic ordering, although with a reduced Ms value of 21.04 emu g−1 (0.5µB), which can be attributed to the non-magnetic (diamagnetic) nature of CeO2 and partial aggregation of Fe3O4 nanoparticles on the CeO2 surface. Here, it is noted that composite reached magnetic saturation at a relatively low applied field (∼5000 Oe), suggesting facile magnetic response. This behavior enables rapid magnetic separation, as demonstrated in the digital images showing the dispersion before and after exposure to an external magnet (Fig. 4b, insets A and B). VSM measurements demonstrated excellent magnetic saturation, facilitating effortless recovery of the catalyst from aqueous media.

The XPS spectrum (Fig. 5a) authenticates the presence of Fe, Ce and O elements in the synthesized Fe3O4/CeO2 nanocomposite without noticeable impurities. The high-resolution Fe 2p spectrum (Fig. 5b) shows two characteristic peaks at approximately 710.6 eV and 724.2 eV corresponding to Fe 2p3/2 and Fe 2p1/2 respectively, with a satellite feature around 718 eV, confirming the Fe3O4 phase. The Ce 3d spectrum (Fig. 5c) exhibits multiple components associated with the coexistence of Ce3+ and Ce4+ states. The Ce3+/(Ce3+ + Ce4+) ratio is ∼0.38, confirming substantial oxygen vacancy formation. Deconvoluted into three components: lattice O2− (529.4 eV, ∼42%), defect/oxygen-vacancy oxygen (531.0 eV, ∼36%), and adsorbed O2/C–O–C (532.8 eV, ∼22%). The defect-oxygen fraction (∼36%) is consistent with the Ce3+ fraction (∼38%), providing cross-validated evidence for oxygen vacancies. These results substantiate the successful formation of the Fe3O4/CeO2 nanocomposite with abundant oxygen vacancies.

Fig. 5. XPS spectrum of Fe3O4/CeO2: (a) survey spectrum, high-resolution XPS spectra of (b) Fe 2p, (c) Ce 3d, (d) O 1s.

Fig. 5

Open in a new tab

3.2. Optimization of adsorption-assisted removal parameters

The temporal profiles for CB, 2,4-DCP removal at varying initial concentrations (50, 100 and 200 ppm) using catalysts as 100 mg L−1 of Fe3O4/CeO2 nanocomposite are presented in Fig. 6a and b. All reactions were performed at ambient temperature, and the solution pH remained close to neutral (pH ∼7). In the absence of stirring, catalyst addition alone did not induce measurable adsorption-assisted removal. Under catalytic conditions, nearly complete removal of CB was achieved within 50 min for an initial concentration of 50 ppm and within 70 min for 100 ppm. Almost complete removal of 200 ppm CB required approximately 80 min, whereas only ∼72% removal was achieved for 500 ppm after the same reaction duration. These results highlight the concentration-dependent kinetics of CB degradation. As the initial CB concentration increases while the catalyst loading and the density of active zones remain constant, competition for adsorption sites intensifies, thereby slowing surface-mediated reactions and reducing overall removal efficiency. Consequently, lower pollutant concentrations facilitate higher removal efficiencies under identical loading conditions. Based on these findings, an initial CB concentration of 100 ppm was decided for subsequent optimization experiments. The removal efficiency of COPs is strongly influenced by the availability of active sites on the catalyst surface, where electrostatic interactions play a significant role. At lower pollutant concentrations, sufficient active sites are available, facilitating efficient adsorption and subsequent removal.41 However, as the concentration of COPs increases, competition among pollutant molecules for the limited adsorption sites becomes more pronounced, leading to reduced removal efficiency. Consequently, higher pollutant concentrations require longer contact time to achieve comparable removal levels.

Fig. 6. Removal of (a) 2,4-DCP (b) CB (c) effect of pH (d) effect of amount of catalysts.

Fig. 6

Open in a new tab

The observed decline in removal percentage at elevated concentrations can be attributed to the saturation of active sites on the adsorbent surface. Therefore, controlling the initial concentration of pollutants is essential to achieve optimal performance.42 Based on these observations, an initial concentration of 100 ppm was selected for subsequent experiments. Under optimized conditions, complete removal of COPs (100 ppm) was achieved within 70 min at pH ∼7 using a catalyst dosage of 100 mg L−1 with continuous stirring at 400 rpm.

