Skip to main content
NCBI home page
RSC Advances logoLink to RSC Advances
. 2026 Jul 6;16(36):37797–37815. doi: 10.1039/d6ra01341d

Synthesis of ZnO, Fe3O4, and ZnFe2O4 nanoparticles/activated carbon derived from Baccaurea ramiflora L. fruit peel for co-adsorption of sulfamethoxazole and alizarin red S

Sang Thi Tuyet Nguyen a,b,c, Thuy Thi Thanh Nguyen c, Thuan Van Tran a,b,✉
PMCID: PMC13334721  PMID: 42441256

Abstract

This study reports the synthesis of ZnO/BAC, Fe3O4/BAC, and ZnFe2O4/BAC composites using activated carbon (BAC) derived from Baccaurea ramiflora (L.) fruit peel. The materials were characterized by XRD, FT-IR, SEM, TEM, and the point of zero charge. Batch adsorption experiments evaluated the effect of contact time, initial concentration, temperature, adsorbent dosage, initial solution pH, and ionic strength on the removal of alizarin red S (ARS) and sulfamethoxazole (SMX) in the single-solute and binary systems. At a contact time of 120 min, a pollutant concentration of 10 mg L−1, a dosage of 0.25 g L−1 and a temperature of 25 °C, ZnFe2O4/BAC exhibited the best adsorption efficiency at 88.4% and 99% for ARS and 96% and 86% for SMX in single and binary solutions, respectively. ARS and SMX adsorptions in single and binary solutions were fitted with the pseudo second-order and Bangham kinetic models. At 25 °C, ARS and SMX adsorptions were fitted with Langmuir and Freundlich isotherm models, respectively. Regeneration tests identified HCl as the most effective desorption agent, enabling the reuse of ZnFe2O4/BAC for four cycles. The adsorption mechanisms involve hydrogen bonding, pore-filling, electrostatic interactions, and π–π interactions. At 25 °C, maximum adsorption capacities obtained from the Langmuir equation were arranged as follows: ARS (binary, 50.2 mg g−1) < ARS (single, 51.3 mg g−1) < SMX (binary, 67.6 mg g−1) < SMX (single, 69.3 mg g−1). Thus, ZnFe2O4/BAC is a promising recyclable material for the removal of ARS and SMX in both single and binary solutions.

Here, we synthesized a series of carbon-based adsorbents derived from Baccaurea ramiflora (L.) peel for removal of sulfamethoxazole and alizarin red S from water. ZnFe2O4/activated carbon exhibited high adsorption performance and reusability.graphic file with name d6ra01341d-ga.jpg

1. Introduction

In recent years, the rapid expansion of the textile and healthcare sectors has led to a significant increase in the release of dyes and antibiotics into aquatic environments.1,2 Among antibiotics, sulfamethoxazole (SMX, a sulfonamide antibiotic), is of major concern because of its persistence and potential ecotoxicological effects.3 With high water solubility (Table S1) and high mobility, antibiotics including SMX can be detected in water sources such as surface water and groundwater.4 Meanwhile, alizarin red S (ARS, an anthraquinone-based dye) can pose potential threats to aquatic ecosystems because the presence of this dye in water reduces the effectiveness of solar light, affecting the photosynthetic processes and metabolisms of aquatic species.5 As a result, removal of sulfamethoxazole antibiotic and alizarin red S dye in water is required.

Adsorption has attracted considerable attention because it is simple to operate, high efficient, and reusability.6 To enhance adsorption performance, adsorbents often possess high BET surface area (SBET, BET = Brunauer–Emmett–Teller) and active functional groups.7 For example, Wu et al. reported a dual-functional Fe-modified ZIF-8 (Fe-ZIF-8) for phosphorus adsorption.8 Although Fe-ZIF-8 obtained a SBET of up to 1249 m2 g−1, high pore volume (0.52 cm3 g−1), and adsorption capacity of 10.7 mg g−1, this adsorbent remains expensive. Among adsorbents, activated carbon (AC) is widely used as a low cost adsorbent with large surface area and well-developed porous structure.9 Zinc ferrite (ZnFe2O4) is a magnetic spinel-type ferrite, and owns thermal stability and magnetic properties. Incorporating ZnFe2O4 into AC can combine high adsorption capacity with magnetic recoverability, facilitating separation and reuse.10,11

Recently, there has been growing interest in developing low-cost and environmentally friendly adsorbents derived from biomass residual and agricultural by-products.12Baccaurea ramiflora L. (common name: Burmese grape) is often cultivated in India and Southeast Asia, with valuable chemical constituents and pharmacological activities.13,14 Rohilla revealed that more than 50% of Baccaurea ramiflora fruit (including seeds, peels, etc.) is non-edible.13 As a result, a majority of the fruit mass and can be discarded during processing. With high cellulose (40%), hemicellulose (11.56%), and pectin (38%) contents, this biowaste can be promising for carbon materials.15

A waste-to-wealth strategy by valorizing Baccaurea ramiflora fruit peel can both reduce environmental management burdens and provide a sustainable precursor to produce adsorbents. For example, Roy et al. converted Baccaurea ramiflora (Burmese grape) seed into biochar via H3PO4 activation and subsequent KOH modification.16 These authors reported that the activated biochar obtained a SBET of 19.9 m2 g−1 and methylene blue dye adsorption capacity of 166.3 mg g−1. Van Phuoc et al. also improved the SBET (up to 478.5 m2 g−1) of activated carbon derived from the Baccaurea ramiflora waste by microwave-assisted NaOH activation.17 They applied response surface methodology for optimization of ciprofloxacin removal with an adsorption capacity of 191.33 mg g−1. However, very few studies explored the use of Baccaurea ramiflora-based adsorbents for water treatment. Moreover, adsorption capacities obtained from these adsorbents remained low. While co-adsorption of contaminants was not investigated, uptake mechanisms were suggested without evidences of interactions before and after adsorption. To our best knowledge, no investigations of composites based on metal oxides, ferrites supported on activated carbon Baccaurea ramiflora waste for simultaneous removal of pharmaceutical and organic dye contaminants were previously reported.

Here, ZnO/BAC, Fe3O4/BAC, and ZnFe2O4/BAC composites were synthesized using Baccaurea ramiflora fruit peel as the carbon source. The materials were evaluated for the simultaneous adsorption of SMX and ARS in single-solute and binary systems. Physicochemical properties, adsorption performance, kinetics, isotherms, regeneration, and plausible adsorption mechanisms were systematically investigated to assess their potential for water treatment. The significance of this study is to develop novel recyclable and efficient composites via the waste-to-wealth approach to solve a dual problem: agricultural waste and water pollution.

2. Materials and methods

2.1. Materials and chemical

Baccaurea ramiflora fruits were collected at Thu Duc Market, Ho Chi Minh City, Vietnam. The peels were then manually separated from the fruits. Sulfamethoxazole (SMX, C10H11N3O3S, purity ≥ 99%), and alizarin red S (ARS, C14H7NaO7S, purity ≥ 99%) were purchased from Sigma-Aldrich (Germany). Zinc chloride (ZnCl2, purity ≥97%) was purchased from Fisher Scientific (UK). Iron(iii) chloride hexahydrate (FeCl3·6H2O, purity ≥ 99.5%) was purchased from Guangdong Chemical Reagent Engineering (China). Absolute ethanol (C2H5OH, 99.5%) was purchased from ChemSol (Vietnam). Other reagents included hydrochloric acid (HCl, 36–38%), potassium hydroxide (KOH, ≥ 85%), sodium hydroxide (NaOH, ≥ 96%) were purchased from Xilong Scientific Co Ltd (China).

2.2. Preparation of adsorbents

2.2.1. Biochar from BP from Baccaurea ramiflora fruit

The Baccaurea ramiflora fruit peels were washed thoroughly with water, cut into small pieces, and sun-dried to reduce moisture content. Subsequently, the peels were oven-dried at 80 °C until they completely dried to obtain raw adsorbent (BD). The dried peels were then pyrolyzed at 500 °C for 3 h, ground into a fine powder to obtain biochar (BBC).