pH is another critical factor governing both adsorption behavior pathways.11 To evaluate its influence, removal reactions were conducted at pH values of 3, 5, 7, 9, and 11 using 100 ppm CB and 2,4-DCP using 100 mg L−1 of Fe3O4/CeO2 (Fig. 6c). pH adjustments were performed using 1 M KOH or 10% HNO3. At pH ∼7, the system achieved nearly complete (∼100%) adsorption-assisted removal within 70 min. In contrast, removal efficiencies were substantially lower in acidic media ∼45% at pH 3 and ∼56% at pH 5 and only moderately improved in alkaline conditions, reaching ∼72% at pH 9 and ∼68% at pH 11. CB is a neutral, non-ionizable chlorinated aromatic compound under the experimental pH range. Therefore, its pH-dependent removal is primarily associated with changes in the surface charge state of the Fe3O4/CeO2 nanocomposite and the availability of surface adsorption sites rather than pollutant ionization. In contrast, 2,4-DCP is a weakly acidic compound, and its degree of deprotonation increases with increasing pH. Consequently, its interaction with the catalyst surface is affected by both the surface charge of Fe3O4/CeO2 and the changing speciation of 2,4-DCP. At low and higher the pHpzc, electrostatic interactions between the increased charge at Fe3O4/CeO2 surface and 2,4-DCP species can become less favourable, potentially reducing adsorption. Therefore, the observed pH dependence of 2,4-DCP removal represents a combined effect of surface charge, pollutant speciation, adsorption, and subsequent surface-associated transformation, rather than a simple electrostatic interaction. These observations indicate that a neutral environment most effectively supports the adsorption-assisted removal process, likely due to favorable surface charge interactions and optimal oxidation states of the Fe3O4/CeO2. Thus, pH ∼7 was identified as the optimum working condition.

The influence of catalyst dosage was further examined by varying the Fe3O4/CeO2 concentration from 50 to 200 mg L−1 for a 100 ppm COPs solution (Fig. 6d). Increasing catalyst loading enhanced removal efficiency by providing a greater number of accessible adsorption and reaction sites. However, beyond 100 mg L−1, further increases produced only marginal improvements, indicating that active-site saturation or light/diffusion limitations may occur. Accordingly, 100 mg L−1 was designated as the optimal catalyst dosage for all remaining experiments.

3.3. Point of zero pH effect

The point of zero charge (pHpzc ≈ 6.8) indicates that the surface of the Fe3O4/CeO2 nanocomposite remains positively charged at pH values below 6.8 and becomes negatively charged at pH values above this threshold (Table S2). As shown in Fig. 6c, the maximum removal efficiency observed at pH ≈ 7, which is close to the pHpzc, suggests that electrostatic repulsion is minimized under these conditions, allowing other interactions such as hydrophobic forces, hydrogen bonding, π–π interactions, and catalytic oxidation at Ce active sites to dominate.

3.4. Kinetic modelling

The combined kinetic and mathematical modelling demonstrates that the adsorption-assisted removal of CB and 2,4-DCP is governed primarily by pseudo-first-order kinetics43 for R2 = 0.99. The H–L model successfully explains the competition for active sites and validates the surface-controlled catalytic mechanism for both CB and 2,4-DCP. The RK4 simulations provide strong numerical validation of the experimentally obtained kinetic parameters. Fig. 7 shows integrated experimental, kinetic, numerical and machine-learning modelling of chlorobenzene and 2,4-DCP adsorption-assisted removal at 100 ppm using Fe3O4/CeO2 (100 mgL−1, pH 7). Fig. 7a–c present a comparison of experimental adsorption-assisted removal profiles with RK4 numerical simulations, pseudo-first-order kinetic fitting, and L–H model for surface reaction controlled behaviour. The kinetic parameters of all types of modelling have been shown in Table 1.

Fig. 7. (a) Comparison of experimental adsorption-assisted removal with Runge–Kutta (RK4) numerical simulations, (b) pseudo-first-order kinetic fitting, (c) Langmuir–Hinshelwood model illustrating surface reaction-controlled behaviour, and (d) parity plot comparing experimental and ANFIS-predicted adsorption efficiencies.

Fig. 7

Open in a new tab

Table 1. Final kinetic parameters for COPs adsorption-assisted removal.

Kinetic model CB 2,4-DCP
Rate constant R 2 Rate constant R 2
PFO 0.072 min−1 0.99 0.0399 min−1 0.99
PSO 0.018 g mg−1 min−1 0.95 0.012 mg−1 min−1 0.81
H–L 4.0 mg L−1 min−1 0.99 3.0 mg L−1 min−1 0.99
RK4 0.072 min−1 0.99 0.0399 min−1 0.99
Open in a new tab

3.5. Proposed mechanism

Mesoporous CeO2, magnetite (Fe3O4) nanoparticles, and the ultrasound-assisted magnetic Fe3O4/CeO2 nanocomposite44–49 were evaluated as catalysts for the adsorption-assisted removal of 100 ppm CB and 2,4-DCP. Pure Fe3O4 exhibited an adsorption efficiency of approximately 21% within 70 minutes, which can be attributed to the presence of surface –OH groups capable of generating reactive hydroxyl species. Mesoporous CeO2, with a large specific surface area, contains abundant oxygen vacancies and surface radicals, enabling a higher adsorption-assisted removal efficiency of about 42% under the same conditions as shown in eqn (8).