2.2.2. Synthesis of composites

ZnCl2 and FeCl3·6H2O were used as both activating agent and precursor to synthesize ZnO/BAC, Fe3O4/BAC and ZnFe2O4/BAC (Fig. 1). For ZnO/BAC, 0.5 g of BBC and 0.5 g of ZnCl2 were dissolved in 10 mL of deionized water. For Fe3O4/BAC, 0.5 g of BBC and 0.5 g FeCl3·6H2O were dissolved in 10 mL of deionized water. For ZnFe2O4/BAC, 0.5 g BBC, 0.5 g ZnCl2 and 1.9 g FeCl3·6H2O were dissolved in 20 mL of deionized water. The mixtures were stirred thoroughly for 30 min and then dried overnight at 80 °C. The dried samples were calcined under a nitrogen atmosphere at 600 °C for 3 h (heating rate: 5 °C min−1). After calcination, the samples were washed with deionized water (3 × 100 mL) and centrifuged, then dried at 120 °C for 2 h. For comparison, BBC was activated with KH2PO4, melamine, and tannic acid (1 : 1, by weight) using the same carbonization procedure, forming different activated carbons, labeled as BP, BM, and BTA, respectively.

Fig. 1. Schematic synthesis of the BP-derived biochar and ZnFe2O4/BAC material.

Fig. 1

Open in a new tab

2.3. Adsorption study

Batch adsorption experiments were conducted to investigate the efficiency of ZnO/BAC, Fe3O4/BAC, and ZnFe2O4/BAC in the removals of ARS and SMX. In a typical run, 0.005 g of adsorbent was added to 20 mL of single solution ARS or SMX or binary (10 mg L−1) and shaken at 200 rpm, and the residual concentrations were determined by UV-vis spectrophotometer (Cary 60 Agilent, Agilent Technologies Inc., Santa Clara, California, USA). The maximum wavelength (λmax), and concentration of ARS and SMX in single and binary systems were described in the SI document. The effect of contact time (0–150 min), initial concentration (10–75 mg L−1), temperature (25–45 °C), adsorbent dosage (0.1–1.5 g L−1), and pH (4–10) were examined. Ionic effects were assessed by adding selected cations and anions (Na+, K+, NH4+, Mg2+, Al3+, Cl−, H2PO4−, CO32−, HCO32−, SO42−, PO43−) at 0.01 M. The adsorption efficiency (eqn (1)) and adsorption capacity (eqn (2)), which were calculated using the following equations.

2.3. 1
2.3. 2

where H% is the adsorption efficiency of ARS or SMX, Q (mg g−1) is the adsorption capacity of ARS or SMX, C0 (mg L−1) is the initial concentration, Ce (mg L−1) is the equilibrium concentration, m (g) is the mass of the absorbent, and V (L) is the solution volume.

2.4. Regeneration and reusability

In the regeneration of ZnFe2O4/BAC, four solvents including 0.1 M HCl, 0.1 M NaOH, 0.1 M KOH, and absolute ethanol were used as desorbing eluents. Typically, 0.02 g of ARS-, SMX-, or ARS or SMX-laden ZnFe2O4/BAC was added into 5 mL each of solvent, and shaken for 60 min for desorption. This procedure was repeated three times. The neutralization was required for washing with 0.1 M HCl, 0.1 M NaOH, and 0.1 M KOH. Finally, the adsorbent was dried before use.

The regenerated ZnFe2O4/BAC was used as a fresh adsorbent for adsorption of ARS, SMX, and mixed ARS and SMX. 0.01 g of regenerated ZnFe2O4/BAC was added into ARS (or SMX or mixed solution). After adsorption occurred for 120 min, 2 mL of the solution was extracted for determination of concentration using the UV-vis spectrophotometer. When the best solvent was selected, the reusability was investigated through four successive adsorption cycles. Each experiment was repeated three times to assess removal efficiency.

3. Results and discussion

3.1. Characterization of BP-derived adsorbents

3.1.1. X-ray diffraction

Fig. 2 shows the XRD patterns of ZnO/BAC, Fe3O4/BAC and ZnFe2O4/BAC. Specifically, the diffraction peaks of crystalline ZnO at 2θ ≈ 31.8°, 36.3°, and 47.5°, which can be indexed to the (100), (101), and (102) planes of ZnO, respectively (Fig. 2a). These observations are in good agreement with the standard JCPDS card No. 36-1451.18 The average crystallite size of ZnO in the ZnO/BAC composite was estimated at 19.51 nm using the Scherrer equation (eqn (3)). Sayed et al. reported a slightly smaller crystallite size of 14.93 nm for sawdust-derived ZnO@AC composite.19 However, the diffraction peaks appear relatively weak, indicating the low crystallinity of ZnO. This phenomenon can be attributed to the amorphous nature of AC that interferes with the metal oxide signals. Worasawat et al. reported the same pattern of ZnO nanoparticles synthesized via the hydrothermal method.20

D = Kλβ cos θ 3

where, D is the crystallite size (nm), K is the Scherrer's constant (0.9). λ is the wavelength of Cu Kα radiation (0.15406 nm), β is the full width at half maximum (FWHM) of the diffraction peak, and θ is the Bragg diffraction angle.

Fig. 2. XRD patterns of (a) ZnO/BAC, (b) Fe3O4/BAC, (c) ZnFe2O4/BAC.

Fig. 2

Open in a new tab

Fig. 2b shows that diffraction peaks at 2θ ≈ 30.1°, 35.5° and 41.3° which can be indexed to the (220), (311) and (400) planes of Fe3O4, respectively. This confirms the presence of a cubic spinel phase in accordance with JCPDS card No. 19-0629.21 The crystallite size of Fe3O4 was determined at 12.35 nm, which was very close to that (11.73 nm) of Fe3O4/SiO2/AC nanocomposite.22 Nevertheless, Fe3O4/BAC had the poor diffraction peak intensities can be ascribed to the amorphous structure of BAC that generates a broad background signal and partially masks the characteristic reflections of crystalline Fe3O4.23

Fig. 2c shows that diffraction peaks that match well with the standard data from JCPDS card No. 22-1012.24 The presence of ZnFe2O4 in the ZnFe2O4/BAC is most prominently evidenced by the diffraction peak at 2θ ≈ 35.5° corresponding to the (311) plane. This peak is recognized as the most characteristic reflection of the spinel structure.25 Additional moderate-intensity peaks are observed at 2θ ≈ 30.1°, 57.0° and 63.8°, corresponding to the (220), (511), and (440) planes, respectively. Compared with ZnO and Fe3O4, ZnFe2O4 had a larger crystallite size of 24.05 nm. The incorporation of Zn2+ and Fe3+ ions during synthesis might promote the formation of a well-ordered cubic spinel ZnFe2O4 phase with larger crystallites. This resulted in stronger Bragg reflections and reduced background noise from amorphous carbon.26 Similarly, Konicki et al. reported the well define XRD patterns of magnetic ZnFe2O4 nanocomposites synthesized via a hydrothermal method.27

3.1.2. Fourier transform infrared spectroscopy

Fig. 3 shows that the functional groups of BBC, ZnO/BAC, Fe3O4/BAC, and ZnFe2O4/BAC. Specifically, the broad peak in the range of 3100–3500 cm−1 is attributed to O–H stretching vibrations. The presence of O–H groups in all four materials indicates that hydroxyl groups originated from cellulose, hemicellulose, and lignin of the Baccaurea ramiflora fruit peels remained even after pyrolysis.28 Notably, the O–H bands of ZnO/BAC, Fe3O4/BAC and ZnFe2O4/BAC are attributable to the activation of BBC with ZnCl2, FeCl3·6H2O and mixture of ZnCl2 and FeCl3·6H2O. The peaks at 2800–2900 cm−1 (C–H) exhibit decreasing intensities from BBC to the modified materials, suggesting that hydrocarbon moieties were decomposed. The C Created by potrace 1.16, written by Peter Selinger 2001-2019 C stretching vibration at 1600 cm−1 corresponds to aromatic rings, indicating that the carbon structure was largely preserved during the BBC and composite syntheses. The C Created by potrace 1.16, written by Peter Selinger 2001-2019 O stretching vibration appears at 1425 cm−1.29 The C–O stretching vibrations in the range of 1050–1250 cm−1 are assigned to ether, phenol, and ester groups. The bands between 450–600 cm−1 are characteristic of metal–oxygen (M–O) bonds (Zn–O, Fe–O).30 These peaks are absent in the pristine biochar sample, indicating that modification was successful.