CB/2,4-DCP(aq) → CB/2,4-DCP(ads) 8

Among all the tested systems, the Fe3O4/CeO2 nanocomposite demonstrated the most rapid catalytic response, achieving nearly 100% adsorption-assisted removal within 70 minutes at neutral pH (∼7) using 100 mg L−1 of catalyst. The composite synthesized under ultrasonic irradiation possessed an even higher surface area of 44 m2 g−1, together with an increased concentration of oxygen vacancies (O2˙−) and ˙OH radicals. The adsorption step, therefore, acts as a precursor to catalytic oxidation, ensuring efficient electron transfer and rapid adsorption-assisted removal. The adsorption efficiency of CB and 2,4-DCP using Fe3O4, CeO2, and Fe3O4/CeO2 nanocomposites has been displayed in Fig. 8a. These characteristics significantly enhanced the adsorption–catalysis interaction at room temperature.

Ce3+ + O2 → Ce4+ + O2˙− 9
H2O → ˙OH + H+ 10
Fe2+ ↔ Fe3+ + e− 11
Ce4+ + e− ↔ Ce3+ 12
O2 + e− → O2˙− 13
H2O + e− → ˙OH 14
CB/2,4-DCP + ˙OH/O2˙− → intermediates 15
Intermediates → CO2 + H2O + Cl− 16

Fig. 8. (a) Adsorption study of synthesized materials (b) scavenger study: effect of different radical scavengers (IPA, BQ, and EDTA) on the removal efficiency of CB and 2,4-DCP using Fe3O4/CeO2. The significant decrease in removal efficiency in the presence of IPA confirms hydroxyl radicals (˙OH) as the dominant reactive species. (c) Reusability for Fe3O4/CeO2 nanocomposite (d) application for the adsorption-assisted removal of other organic pollutants.

Fig. 8

Open in a new tab

The exceptional adsorption-assisted catalytic performance of the Fe3O4/CeO2 composite can be ascribed to a strong synergistic interplay between the Fe3O4 and CeO2 phases. The interfacial contact between the two oxides promotes efficient separation of photon or chemically generated electrons, preventing recombination and thereby enhancing redox activity. The UV-vis spectra for CB adsorption-assisted removal (in Fig. S3) show the main absorbance peak at 278 nm. Initially, Fe3O4 captures CB molecules on its surface; subsequently, electron exchange between Fe3+/Fe2+ and Ce4+/Ce3+ pairs accelerate the production of hydroxyl radicals (˙OH). The progressive decline of the 278 nm peak after 50 minutes reflects the oxidation of CB, while the gradual shift and disappearance of π → π* aromatic transitions indicate cleavage of the C–Cl bonds.

After 60 min, depletion at the absorption band 280 nm (n → π* transitions) shows oxidation of the 2,4-DCP.50 Band shift presents π → π* transitions of the aromatic ring disappearing slowly, which presents the breakage of C–Cl bonds. The Fourier-transform infrared (FTIR) spectrum of the calcined Fe3O4/CeO2 nanocomposite exhibits a broad absorption band around 3437 cm−1, which is attributed to surface hydroxyl (–OH) groups. The band observed at approximately 1549 cm−1 is associated with adsorbed water molecules. A distinct peak at 1079 cm−1 corresponds to Fe–Ce–O vibrational modes, indicating the interaction between Fe3O4 and CeO2 within the composite structure. Furthermore, the absorption region between 450–700 cm−1 is assigned to Ce–O stretching vibrations; the broad nature of this band arises from the overlap of Ce–O and Fe–O vibrations, confirming the successful integration of Fe3O4 onto the CeO2 surface. These observations validate the formation of the magnetic Fe3O4/CeO2 nanocomposite. After the adsorption-assisted removal of 2,4-DCP, noticeable changes in the FTIR spectrum were observed (Fig. S4b). The appearance of a new band near 1650 cm−1, along with the broadening of the hydroxyl band around 3400 cm−1, indicates the presence of adsorbed intermediates and increased surface hydroxyl interactions. Additionally, the emergence of a peak at approximately 743 cm−1 is attributed to C–Cl vibrations, suggesting the interaction of 2,4-DCP or its transformation products with the catalyst surface. These spectral changes confirm the involvement of surface functional groups during the catalytic adsorption-assisted removal.34,35 After the catalytic reaction with CB, the FTIR spectrum displays a more pronounced –OH stretching band near 3400 cm−1 and the appearance of a new band at around 746 cm−1, which is characteristic of H–O–H and C–Cl related vibrations, indicating the participation of CB molecules and subsequent structural transformation during adsorption-assisted removal as shown in Fig. S2. HPLC analysis further confirmed the removal of the parent 2,4-DCP and CB (Fig. S5 and S6 in SI) peaks following treatment with Fe3O4/CeO2, together with the appearance of additional chromatographic peaks attributable to possible transformation products. These results support the proposed adsorption-assisted oxidative transformation pathway.

The adsorption-assisted removal efficiency of CB and 2,4-DCP decreased significantly in the presence of scavengers, indicating the crucial role of reactive oxygen species in parity with HPLC results. The addition of IPA resulted in the most pronounced suppression of adsorption-assisted removal efficiency (reduction from ∼100% to ∼45–50%), suggesting that hydroxyl radicals (˙OH) are the primary active species governing the adsorption-assisted removal process. In the presence of benzoquinone, the degradation efficiency decreased to ∼60–65%, confirming the involvement of superoxide radicals (O2˙−) as secondary reactive species. A comparatively smaller reduction (∼75–80%) was observed with EDTA, indicating a limited contribution of surface-generated holes (h+) as shown in Fig. 8b. These results demonstrate that the adsorption-assisted removal mechanism is predominantly driven by ˙OH radicals, with additional contributions from O2˙− species. The generation of these reactive species can be attributed to the synergistic interaction between Fe3O4 and CeO2. The redox cycling between Fe2+/Fe3+ and Ce3+/Ce4+ facilitates electron transfer. This behaviour strongly supports a stepwise oxidative adsorption-assisted removal pathway, as shown in Fig. S7. Magnetic recovery allows efficient reutilization (after washing and drying) without structural collapse.51

Fe3O4/CeO2 nanocomposite is a highly stable composite with ∼100% adsorption-assisted removal efficiency for multiple organic pollutants.52 It has a combination of adsorption and catalytic synergy that boosts overall performance. Magnetic separation allows easy recovery and almost ∼98% reusability was found over repeated cycles. Recyclability tests were performed to assess catalyst stability and practice al applicability. After each adsorption-assisted removal cycle, the magnetic Fe3O4/CeO2 nanocomposite was rapidly separated using an external magnet, washed, and dried at 85 °C for 3 h before reuse. The catalyst maintained high activity over five consecutive cycles (Fig. 8c), demonstrating robust magnetic recoverability and operational durability suitable for repeated environmental applications. The post reaction XRD patterns have been compared with fresh Fe3O4/CeO2 nanocomposite. Almost all peaks are recovered after 5 cycle reaction of Fe3O4/CeO2 with 2,4-DCP and CB as shown in Fig. S10. It represents stability of nanocomposite and very less leaching of material. Synthesized Fe3O4/CeO2 nanocomposite is effectively applicable for the adsorption-assisted removal of bisphenol A (BPA) and crystal violet (CV) dye as shown in Fig. 8d. UV-vis based adsorption-assisted removal results have been shown in Fig. S6 and S7. Overall, Fe3O4/CeO2 has strong potential for industrial-scale wastewater treatment applications. Table S3 (in SI) shows a performance comparison of prepared magnetic nanocomposite Fe3O4/CeO2 for the degradation of CB and 2,4-DCP with the literature.

4. DFT and ML modelling insights

For DFT calculations in QuantumATK, Fe3O4/CeO2 nanocomposites were designed at 1 : 1, 1 : 2 and 2 : 1 ratios by selecting unit cells with the lowest lattice mismatch between the layers. The Fe3O4/CeO2 nanocomposites were constructed by interfacing Fe3O4 and CeO2 unit cells.

Fig. 9a–c illustrate the structure of the different ratios of Fe3O4/CeO2 nanocomposites. To know the ground-state stacking stability of these Fe3O4/CeO2 nanocomposites, we moved the Fe3O4 over the CeO2 unit cell on a uniform mesh, respectively, we were able to compute the binding energies of the various packing geometries. The interatomic distance between Fe3O4 and CeO2 in the different composites is 3.2 Å. To minimize the lattice mismatch, the mean absolute strain is applied on the second surfaces and is kept at 1% with a vacuum distance of 30 Å. The binding energies of the Fe3O4/CeO2 composites at 1 : 1, 1 : 2, and 2 : 1 ratio were calculated as −15.7078, −8.1174, and −7.7145 eV, respectively, as shown in Fig. 9d. The binding energy was computed by using eqn (17).

EBind = EFe3O4/CeO2 − EFe3O4 − ECeO2 17

Fig. 9. Structure of Fe3O4/CeO2 nanocomposites (a) 1 : 1 (b) 1 : 2 (c) 2 : 1 ratio (d) binding energy (e) projected DOS Fe3O4/CeO2 nanocomposites (1 : 1).

Fig. 9

Open in a new tab

The PDOS results (Fig. 9e) showed a clear hybridization between Fe 3d and O 2p states in the valence region, reflecting strong bonding interactions within the Fe3O4 component. A finite density of states at the Fermi level was observed, indicating the metallic character of the heterostructure and its suitability for efficient charge transport. In the conduction region, the electronic states were mainly contributed by Ce 4f orbitals, appearing as distinct peaks just above the Fermi level due to the localization effects introduced by the Hubbard U correction. Furthermore, the partial overlap among Fe d, O p, and Ce f states pointed to significant electronic interactions at the interface, which are expected to improve the electrochemical response of the heterostructure.