Fig. 3. FT-IR spectra of (a) BBC, (b) ZnO/BAC, (c) Fe3O4/BAC, (d) ZnFe2O4/BAC.

Fig. 3

Open in a new tab

3.1.3. Point of zero charge

Fig. 4 shows that pHpzc of ZnO/BAC, Fe3O4/BAC, ZnFe2O4/BAC. At pHsolution > pHpzc the material surface carries negative charge, the surface carries a positive charge at pHsolution < pHpzc.31 Here, the pHpzc was determined at 7.3 for ZnO/BAC. Adeoye et al. reported the same pHpzc (7.25) of ZnO-AC.32 Fe3O4/BAC had a pHpzc of 6.6, which was as same as pHpzc for Fe3O4/AC reported previously.33 Meanwhile, ZnFe2O4/BAC had a pHpzc of 6, which was lower than that of ZnO/BAC and Fe3O4/BAC. However, this result was higher than that of ZnFe2O4-based composites such as Mo2C-ZnFe2O4@PPNB (pHpzc of 5.8),34 and ZnFe2O4 nanoparticles (pHpzc of 6).35

Fig. 4. The pHpzc of (a) ZnO/BAC, (b) Fe3O4/BAC, (c) ZnFe2O4/BAC.

Fig. 4

Open in a new tab

3.1.4. Quantification of surface functional groups

Fig. 5 shows that the quantification of carboxylic, lactonic, phenolic, and basic functional groups on ZnFe2O4/BAC. Through Boehm titration, we found that this adsorbent possesses 0.82 ± 0.04 mmol g−1 carboxylic group, 0.35 ± 0.06 mmol g−1 lactone group, 0.08 ± 0.06 mmol g−1 phenolic group, and 0.03 ± 0.02 mmol g−1 basic group. The total basic group content was so far lower than the total acidic (oxygenated) group content. This result can be because activating agents are Lewis acids (ZnCl2, FeCl3). When they reacted with BBC, acidic active sites (Zn2+, Fe3+) attached on the surface of ZnFe2O4/BAC, resulting in more acidic groups. Djukic et al. also reported the higher amount of acidic functional groups than basic groups in mechanochemically activated carbon.36 Moreover, the total functional group content of ZnFe2O4/BAC in this study was higher than that of the AC sample synthesized previously.37

Fig. 5. Quantification of surface functional groups of ZnFe2O4/BAC by the Boehm titration.

Fig. 5

Open in a new tab

3.1.5. Morphological analysis

The morphological structure of ZnFe2O4/BAC was examined by SEM images (Fig. 6). Specifically, Fig. 6a–c shows that the carbon surface exhibits a porous and irregular texture, containing large voids. Similarly, Alam et al. reported that surface irregularity with large pores in the AC structure.38 ZnFe2O4 with rod-like shapes appear relatively aggregated and attached to the AC surface. Some ZnFe2O4 nanorods are dispersed on the AC matrix (Fig. 6d and e). These ZnFe2O4 nanorods exhibited less than 10 nm in width and up to 400 nm in length. Meanwhile, several studies observed a spherical and aggregated morphology of ZnFe2O4 particles.39,40 Here, AC might affect the formation of ZnFe2O4, resulting in the nanorod morphology.

Fig. 6. SEM (a–c) and TEM (d and e) images of (a) ZnFe2O4/BAC.

Fig. 6

Open in a new tab

3.2. Effect of factors on adsorption

3.2.1. Effect of the activating agent

Fig. 7 shows the removal efficiencies (H%) of ARS and SMX from single and binary solutions. Overall, BD and BBC exhibited very low removal percentages (0–8.5%) in all four adsorption modes. This is mainly attributed to their low surface area and the absence of surface functional groups. In contrast, BP, BM, and BTA exhibited minor improvements (0–23%). Chemical activation of BBC with KH2PO4, melamine, and tannic acid might generate additional porosity and surface sites.41 However, the ARS and SMX removal efficiencies by these adsorbents were still relatively low compared with ZnO/BAC, Fe3O4/BAC, and ZnFe2O4/BAC composites. Indeed, the composites obtained considerably higher removal percentages. ZnFe2O4/BAC activated with ZnCl2 and FeCl3 led to higher removal efficiencies (86–99%) compared to Fe3O4/BAC (66–84%) and ZnO/BAC (26–57%). The activation with ZnCl2 and/or FeCl3 provided additional surface-active sites, and enhanced pore development.

Fig. 7. Effect of the activation agents on adsorption efficiencies of (a) ARS (single), (b) SMX (single), (c) ARS (binary), and (d) SMX (binary). Experimental conditions: contact time = 120 min, C0 = 10 mg L−1, dosage = 0.25 g L−1, unadjusted pH, 25 °C. Note: BD: Baccaurea ramiflora peel raw adsorbent (un-activated adsorbent), BBC: Baccaurea ramiflora peel biochar, BP: BBC activated with KH2PO4, BM: BBC activated with melamine, and BTA: BBC activated with tannic acid, BAC: Baccaurea ramiflora activated carbon.

Fig. 7

Open in a new tab

3.2.2. Effects of contact time

Fig. 8 presents the adsorption efficiencies (H%) of ARS and SMX in single-solute and binary systems using ZnO/BAC, Fe3O4/BAC, and ZnFe2O4/BAC over contact times of 10–150 min. Overall, H% increased with increasing contact time, which can be attributed to the progressive diffusion of ARS and SMX molecules from the bulk solution to the adsorbent surface and into the pore network until equilibrium was reached.16 The difference in H% between 120 and 150 min was negligible, indicating that adsorption equilibrium was achieved at approximately 120 min. Machado et al. reported that ARS adsorption on multiwalled carbon nanotubes increased over 0–300 min and reached equilibrium at 100 min.42 Insuwan et al. also reported an optimal contact time of 120 min for ARS removal using Zeolite L.43 For SMX, Zhang et al. demonstrated that adsorption equilibrium on magnetic biochar derived from sugarcane bagasse impregnated with 300 mM FeCl3 and pyrolyzed at 800 °C was reached at 240 min.44

Fig. 8. UV-vis absorption spectra of (a) ARS and (b) SMX adsorption by ZnFe2O4/BAC. Effect of contact time on removal efficiencies of (c) ARS (single), (d) SMX (single), (e) ARS (binary), (f) SMX (binary). Experimental conditions: C0 = 10 mg L−1, dosage = 0.25 g L−1, unadjusted pH, 25 °C.

Fig. 8

Open in a new tab

Notably, ARS removal in the binary system was higher than in the single-solute system. At 120 min, ARS removal by ZnO/BAC increased from 36% (single) to 57% (binary), while Fe3O4/BAC increased from 66% to 84%. ZnFe2O4/BAC exhibited the highest performance, with ARS removal increasing from 88.4% (single) to 99.0% (binary). Rajapaksha et al. suggested that the coexistence of dyes and antibiotics can promote adsorption through electrostatic and π–π interactions, leading to synergistic binding effects.45 Budhiraja et al. also showed that the adsorption behavior of one contaminant can be substantially altered by the presence of co-contaminant.46 In their study, methylparaben enhanced triclosan adsorption on polyethylene and oxidatively degradable polyethylene, increasing the maximum adsorption capacity from 1.06 to 3.667 mg g−1 for polyethylene and from 1.506 to 4.421 mg g−1 for oxidatively degradable polyethylene. These findings support that co-solutes can improve adsorption performance in mixed systems. Here, 120 min was selected as the optimal contact time, and ZnFe2O4/BAC was chosen as the most effective adsorbent for subsequent experiments on the simultaneous removal of ARS and SMX.

3.2.3. Effects of concentration

Fig. 9 shows that the adsorption performance of ARS and SMX in both single and binary solutions by ZnFe2O4/BAC at initial concentrations of 10–75 mg L−1 and temperatures of 25 °C, 35 °C, and 45 °C. In general, the uptake efficiency decreased with increase initial solution concentration. Similar trends were reported for ARS adsorption on activated carbon/γ-Fe2O3 nano-composite at 50–150 mg L−1,47 and for SMX adsorption on core–shell activated carbon at 5–100 mg L−1.48 At low concentrations, ARS and SMX are present at lower molecular densities, and they can be captured by a large number of available adsorption sites in the ZnFe2O4/BAC pore structure. At higher concentrations, adsorption sites become progressively occupied, leading to lower removal efficiencies.49,50 The highest removal efficiencies were obtained at an initial concentration of 10 mg L−1 in both single-solute and binary systems. Under these conditions, ZnFe2O4/BAC achieved ARS removal efficiencies of 88.4% (single) and 99.0% (binary), and SMX removal efficiencies of 96.0% (single) and 86.0% (binary).