The negative binding energies confirm the structural stability of all three composite configurations. The predictive performance of the developed ANFIS model was evaluated using statistical error parameters, including SSE, MSE, RMSE and the coefficient of determination (R2). The obtained results are summarized in Table 2. The model demonstrated predictive agreement within the experimental domain represented by the available dataset.

Table 2. Statistical performance of the ANFIS model for the prediction of CB and 2,4-DCP degradation.

Pollutants SSE MSE RMSE R 2
CB 1.84 0.0048 0.069 0.991
2,4-DCP 2.12 0.0056 0.075 0.989
Open in a new tab

For CB, the model produced very low prediction errors (RMSE = 0.069) with a high correlation coefficient (R2 = 0.991). Similarly, for 2,4-DCP the ANFIS model achieved an RMSE of 0.075 and R2 of 0.989. These results confirm that the developed ANFIS model can accurately predict the adsorption-assisted removal efficiency under varying operational conditions such as pH, catalyst dosage, pollutant concentration and reaction time. The very high R2 values indicate a strong agreement between experimental and predicted values, which is also evident from the parity plot.

5. Conclusion

A novel Fe3O4/CeO2 nanocomposite was successfully synthesized and demonstrated exceptional efficiency for the adsorption-assisted removal of chlorobenzene and 2,4-DCP under mild conditions. The composite's high surface area, strong magnetic recoverability and synergistic redox activity enabled ∼100% pollutant removal and excellent stability across repeated cycles. These findings showcase the nanocomposite as a promising material for scalable, high-performance environmental remediation technologies. Kinetic analysis revealed that the adsorption-assisted removal process follows pseudo-first-order behaviour for R2 = 0.99. Langmuir–Hinshelwood model confirmed that the reaction proceeds through a surface-mediated catalytic mechanism. Numerical simulations using the Runge–Kutta method showed excellent agreement with experimental results, validating the reliability of the kinetic parameters. Furthermore, the ANFIS machine learning model successfully predicted the adsorption-assisted removal performance with high accuracy, demonstrating the potential of integrating artificial intelligence with nanomaterial-based remediation strategies. Further research can include pilot-scale continuous-flow reactor evaluations. This study can be extended to other emerging contaminants (e.g., pharmaceuticals, pesticides, dyes and other organic pollutants).

Author contributions

Monika Dubey: conceptualization, investigation, methodology, writing – original draft, review and editing Sunita: Simulation DFT, Monika Joshi: resources and review, Piyush Kuchhal: resources, review writing and editing, Ranjit Kumar: conceptualization, review writing and editing.

Conflicts of interest

There are no conflicts to declare.

Supplementary Material

RA-OLF-D6RA05504D-s001

Acknowledgments

This study is supported by the Visvesvaraya PDF Scheme for Electronics and IT, implemented by the Digital India Corporation (DIC), Ministry of Electronics and Information Technology (MeitY), Government of India (Implementation Order No. PHD-01/PDF-2025/93). Authors gratefully acknowledge the UPES CIC institutional Material Synthesis, characterization and analysis facilities and technical support that enabled the successful completion of this research work. During the preparation of this work, the authors used Microsoft Copilot to assist in the visual enhancement of the graphical abstract. In compliance with journal guidelines, the authors confirm that the tool utilizes fully licensed datasets and its licensing terms permit commercial reuse. After utilizing the tool, the authors reviewed, edited, and verified the visual and scientific accuracy of the graphic, and take full responsibility for the final content.

Data availability

The authors declare that all 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.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d6ra05504d.

References

  1. Nikolaivits E. et al., Degradation mechanism of 2,4-dichlorophenol by fungi isolated from marine invertebrates. Int. J. Mol. Sci. 2020;21:3317. doi: 10.3390/ijms21093317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Sun Z. et al., Mechanical enhancement of carbon fiber-reinforced polymers: From interfacial regulating strategies to advanced processing technologies. Prog. Mater. Sci. 2024;142:101221. doi: 10.1016/j.pmatsci.2023.101221. [DOI] [Google Scholar]
  3. Wang J. et al., Effect of weak electrical stimulation on m-dichlorobenzene biodegradation in biotrickling filters: Insights from performance and microbial community analysis. Bioresour. Technol. 2023;390:129881. doi: 10.1016/j.biortech.2023.129881. [DOI] [PubMed] [Google Scholar]
  4. Liu N. et al., Study on Gaseous Chlorobenzene Treatment by a Bio-Trickling Filter: Degradation Mechanism and Microbial Community. Processes. 2022;10:1483. doi: 10.3390/pr10081483. [DOI] [Google Scholar]
  5. Mirabediny M. et al., Fast degradation of chlorinated volatile organic compounds using soluble cobalt complexes. Process Saf. Environ. Prot. 2025;200:107356. doi: 10.1016/j.psep.2025.107356. [DOI] [Google Scholar]
  6. Zhang Y. Chen R. Riviere J. E. Comer J. Extraction of chlorobenzenes and pcbs from water by zno nanoparticles. Processes. 2021;9:1764. doi: 10.3390/pr9101764. [DOI] [Google Scholar]
  7. Mathur J. et al., Nanomaterials for Water Remediation: An Efficient Strategy for Prevention of Metal(loid) Hazard. Water. 2022;14:3998. doi: 10.3390/w14243998. [DOI] [Google Scholar]
  8. Qu M. et al., Mechanism of chlorobenzene removal in biotrickling filter enhanced by non-thermal plasma: Insights from biodiversity and functional gene perspectives. Bioresour. Technol. 2025;418:131931. doi: 10.1016/j.biortech.2024.131931. [DOI] [PubMed] [Google Scholar]
  9. Dutschke M. Schnabel T. Schütz F. Springer C. Degradation of chlorinated volatile organic compounds from contaminated ground water using a carrier-bound TiO2/UV/O3-system. J. Environ. Manage. 2022;304:114236. doi: 10.1016/j.jenvman.2021.114236. [DOI] [PubMed] [Google Scholar]
  10. Feng Y. C. et al., Remediation of groundwater contaminated by chlorinated organic compounds using alkaline-activated persulfate. Sustainable Environ. Res. 2025;35:13. doi: 10.1186/s42834-025-00249-x. [DOI] [Google Scholar]
  11. Zhang S. et al., Superior performance and mechanism of chlorobenzene degradation by a novel bacterium. RSC Adv. 2019;9:15004–15012. doi: 10.1039/C9RA01229J. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Sillanpää M. Ncibi M. C. Matilainen A. Advanced oxidation processes for the removal of natural organic matter from drinking water sources: A comprehensive review. J. Environ. Manage. 2018;208:56–76. doi: 10.1016/j.jenvman.2017.12.009. [DOI] [PubMed] [Google Scholar]
  13. Trueba-Santiso A. Palau J. Soder-Walz J. M. Vicent T. Marco-Urrea E. Assessment of aerobic biodegradation of lower-chlorinated benzenes in contaminated groundwater using field-derived microcosms and compound-specific carbon isotope fractionation. J. Environ. Sci. 2022;118:204–213. doi: 10.1016/j.jes.2021.12.025. [DOI] [PubMed] [Google Scholar]
  14. Qian Y. Zhang J. Zhang Y. Chen J. Zhou X. Degradation of 2,4-dichlorophenol by nanoscale calcium peroxide: Implication for groundwater remediation. Sep. Purif. Technol. 2016;166:222–229. doi: 10.1016/j.seppur.2016.04.010. [DOI] [Google Scholar]
  15. Falciglia P. P. et al., Glycerol-enhanced microwave heating for ultra-rapid effective remediation of marine sediments highly contaminated with hydrocarbons. Sep. Purif. Technol. 2017;189:11–19. doi: 10.1016/j.seppur.2017.07.066. [DOI] [Google Scholar]
  16. Li X. et al., Preparation of 3D PbO2 nanospheres@SnO2 nanowires/Ti electrode and its application in methyl orange degradation. Electrochim. Acta. 2014;146:15–22. doi: 10.1016/j.electacta.2014.08.150. [DOI] [Google Scholar]
  17. Pandis P. K. et al., Key Points of Advanced Oxidation Processes (AOPs) for Wastewater, Organic Pollutants and Pharmaceutical Waste Treatment: A Mini Review. ChemEngineering. 2022;6:8. doi: 10.3390/chemengineering6010008. [DOI] [Google Scholar]
  18. Katowah D. F. Mohammed G. I. Al-Eryani D. A. Sobahi T. R. Hussein M. A. Rapid and sensitive electrochemical sensor of cross-linked polyaniline/oxidized carbon nanomaterials core-shell nanocomposites for determination of 2,4-dichlorophenol. PLoS One. 2020;15:e0234815. doi: 10.1371/journal.pone.0234815. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Tkáč M. et al., Effect of ferrite content on the soft magnetic properties enhancement of Fe-based composite materials. J. Magn. Magn. Mater. 2026;645:173955. doi: 10.1016/j.jmmm.2026.173955. [DOI] [Google Scholar]