Fig. 9. Effect of concentration and temperature on adsorption efficiencies of (a) ARS (single), (b) SMX (single), (c) ARS (binary), and (d) SMX (binary). Experimental conditions: contact time = 120 min, dosage = 0.25 g L−1, unadjusted pH.

Fig. 9

Open in a new tab

3.2.4. Effects of temperature

The effect of temperature was evaluated at 25–45 °C (Fig. 9). Overall, H% increased slightly with increasing temperature, which may be attributed to enhanced surface activity and higher kinetic energy of the adsorbate molecules.50 This suggests adsorption is an endothermic process. However, adsorption between 25 °C and higher temperatures (35 °C and 45 °C) were not substantially different for either ARS or SMX. This may be due to weakened intermolecular interactions between the adsorbates and ZnFe2O4/BAC at elevated temperatures, which can reduce adsorption affinity and may adversely affect adsorption sites or the physical structure of the adsorbent.50 Khamis et al. similarly reported that ARS removal remained nearly constant over 25–50 °C.51 Consistently, Delpiano et al. reported favorable ARS adsorption by an iron-based metal–organic framework (Fe-BTC) was at 25 °C.52 Zeng et al. also identified 25 °C as an optimal condition for SMX adsorption using FeCl3-activated biochar derived from bermudagrass.53 The same trend was reported in previous works.44,47,49,51 Hence, an initial concentration of 10 mg L−1 and temperature of 25 °C were selected for subsequent experiments.

3.2.5. Effects of dosage

Fig. 10 illustrates the effect of ZnFe2O4/BAC dosage on the removal efficiency (H%) of ARS and SMX in both single-solute and binary systems over the range of 0.1–1.5 g L−1. It can be observed that H% increased proportionally with the adsorbent dosage. This is due to the greater number of available adsorption sites provided by a larger adsorbent mass.49 Khamis et al. reported that the removal efficiency of ARS increased from approximately 43% to approximately 93% with an increase in wool dosage from 1 to 8 g L−1.51 Zhao et al. also observed enhanced SMX adsorption (from 5.9% from 70.3%) when dosages of biochar derived from wheat straw-derived biochars and graphite increased dosages from 25 mg to 100 mg per 7 mL of SMX.54 In the present study, however, the improvement in removal efficiency became marginal when the dosage was higher than 0.25 g L−1. Fayazi et al. reported a comparable pattern for ARS adsorption of magnetic activated carbon nanocomposites.47 At higher adsorbent dosages, particle agglomeration may occur and consequently diminish the removal efficiency.55

Fig. 10. Effect of ZnFe2O4/BAC dosage on removal efficiencies of (a) ARS (single), (b) SMX (single), (c) ARS (binary), and (d) SMX (binary). Experimental conditions: contact time = 120 min, C0 = 10 mg L−1, unadjusted pH, 25 °C.

Fig. 10

Open in a new tab

At an adsorbent dosage of 0.25 g L−1, ZnFe2O4/BAC achieved maximum removal efficiencies (Hmax) for ARS of 88.4% (single) and 99.0% (binary), and for SMX of 96.0% (single) and 86.0% (binary). Machado et al. reported that ARS adsorption on single-walled and multi-walled carbon nanotubes reached maximum efficiencies at a higher dosage of 1.5 g L−1.42 Similarly, Wagh et al. reported an optimal dosage of 4 g L−1 for ARS removal using coconut shell-derived activated carbon.56 However, this dosage was substantially higher than that in the present work. For SMX, Zhang et al. reported an optimal dosage of 0.2 g L−1 using magnetic biochar derived from sugarcane bagasse activated with FeCl3 and pyrolyzed at 800 °C.44 Therefore, 0.25 g L−1 was selected as the optimal dosage in this study.

3.2.6. Effects of solution pH

Fig. 11a shows that the H% of ARS in single-component solution by ZnFe2O4/BAC at different pH values. As pH increased from 4 to 6, H% also increased and reached a peak of 93.50% at an optimal pH of 6. ARS predominantly exists in an anionic form (ARS−) due to sulfonate group and the disassociation of hydroxyl groups (pKa1: 5.82, pKa2: 10.78).57 The pHpzc of ZnFe2O4/BAC was found at 6.0 (Fig. 4). At pH < 6.0, the abundance of protons (H+) results in a positively charged ZnFe2O4/BAC surface (e.g., –OH2+, Fe–OH2+), thereby enhancing the electrostatic attraction between ARS− and positively charged ZnFe2O4/BAC. In contrast, at pH higher than 6.0, surface deprotonation increases the negative surface charge and enhances electrostatic repulsion between ARS− and negatively charged ZnFe2O4/BAC.49 The adsorption of ARS was therefore unfavored under pHs higher 6. Optimal ARS removal was evidenced at low pH values.42,47,51 In this work, ZnFe2O4/BAC exhibited the highest ARS removal at pH 6. But at this pH, ZnFe2O4/BAC surface was not charged, hence, electrostatic attraction was not formed between ARS− and ZnFe2O4/BAC. It was suggested that other interactions such as π–π donor–acceptor interactions and hydrogen bonding,58 and pore-filling59 might play vital roles.

Fig. 11. Effect of pH on removal efficiencies of (a) ARS (single), (b) SMX (single), (c) ARS (binary), and (d) SMX (binary). Experimental conditions: contact time = 120 min, C0 = 10 mg L−1, dosage = 0.25 g L−1, 25 °C.

Fig. 11

Open in a new tab

Fig. 11b shows the removal efficiency of SMX in a single-solute system as a function of pH. Similar to ARS, H% increased from 67.0% to 95.5% as pH increased from 4 to 6, and then dropped at pH 7–10. In SMX adsorption, pH affects not only the surface charge of the ZnFe2O4/BAC but also the aqueous speciation of SMX.48 Across a wide pH range, SMX can exist as cationic (SMX+), neutral (SMX0), and anionic (SMX−) species, governed mainly by protonation of the aniline N and deprotonation of the sulfonamide (CONH) group (pKa1: 1.7, pKa2: 5.6).60 At pH below pHpzc (6.0), ZnFe2O4/BAC is positively charged, while SMX is predominantly neutral. Thus, it is not possible to form electrostatic interaction. Nevertheless, appreciable adsorption observed can be due to other interactions such as hydrogen bonding or π–π interactions.16,61 At pH higher than pHpzc, adsorption decreases due to increased electrostatic repulsion between SMX− and negatively charged ZnFe2O4/BAC. This finding was in agreement with previous reports.44,61,62

In addition, the mutual influence of ARS and SMX in the binary solution under different pH conditions was also examined as shown in Fig. 11c. At pH 4, the competitive adsorption between ARS and SMX on ZnFe2O4/BAC remained strong. As a result, H% was observed at 69.2%, which was lower than that (80.6%) in the single solution. However, at other pHs (5–10), SMX supported ARS adsorption, increasing ARS removal efficiency. Especially, at pH 5–6, H% was the highest at 100%. At these pHs, ZnFe2O4/BAC surface is positively charged, while SMX is predominantly neutral and ARS exists in an anionic form (ARS−). Thus, electrostatic interaction between positively charged ZnFe2O4/BAC surface and ARS− occurred. SMX did not electrostatically interact with both ZnFe2O4/BAC surface and ARS, but it had a synergistic effect on this adsorption process.63

Meanwhile, SMX adsorption on ZnFe2O4/BAC decreased in the presence of ARS (Fig. 11d). At pH 4–10, the effect of ARS on SMX removal was relatively considerable due to the competitive adsorption between ARS and SMX on ZnFe2O4/BAC.64 At pH 4, there is an electrostatic attraction between ARS− and positively charged ZnFe2O4/BAC, influencing the SMX adsorption on ZnFe2O4/BAC. As a result, H% was found at 52.8%, lower than that (67%) in the single system. At pH 5–6, SMX removal efficiency remained high (87.7–90.0%), which could be attributed to other stronger interactions (e.g., hydrogen bonding and π–π interaction. At pH 7–10, the deprotonation of SMX in the presence of ARS− weakened hydrogen bonding force, leading to reduced adsorption performance. Overall, the adsorption of ARS and SMX on ZnFe2O4/BAC in the binary systems was the complex process under various pHs, which synergistic effect and competitive adsorption played vital roles.

3.2.7. Effects of ionic strength

Fig. 12 shows the adsorption efficiencies of ARS and SMX in single-solute and binary systems using ZnFe2O4/BAC under different ionic environments. Overall, the efficiencies of all four adsorption modes were affected by the presence of electrolytes. For ARS in the single-solute system, ZnFe2O4/BAC achieved a Hmax of 88.4% in the absence of added salts. However, this value decreased to 56.9% and 60.4% upon the addition of cations and anions, respectively. Similarly, SMX removal in the single-solute system decreased from 96% to below 70.5% (with added cations) and 72.5% (with added anions). Because electrostatic interactions play a certain role in the adsorption process, dissolved ions can either promote or inhibit adsorption by screening surface charges and competing for active sites. In many cases, multivalent ions tend to exert stronger inhibitory effects due to their higher charge density. In addition, hydrolysis of certain cations (e.g., Al3+, NH4+, Mg2+) can also alter solution chemistry and alter SMX speciation, thereby further influencing adsorption behavior. Lu et al. reported that the Al3+ addition reduced SMX removal efficiency.65 Moreover, some anions can increase solution pH, making the ZnFe2O4/BAC surface more negatively charged. Consequently, electrostatic repulsion between the adsorbent and anionic forms of ARS and SMX is intensified, leading to reduced adsorption efficiency.

Fig. 12. Effect of ions on ZnFe2O4/BAC adsorption: (a) ARS (single), (b) SMX (single), (c) ARS (binary), and (d) SMX (binary). Experimental conditions: contact time = 120 min, C0 = 10 mg L−1, dosage = 0.25 g L−1, 25 °C.

Fig. 12

Open in a new tab

3.3. Adsorption model

3.3.1. Adsorption kinetic

Table 1 summarizes the adsorption of ARS and SMX by ZnO/BAC in both single and binary solutions fitted with the four kinetic models. The results showed that ARS in single-component solution followed the pseudo second-order with the highest (Radj)2 of 0.989. This adsorption process occurred through a chemical mechanism.66 Meanwhile, SMX (single) and ARS (binary) adsorptions follow the Elovich model, with (Radj)2 = 0.994–0.998. The initial SMX adsorption rate constants (α) in the single model is found at 17.456 mg g−1 min−1. In the binary mode, α for ARS adsorption is lower, at 10.921 mg g−1 min−1. Thamer et al. reported that adsorption of crystal violet dye onto sulfonated activated carbon derived from pomegranate peel waste was well fitted with Elovich model.67 They found very high correlation coefficients at R2 = 0.99978 for activated carbon, and at R2 = 0.99996 for sulfonated activated carbon. SMX adsorption by ZnO/BAC in the binary system was best described by Bangham model ((Radj)2 = 0.997).

Table 1. Kinetic models for adsorption of ARS and SMX by ZnO/BAC.
Kinetic model ARS (single) SMX (single) ARS (binary) SMX (binary)
Pseudo first-order
Q 1 (mg g−1) 17.028 ± 0.854 30.653 ± 0.837 30.343 ± 0.009 27.691 ± 0.170
k 1 (1 min−1) 0.04 ± 0.003 0.071 ± 0.002 0.061 ± 0.006 0.117 ± 0.005
MRE 4.712 ± 0.963 8.520 ± 0.439 7.485 ± 0.449 15.287 ± 0.212
(Radj)2 0.984 ± 0.005 0.952 ± 0.004 0.957 ± 0.005 0.968 ± 0.004
 
Pseudo second-order
Q 2 (mg g−1) 20.132 ± 0.257 33.966 ± 0.295 31.799 ± 0.139 29.499 ± 0.290
k 2 (g mg−1 min−1) 0.002 ± 0.0003 0.003 ± 0.0002 0.002 ± 0.0002 0.006 ± 0.0005
MRE 4.324 ± 0.912 4.339 ± 0.325 3.793 ± 0.588 16.495 ± 0.406
(Radj)2 0.989 ± 0.006 0.987 ± 0.001 0.988 ± 0.004 0.983 ± 0.002
 
Elovich
α (mg g−1 min−1) 1.638 ± 0.166 17.456 ± 0.055 10.921 ± 0.678 55.375 ± 0.500
β (g mg−1) 0.223 ± 0.016 00.188 ± 0.010 0.170 ± 0.07 0.259 ± 0.014
MRE 6.273 ± 0.326 1.370 ± 0.162 2.482 ± 0.534 16.536 ± 0.365
(Radj)2 0.980 ± 0.012 0.998 ± 0.001 0.994 ± 0.003 0.995 ± 0.001
 
Bangham
k B 3.575 ± 0.204 12.138 ± 0.697 10.339 ± 0.750 12.872 ± 0.563
α B 0.326 ± 0.015 0.202 ± 0.004 0.232 ± 0.020 0.166 ± 0.010
MRE 9.251 ± 0.065 1.729 ± 0.517 3.485 ± 0.248 16.459 ± 0.353
(Radj)2 0.963 ± 0.017 0.997 ± 0.001 0.992 ± 0.003 0.997 ± 0.001
Open in a new tab

The adsorptions of ARS and SMX by Fe3O4/BAC in single and binary solutions were shown in Table 2. ARS adsorption in the binary mode followed the pseudo second-order with a high (Radj)2 of 0.994. This reflects a high degree of reliability of the model. Meanwhile, ARS (single), SMX (single), and SMX (binary) adsorptions follow the Bangham model, with (Radj)2 = 0.985–0.994. Thus, these adsorptions could be complex processes, which chemisorption might occur on a heterogeneous surface. For adsorption of ARS and SMX by ZnFe2O4/BAC, Table 3 shows that the pseudo second-order was the best model to describe adsorptions of SMX and ARS in the single and binary modes, respectively. For adsorptions of ARS (single) and SMX (binary), Bangham model is the best fit.

Table 2. Kinetic models for adsorption of ARS and SMX by Fe3O4/BAC.
Kinetic model ARS (single) SMX (single) ARS (binary) SMX (binary)
Pseudo first-order
Q 1 (mg g−1) 30.095 ± 0.649 13 ± 0.094 42.361 ± 0.284 19.397 ± 0.092
k 1 (1 min−1) 0.020 ± 0.0017 0.039 ± 0.0013 0.031 ± 0.0001 0.078 ± 0.0022
MRE 14.499 ± 0.840 13.106 ± 0.645 4.224 ± 0.691 8.637 ± 0.595
(Radj)2 0.943 ± 0.011 0.900 ± 0.006 0.988 ± 0.004 0.951 ± 0.002
 
Pseudo second-order
Q 2 (mg g−1) 38.947 ± 0.820 15.348 ± 0.158 51.547 ± 0.445 21.165 ± 0.154
k 2 (g mg−1 min−1) 0.0005 ± 0.0001 0.0032 ± 0.0002 0.0006 ± 0.0002 0.0060 ± 0.0005
MRE 11.376 ± 0.977 10.187 ± 0.284 2.649 ± 0.827 4.791 ± 0.427
(Radj)2 0.961 ± 0.011 0.942 ± 0.004 0.994 ± 0.003 0.982 ± 0.0053
 
Elovich
α (mg g−1 min−1) 0.990 ± 0.109 1.624 ± 0.104 2.644 ± 0.164 16.722 ± 0.826
β (g mg−1) 0.092 ± 0.006 0.314 ± 0.007 0.078 ± 0.002 0.334 ± 0.015
MRE 8.346 ± 0.856 7.093 ± 0.342 3.443 ± 0.911 3.101 ± 0.703
(Radj)2 0.975 ± 0.010 0.972 ± 0.002 0.986 ± 0.005 0.994 ± 0.002
 