  20. Kumar R. Park J. G. van der Waals antiferromagnets: From early discoveries to future directions in the 2D limit. J. Magn. Magn. Mater. 2026;645:173977. doi: 10.1016/j.jmmm.2026.173977. [DOI] [Google Scholar]
  21. Yan X. Zhang A. Gao M. Zeng S. CeO2/CuO catalysts using different template agent for preferential CO oxidation in H2-rich stream. J. Rare Earths. 2017;35:1216–1220. doi: 10.1016/j.jre.2017.05.015. [DOI] [Google Scholar]
  22. Zhou M. Zhang K. Chen F. Chen Z. Synthesis of biomimetic cerium oxide by bean sprouts bio-template and its photocatalytic performance. J. Rare Earths. 2016;34:683–688. doi: 10.1016/S1002-0721(16)60079-0. [DOI] [Google Scholar]
  23. Voskanyan A. A. Chan K. Y. Kilogram-scale fabrication of 3D CeO2 active catalytic support with tailored 12 nm spherical mesopores via colloidal solution combustion synthesis. Microporous Mesoporous Mater. 2019;286:182–186. doi: 10.1016/j.micromeso.2019.04.015. [DOI] [Google Scholar]
  24. Lee S. K. Jo C. Kim J. Ryoo R. Soft-to-hard consecutive templating one-pot route from metal nitrate/phenol resin/surfactant to mesoporous metal oxides with enhanced thermal stability. Microporous Mesoporous Mater. 2020;293:109767. doi: 10.1016/j.micromeso.2019.109767. [DOI] [Google Scholar]
  25. Dubey M. Wadhwa S. Mathur A. Kumar R. Progress in mesoporous ceria: A review on synthesis strategies and catalytic applications. Appl. Surf. Sci. Adv. 2022;12:100340. doi: 10.1016/j.apsadv.2022.100340. [DOI] [Google Scholar]
  26. Nazeer S. et al., Mesoporous ceria-supported vanadia catalysts for selective aerobic oxidation of ethylbenzene. New J. Chem. 2024;48:10391–10400. doi: 10.1039/D4NJ00529E. [DOI] [Google Scholar]
  27. Bbumba S. Karume I. Comparative analysis of ANFIS, ANN, and BBD for enhanced prediction of methyl orange adsorption in water treatment. Sci. Rep. 2026;16:12822. doi: 10.1038/s41598-026-43445-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Sadek M. S. Khedr G. E. Messih M. F. A. Ismail M. A. H. Experimental and DFT study of photocatalytic activity of reduced graphene oxide/copper sulfide composite for removal of organic dyes from water. Sci. Rep. 2023;13:15636. doi: 10.1038/s41598-023-42680-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Dubey M. Joshi M. Thota C. Kumar R. Novel template free synthesis of high surface area mesoporous ceria and adaptive neuro fuzzy interference system (ANFIS) modelling for Cr(VI) adsorption. Mater. Sci. Eng., B. 2023;298:116819. doi: 10.1016/j.mseb.2023.116819. [DOI] [Google Scholar]
  30. Poddar M. K. Arjmand M. Sundararaj U. Moholkar V. S. Ultrasound-assisted synthesis and characterization of magnetite nanoparticles and poly(methyl methacrylate)/magnetite nanocomposites. Ultrason. Sonochem. 2018;43:38–51. doi: 10.1016/j.ultsonch.2017.12.035. [DOI] [PubMed] [Google Scholar]
  31. Khatamian M. Divband B. Shahi R. Ultrasound assisted co-precipitation synthesis of Fe3O4/bentonite nanocomposite: Performance for nitrate, BOD and COD water treatment. J. Water Process Eng. 2019;31:100870. doi: 10.1016/j.jwpe.2019.100870. [DOI] [Google Scholar]
  32. Aliakbari A. Amiri P. Salehi H. First-principles investigation of the structural and dynamical stability, electronic and thermal properties of two-dimensional Yn+1Cn (n = 1, 2, and 3) MXenes. FlatChem. 2022;31:100328. doi: 10.1016/j.flatc.2021.100328. [DOI] [Google Scholar]
  33. Çimen Mesutoğlu Ö. Predictive modeling of MB adsorption on activated olive stone through artificial neural networks. Sci. Rep. 2025;15:25084. doi: 10.1038/s41598-025-90143-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Dubey M. et al., Algae-synthesized cerium oxide nanoparticles for antibiotic degradation in water and subsequent bioenergy production. 3 Biotech. 2024;14:318. doi: 10.1007/s13205-024-04134-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mitra S. Sarkar A. Sen S. Removal of chromium from industrial effluents using nanotechnology: a review. Nanotechnol. Environ. Eng. 2017;2:1–14. [Google Scholar]
  36. Vasanth Kumar K. Porkodi K. Selvaganapathi A. Constrain in solving Langmuir-Hinshelwood kinetic expression for the photocatalytic degradation of Auramine O aqueous solutions by ZnO catalyst. Dyes Pigm. 2007;75:246–249. doi: 10.1016/j.dyepig.2006.05.035. [DOI] [Google Scholar]