Bangham
k B 2.682 ± 0.288 2.981 ± 0.087 6.714 ± 0.284 8.679 ± 0.337
α B 0.482 ± 0.022 0.311 ± 0.006 0.378 ± 0.008 0.176 ± 0.008
MRE 4.766 ± 0.804 5.596 ± 0.048 6.733 ± 0.867 2.931 ± 0.816
(Radj)2 0.988 ± 0.008 0.985 ± 0.002 0.959 ± 0.009 0.994 ± 0.001
Open in a new tab
Table 3. Kinetic models for adsorption of ARS and SMX by ZnFe2O4/BAC.
Kinetic model ARS (single) SMX (single) ARS (binary) SMX (binary)
Pseudo first-order
Q 1 (mg g−1) 37.682 ± 0.291 45.197 ± 0.319 45.228 ± 0.178 35.878 ± 0.062
k 1 (1 min−1) 0.027 ± 0.001 0.137 ± 0.008 0.047 ± 0.002 0.085 ± 0.001
MRE 11.365 ± 0.840 4.312 ± 0.387 6.498 ± 0.645 9.776 ± 0.176
(Radj)2 0.973 ± 0.008 0.984 ± 0.003 0.979 ± 0.005 0.934 ± 0.002
 
Pseudo second-order
Q 2 (mg g−1) 46.693 ± 0.557 47.991 ± 0.393 51.688 ± 0.280 39.432 ± 0.084
k 2 (g mg−1 min−1) 0.001 ± 0.00004 0.006 ± 0.00047 0.001 ± 0.00008 0.003 ± 0.00004
MRE 9.157 ± 1.966 1.205 ± 0.581 5.366 ± 0.559 5.507 ± 0.172
(Radj)2 0.986 ± 0.006 0.998 ± 0.001 0.982 ± 0.003 0.978 ± 0.001
 
Elovich
α (mg g−1 min−1) 1.868 ± 0.108 2375.066 ± 0.944 7.399 ± 0.943 37.611 ± 0.754
β (g mg−1) 0.084 ± 0.002 0.234 ± 0.012 0.098 ± 0.003 0.178 ± 0.001
MRE 3.570 ± 0.857 1.816 ± 0.493 5.384 ± 0.623 1.693 ± 0.165
(Radj)2 0.9938 ± 0.003 0.9968 ± 0.002 0.971 ± 0.007 0.9971 ± 0.002
 
Bangham
k B 4.865 ± 0.273 27.892 ± 0.512 12.333 ± 0.752 16.011 ± 0.060
α B 0.415 ± 0.011 0.111 ± 0.013 0.272 ± 0.012 0.179 ± 0.001
MRE 4.587 ± 0.413 2.730 ± 0.839 7.033 ± 0.617 0.874 ± 0.104
(Radj)2 0.996 ± 0.001 0.932 ± 0.089 0.958 ± 0.010 0.999 ± 0.001
Open in a new tab

3.3.2. Adsorption isotherms

Table 4 summarizes the adsorption isotherms of ARS and SMX onto ZnFe2O4/BAC in single and binary solutions at 25 °C. ARS adsorptions in both single and binary modes are best described by the Langmuir model with (Radj)2 higher than 0.9. There is also no significant difference in Qm (50.2–51.3 mg g−1). Thus, the ARS adsorption onto ZnFe2O4/BAC at 25 °C was a monolayer process, which each adsorption site interacted with a single molecule. Gholivand et al. reported that ARS adsorption onto polypyrrole-coated magnetic nanoparticles also followed the Langmuir model (R2 = 0.9644).68 For SMX adsorption, Freundlich was the best describing model ((Radj)2 = 0.89–0.916), suggesting that this was the multilayer adsorption process.

Table 4. Isotherm models for adsorption of ARS and SMX by ZnFe2O4/BAC at 25 °C.
Isotherm model ARS (single) SMX (single) ARS (binary) SMX (binary)
Langmuir
Q m (mg g−1) 51.337 ± 0.320 69.284 ± 0.847 50.237 ± 0.003 67.635 ± 0.989
k L (L mg−1) 3.112 ± 0.949 2.453 ± 0.832 71.124 ± 2.837 0.712 ± 0.063
R L 0.0047 ± 0.001 0.0063 ± 0.003 0.0002 ± 0.001 0.019 ± 0.002
MRE 2.407 ± 0.044 5.2688 ± 0.849 4.625 ± 0.004 7.283 ± 1.005
(Radj)2 0.907 ± 0.002 0.874 ± 0.063 0.903 ± 0.001 0.802 ± 0.034
 
Freundlich
k F 40.132 ± 0.471 43.210 ± 1.827 40.132 ± 0.471 37.075 ± 0.962
n −1 0.069 ± 0.003 0.143 ± 0.017 0.069 ± 0.003 0.166 ± 0.014
MRE 3.231 ± 0.330 12.109 ± 0.821 3.231 ± 0.330 4.166 ± 0.554
(Radj)2 0.823 ± 0.033 0.890 ± 0.041 0.823 ± 0.033 0.916 ± 0.045
Open in a new tab

At 35 °C, the adsorption of ARS and SMX onto ZnFe2O4/BAC follow the Freundlich model (Table 5), except for the ARS in binary mode (follows the Langmuir model). By contrast, Table 6 exhibits the adherence of Langmuir model for adsorption of ARS and SMX on ZnFe2O4/BAC in both single and binary modes at 45 °C, except for the ARS in binary mode (follows the Freundlich model). The results reveal that Qm obtained from the Langmuir model can be ranked as ARS (binary, 58.878 mg g−1) < ARS (single, 64.252 mg g−1) < SMX (binary, 73.401 mg g−1) > SMX (single, 85.708 mg g−1). Han et al. also reported that the adsorption of SMX onto rice straw-derived biochar followed the Langmuir model.69 Importantly, RL and n−1 of all models are in range from zero to one, suggesting the all adsorption processes at 25–45 °C were favorable.

Table 5. Isotherm models for adsorption of ARS and SMX by ZnFe2O4/BAC at 35 °C.
Isotherm model ARS (single) SMX (single) ARS (binary) SMX (binary)
Langmuir
Q m (mg g−1) 59.610 ± 0.558 80.841 ± 0.655 55.678 ± 0.009 77.247 ± 0.680
k L (L mg−1) 8.880 ± 0.467 1.879 ± 0.113 232.532 ± 0.139 0.840 ± 0.151
R L 0.002 ± 0.0002 0.007 ± 0.0004 0.001 ± 0.0003 0.016 ± 0.0030
MRE 2.526 ± 0.497 7.785 ± 0.363 3.018 ± 0.013 8.202 ± 0.543
(Radj)2 0.764 ± 0.010 0.792 ± 0.028 0.818 ± 0.002 0.795 ± 0.077
 
Freundlich
k F 51.889 ± 0.081 53.596 ± 0.425 49.711 ± 0.123 39.159 ± 0.465
n −1 0.043 ± 0.002 0.121 ± 0.005 0.038 ± 0.001 0.189 ± 0.021
MRE 1.069 ± 0.453 7.682 ± 0.482 4.251 ± 0.165 8.458 ± 0.101
(Radj)2 0.963 ± 0.003 0.849 ± 0.037 0.765 ± 0.017 0.884 ± 0.026
Open in a new tab
Table 6. Isotherm models for adsorption of ARS and SMX by ZnFe2O4/BAC at 45 °C.
Isotherm model ARS (single) SMX (single) ARS (binary) SMX (binary)
Langmuir
Q m (mg g−1) 64.252 ± 0.022 85.708 ± 1.063 58.878 ± 0.020 73.401 ± 0.432
k L (L mg−1) 8.503 ± 0.483 1.871 ± 0.090 186.547 ± 0.338 1.040 ± 0.051
R L 0.0016 ± 0.001 0.0079 ± 0.001 0.0001 ± 0.001 0.0127 ± 0.001
MRE 4.137 ± 0.005 8.002 ± 0.332 7.487 ± 0.010 4.193 ± 1.552
(Radj)2 0.901 ± 0.001 0.821 ± 0.008 0.842 ± 0.002 0.928 ± 0.026
 