  37. Dubey M. Singh A. K. Saxena V. Alginate embedded 1-D nanocomposite for environmental remediation: AI modeling, kinetics and thermodynamics. Nano Trends. 2026;13:100175. doi: 10.1016/j.nwnano.2025.100175. [DOI] [Google Scholar]
  38. Moreno de la Calle Á. Vizcaíno A. J. Carrero A. Calles J. A. Megía P. J. Study of Mesostructured CeO2 Synthesis via Nanocasting Using SBA-15 as a Template: Influence of the Cerium Precursor. Int. J. Mol. Sci. 2024;25:13016. doi: 10.3390/ijms252313016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ilyas S. Heryanto Abdullah B. Tahir D. X-ray diffraction analysis of nanocomposite Fe3O4/activated carbon by Williamson–Hall and size-strain plot methods. Nano-Struct. Nano-Objects. 2019;20:100396. doi: 10.1016/j.nanoso.2019.100396. [DOI] [Google Scholar]
  40. Dubey M. Joshi M. Thota C. Kumar R. Novel template free synthesis of high surface area mesoporous ceria and adaptive neuro fuzzy interference system (ANFIS) modelling for Cr(VI) adsorption. Mater. Sci. Eng., B. 2023;298:116819. doi: 10.1016/j.mseb.2023.116819. [DOI] [Google Scholar]
  41. Yang Z. et al., Mesoporous CeO2 hollow spheres prepared by ostwald ripening and their environmental applications. Eur. J. Inorg. Chem. 2010:3354–3359. doi: 10.1002/ejic.201000030. [DOI] [Google Scholar]
  42. Calvache-Muñoz J. Prado F. A. Rodríguez-Páez J. E. Cerium oxide nanoparticles: Synthesis, characterization and tentative mechanism of particle formation. Colloids Surf., A. 2017;529:146–159. doi: 10.1016/j.colsurfa.2017.05.059. [DOI] [Google Scholar]
  43. Mohamed S. K. Hegazy S. H. Abdelwahab N. A. Ramadan A. M. Coupled adsorption-photocatalytic degradation of crystal violet under sunlight using chemically synthesized grafted sodium alginate/ZnO/graphene oxide composite. Int. J. Biol. Macromol. 2018;108:1185–1198. doi: 10.1016/j.ijbiomac.2017.11.028. [DOI] [PubMed] [Google Scholar]
  44. Hernández S. and Ormsbee L. E., Evaluation of Different Nanoparticle Support Systems Applied in the Detoxification of Chlorinated Organic Compounds Chloro-Organic Degradation by Polymer Membrane Immobilized Iron-Based Particle Systems View Project Translational Graphene Research Program View Project, 2015 [Google Scholar]
  45. Moodley K. G. Chetty D. K. Ramphal S. R. Gericke G. A rapid method for determining chlorobenzenes in dam water systems. Water SA. 2013;39(1):23–30. [Google Scholar]
  46. Bekhterev V. N. Kabina E. A. Loginova S. A. Removal of chloromethanes from water by the method of vapor-phase extraction. J. Water Chem. Technol. 2014;36:134–138. doi: 10.3103/S1063455X14030060. [DOI] [Google Scholar]
  47. Zhang J. Lei J. Ju H. Wang C. Electrochemical sensor based on chlorohemin modified molecularly imprinted microgel for determination of 2,4-dichlorophenol. Anal. Chim. Acta. 2013:2–7. doi: 10.1016/j.aca.2013.05.012. [DOI] [PubMed] [Google Scholar]
  48. Kartal N. Sökmen M. New catalyst systems for the degradation of chlorophenols. Desalination. 2011;281:209–214. doi: 10.1016/j.desal.2011.07.066. [DOI] [Google Scholar]
  49. Czaplicka M. Sources and transformations of chlorophenols in the natural environment. Sci. Total Environ. 2004;322:21–39. doi: 10.1016/j.scitotenv.2003.09.015. [DOI] [PubMed] [Google Scholar]
  50. Castañeda C. Tzompantzi F. Gómez R. Rojas H. Enhanced photocatalytic degradation of 4-chlorophenol and 2,4-dichlorophenol on in situ phosphated sol-gel TiO2. J. Chem. Technol. Biotechnol. 2016;91:2170–2178. doi: 10.1002/jctb.4943. [DOI] [Google Scholar]
  51. Mishra P. K. Kumar R. Rai P. K. Surfactant-free one-pot synthesis of CeO2, TiO2 and Ti@Ce oxide nanoparticles for the ultrafast removal of Cr(vi) from aqueous media. Nanoscale. 2018;10:7257–7269. doi: 10.1039/C7NR09563E. [DOI] [PubMed] [Google Scholar]
  52. Xu L. Wang J. Magnetic nanoscaled Fe3O4/CeO2 composite as an efficient fenton-like heterogeneous catalyst for degradation of 4-chlorophenol. Environ. Sci. Technol. 2012;46:10145–10153. doi: 10.1021/es300303f. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

RA-OLF-D6RA05504D-s001
RA-OLF-D6RA05504D-s001.pdf (1.5MB, pdf)

Data Availability Statement

The authors declare that all 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.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d6ra05504d.

Articles from RSC Advances are provided here courtesy of Royal Society of Chemistry

RESOURCES