Freundlich
k F 55.834 ± 0.109 52.497 ± 0.800 51.491 ± 0.163 41.343 ± 0.804
n −1 0.041 ± 0.001 0.141 ± 0.014 0.050 ± 0.001 0.171 ± 0.005
MRE 4.163 ± 0.006 7.456 ± 0.571 7.993 ± 0.209 16.443 ± 3.154
(Radj)2 0.764 ± 0.004 0.810 ± 0.035 0.924 ± 0.012 0.895 ± 0.035
Open in a new tab

3.4. Recyclability of ZnFe2O4/BAC

3.4.1. Regeneration solvents

Fig. 13 shows that the adsorption efficiency of ARS and SMX in single and binary solutions by ZnFe2O4/BAC regenerated with different desorption solvents. It can be observed that 0.1 M HCl was the best eluent for regeneration of ZnFe2O4/BAC for all four adsorption modes. For ARS in single and binary solutions, H% were found at 56.1% and 68.5%, respectively. For SMX, H% values were higher at 78.8% and 66.6% in single and binary solution, respectively. Under strongly acidic conditions. At low pH, ARS was decolored and degraded from ZnFe2O4/BAC structure. Meanwhile, SMX can form salt, enhancing its aqueous solubility, leading to effective desorption.70,71

Fig. 13. Effect of different solvents (0.1 M HCl. 0.1 M NaOH, 0.1 M KOH, and ethanol) on the regeneration performance of ZnFe2O4/BAC.

Fig. 13

Open in a new tab

3.4.2. Reusability study

Fig. 14 shows that the adsorption performance of ARS and SMX in single and binary solutions over multiple reuse cycles. In general, the adsorption efficiency decreased with each cycle. It can be explained that a number of ARS and SMX molecules were trapped into pores of ZnFe2O4/BAC after the desorption, leading to the unavailability of active sites for the next runs. Fayazi et al. also reported that the adsorption results decreased after many cycles.47 After the 4th reuse, ZnFe2O4/BAC achieved low adsorption efficiencies for ARS of 32.4% (single) and 39.6% (binary). For SMX adsorption, H% at the 4th reuse was 23.0% and 32.9% in single and binary solutions, respectively. Thus, ZnFe2O4/BAC can be reusable adsorbent for water treatment.

Fig. 14. ZnFe2O4/BAC absorbs contaminants through reuse (a) ARS (single), (b) SMX (single), (c) ARS (binary), and (d) SMX (binary).

Fig. 14

Open in a new tab

3.5. Adsorption mechanism

Hydrogen bonding plays a crucial role in the adsorption of ARS and SMX onto ZnFe2O4/BAC as illustrated in Fig. 15. According to Fig. 5, ZnFe2O4/BAC surface possesses many functional groups (carboxylic, lactone, and phenolic). Therefore, it is possible that –OH and –COOH groups on the ZnFe2O4/BAC surface forms hydrogen bonds with ARS and SMX. Specifically, the –OH, –CO–, and –SO3− groups of ARS molecules can establish hydrogen bonds (e.g., –OH⋯HO–, –OH⋯OC–, –CO⋯HOOC–, etc.) with the –OH and –COOH groups present on the ZnFe2O4/BAC surface. Meanwhile, SMX molecules containing –NH2 and –SO2NH– groups are also capable of forming hydrogen interactions (i.e., –NH2⋯OH, –SO2NH⋯OH, etc.) at the same way. To confirm, FTIR analysis of ZnFe2O4/BAC before and after SMX, ARX, and mixed SMX/ARS adsorption was conducted (Fig. S3). A peak at 3147 cm−1 (before adsorption) shifted to new locations (after adsorption) at 3139 cm−1 (ARS), 3158 cm−1 (SMX), and 3190 cm−1 (binary), suggesting the formation of hydrogen bonding.72 Sun et al. also reported hydrogen bonding between SMX molecules and the surface of biochar derived from coconut shells activated with KOH and FeCl3·6H2O.73 In addition, the hydrogen atoms of hydroxyl groups on ZnFe2O4/BAC can interact with the π-electron aromatic rings of ARS and SMX, referred to as Yoshida hydrogen bonding interactions.74 This type of interaction was also reported in the adsorption mechanism suggested between methylene blue dye and vegetable waste-based biocomposites.75

Fig. 15. Adsorption mechanisms of SMX and ARS onto ZnFe2O4/BAC.

Fig. 15

Open in a new tab

Pore-filling can be another common mechanism in the adsorption process, which ARS and SMX molecules diffuse into the pore structures of the adsorbent.76,77 Indeed, surface area results (SBET) of ZnFe2O4/BAC revealed a significant decrease from 410 m2 g−1 (before adsorption) to 267 m2 g−1 (after mixed SMX-ARS adsorption). This suggests that the ARS and SMX molecules diffused into and occupied the internal cavities of the carbon matrix, leading to pore-blockage. Thus, filled or sterically blocked smaller voids can no longer accommodate N2 molecules during gas physisorption measurements. Chen et al. also reported that pore-filling mechanism occurred between ciprofloxacin and cotton-based KOH-activated carbon, and antibiotics with small molecular size tend to easily enter mesopores of adsorbent compared to dyes (such as Rhodamine B and methyl orange) with larger sizes.78

Electrostatic interaction is one of the main mechanisms enhance the adsorption between SMX−/ARS− and the positively ZnFe2O4/BAC surface. This interaction is greatly dependent on pH, which was discussed in Section 3.2.6. Finally, the benzene rings of amino and sulfonamide groups in SMX act as electron acceptors, while the π orbitals on the carbon surface act as donors, form π–π donor–acceptor interactions.79 According to Fig. S3, shifts from 1615 cm−1 (before adsorption) to 1612 cm−1 (after ARS adsorption), 1562 cm−1 (after SMX adsorption), and 1554 cm−1 (after binary ARS-SMX adsorption) are observed, suggesting that π–π interactions could contribute to ARS and SMX adsorptions.80

3.6. Comparative study

Table S1 summarizes the adsorption conditions and the removal efficiencies of ARS and SMX using various adsorbents (e.g., oxide-modified activated carbon, carbon nanotubes, and biochar-based composites derived from different biomass sources, etc.) reported in previous studies. Almost results showed the optimal pHs of 2–4. Meanwhile, this study reported an optimal pH 6, which was suitable for the real treatment condition. Compared with adsorbents such as activated carbon/γ-Fe2O3 nano-composite (7 cycles), Fe3O4@PPy NPs (5 cycles), NiFe2O4/polyaniline magnetic composite (6 cycles), etc., the reusability of ZnFe2O4/BAC was still low with four cycles.

4. Conclusions

Herein, Baccaurea ramiflora L. fruit peels were used as a biomass precursor for synthesis of ZnFe2O4/BAC. The ZnFe2O4/BAC possessed a porous structure and surface functional groups. ZnFe2O4/BAC could achieve high removal efficiencies for ARS and SMX in both single and binary modes. The adsorption kinetics of ARS and SMX in single and binary solutions on ZnFe2O4/BAC followed the pseudo second-order and Bangham models. The adsorption isotherm of ARS fitted well with the Langmuir model, while SMX adsorption processes followed the Freundlich model at 25 °C. For regeneration, HCl was selected as the most effective eluent agent. ZnFe2O4/BAC maintained reusability for four cycles. ARS and SMX adsorption on ZnFe2O4/BAC occurred via hydrogen bonding, pore-filling, electrostatic interactions, π–π interactions, and Yoshida hydrogen bonding. Thus, ZnFe2O4/BAC can be a potential adsorbent for the removal of ARS and SMX in both single and binary systems.

Author contributions

Sang Thi Tuyet Nguyen contributed to conceptualization, investigation, writing – original draft. Thuy Thi Thanh Nguyen contributed to methodology, validation, data curation. Thuan Van Tran contributed to writing – original draft, writing – review & editing, supervision, project administration. All authors read and approved the final manuscript.

Conflicts of interest

The authors report no declarations of interest.

Supplementary Material

RA-016-D6RA01341D-s001

Acknowledgments

This research is funded by Nguyen Tat Thanh University, Ho Chi Minh City, Vietnam under grant number 2026.01.25/HĐ-KHCN. This manuscript is based upon research originally conducted for an Undergraduate Thesis, authored by Nguyen Thi Tuyet Sang, supervised by Tran Van Thuan and Nguyen Thi Thanh Thuy, and completed at Nong Lam University Ho Chi Minh City, Vietnam.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d6ra01341d.

References

  1. Tkaczyk A. Mitrowska K. Posyniak A. Sci. Total Environ. 2020;717:137222. doi: 10.1016/j.scitotenv.2020.137222. [DOI] [PubMed] [Google Scholar]
  2. Duan W. Cui H. Jia X. Huang X. Sci. Total Environ. 2022;820:153178. doi: 10.1016/j.scitotenv.2022.153178. [DOI] [PubMed] [Google Scholar]
  3. Mu X. Chen C. Fan Q. Zhang W. Liu F. Guo J. Qi W. Liu H. Sci. Total Environ. 2025;958:178034. doi: 10.1016/j.scitotenv.2024.178034. [DOI] [PubMed] [Google Scholar]
  4. Sanusi I. O. Olutona G. O. Wawata I. G. Onohuean H. Environ. Sci. Pollut. Res. 2023;30:90595–90614. doi: 10.1007/s11356-023-28802-4. [DOI] [PubMed] [Google Scholar]
  5. Rodrigues A. R. Almeida G. C. F. de Farias N. O. de Albuquerque A. F. de Oliveira A. C. Cardoso J. C. M. Militão G. F. G. Cabral I. M. Faustino C. A. Vitorino J. D. Botelho M. T. Costa P. M. Umbuzeiro G. de A. Environ. Mol. Mutagen. 2026;67:e70046. doi: 10.1002/em.70046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Nguyen D. T. C. Jalil A. A. Nguyen L. M. Nguyen D. H. Environ. Res. 2025;279:121807. doi: 10.1016/j.envres.2025.121807. [DOI] [PubMed] [Google Scholar]
  7. Dutta S. Gupta B. Srivastava S. K. Gupta A. K. Mater. Adv. 2021;2:4497–4531. doi: 10.1039/D1MA00354B. [DOI] [Google Scholar]
  8. Wu X. Lin Y. Pi Z. Tan X. Sep. Purif. Technol. 2025;377:134236. doi: 10.1016/j.seppur.2025.134236. [DOI] [Google Scholar]
  9. Lesbayev B. Auyelkhankyzy M. Ustayeva G. Yeleuov M. Rakhymzhan N. Maltay A. Maral Y. S. Afr. J. Chem. Eng. 2023;43:327–336. [Google Scholar]
  10. Algarni T. S. Al-Mohaimeed A. M. Al-Odayni A.-B. Abduh N. A. Y. Nanomaterials. 2022;12:3224. doi: 10.3390/nano12183224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Li Y. X. Wang L. Chai F. F. Jing H. F. Gao Z. Q. Zhang Q. H. Zhao X. D. Water Sci. Technol. 2020;82:877–886. doi: 10.2166/wst.2020.389. [DOI] [PubMed] [Google Scholar]
  12. Shabbirahmed A. M. Jacob A. Dey P. Somu P. Haldar D. Discover Appl. Sci. 2025;7:771. doi: 10.1007/s42452-025-07463-7. [DOI] [Google Scholar]
  13. Rohilla S. Food Bioeng. 2023;2:339–349. doi: 10.1002/fbe2.12069. [DOI] [Google Scholar]
  14. Goyal A. K. Middha S. K. Usha T. Adv. Free Radic. Biol. Med. 2022;22:231–249. [Google Scholar]
  15. Rafi K. N. Sen P. Uddin Z. MG A. Food Sci. Eng. 2020;1:105–117. [Google Scholar]
  16. Roy H. Sarkar D. Pervez M. N. Paul S. Cai Y. Naddeo V. Firoz S. H. Islam M. S. Water. 2023;15:394. doi: 10.3390/w15030394. [DOI] [Google Scholar]
  17. Van Phuoc N. Le Quoc H. Tran T. T. Van Tan L. Hoang B. N. J. Chem. Res. 2023;47:17475198231212144. [Google Scholar]
  18. Sedefoglu N. Optik. 2023;288:171217. doi: 10.1016/j.ijleo.2023.171217. [DOI] [Google Scholar]
  19. Sayed N. S. M. Ahmed A. S. A. Abdallah M. H. Gouda G. A. Sci. Rep. 2024;14:5384. doi: 10.1038/s41598-024-55158-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Worasawat S. Masuzawa T. Hatanaka Y. Neo Y. Mimura H. Pecharapa W. Mater. Today: Proc. 2018;5:10964–10969. [Google Scholar]
  21. Namikuchi E. A. Gaspar R. D. L. da Silva D. S. Raimundo I. M. Mazali I. O. Nano Express. 2021;2:020022. doi: 10.1088/2632-959X/ac0596. [DOI] [Google Scholar]
  22. Zainullah M. Herawati L. R. Subadra S. U. I. Utomo J. Hidayat N. Khamsan N. E. Taufiq A. BIO Web Conf. 2025;183:01010. doi: 10.1051/bioconf/202518301010. [DOI] [Google Scholar]
  23. Ali A. A. Abdul-Hameed H. M. Desalin. Water Treat. 2025;321:100951. doi: 10.1016/j.dwt.2024.100951. [DOI] [Google Scholar]
  24. Mueller F. Bresser D. Paillard E. Winter M. Passerini S. J. Power Sources. 2013;236:87–94. doi: 10.1016/j.jpowsour.2013.02.051. [DOI] [Google Scholar]
  25. Andhare D. D. Jadhav S. A. V Khedkar M. Somvanshi S. B. More S. D. Jadhav K. M. J. Phys.: Conf. Ser. 2020;1644:012014. doi: 10.1088/1742-6596/1644/1/012014. [DOI] [Google Scholar]
  26. Qin M. Shuai Q. Wu G. Zheng B. Wang Z. Wu H. Mater. Sci. Eng., B. 2017;224:125–138. doi: 10.1016/j.mseb.2017.07.016. [DOI] [Google Scholar]
  27. Konicki W. Siber D. Narkiewicz U. Pol. J. Chem. Technol. 2017;19:65–74. doi: 10.1515/pjct-2017-0069. [DOI] [Google Scholar]
  28. Tripathi N. Rodriguez Uribe A. Weldekidan H. Misra M. Mohanty A. K. Mater. Adv. 2022;3:9071–9082. doi: 10.1039/D2MA00678B. [DOI] [Google Scholar]
  29. Nguyen D.-K. Ly-Tran Q.-B. Dinh V.-P. Duong B.-N. Nguyen T.-P.-T. Nguyen Kim Tuyen P. RSC Adv. 2024;14:39205–39218. doi: 10.1039/D4RA07455F. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Oliveira T. P. Marques G. N. Macedo Castro M. A. Viana Costa R. C. Rangel J. H. G. Rodrigues S. F. dos Santos C. C. Oliveira M. M. J. Mater. Res. Technol. 2020;9:15001–15015. doi: 10.1016/j.jmrt.2020.10.080. [DOI] [Google Scholar]
  31. Borba L. L. Cuba R. M. F. Cuba Terán F. J. Castro M. N. Mendes T. A. Braz. Arch. Biol. Technol. 2019;62:e19180450. doi: 10.1590/1678-4324-2019180450. [DOI] [Google Scholar]
  32. Adeoye J. B. Lau S. Y. Tan Y. H. Tan Y. Y. Chiong T. Mubarak N. M. Khalid M. Ng J. T. W. Sci. Rep. 2025;15:31754. doi: 10.1038/s41598-025-17464-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Qin N. Tian C. Carter L. Tao D. Zhou Y. Zhang F. Next Sustain. 2024;4:100057. doi: 10.1016/j.nxsust.2024.100057. [DOI] [Google Scholar]
  34. Althobaiti S. A. Nabil G. M. Mahmoud M. E. J. Mol. Liq. 2025;429:127597. doi: 10.1016/j.molliq.2025.127597. [DOI] [Google Scholar]
  35. Ullah R. Khitab F. Gul H. Khattak R. Ihsan J. Khan M. Khan A. Vincevica-Gaile Z. Aouissi H. A. Catalysts. 2023;13:1061. doi: 10.3390/catal13071061. [DOI] [Google Scholar]
  36. Đukić A. B. Vukelić N. S. Paskaš-Mamula B. M. Novaković N. B. Vujasin R. T. Milošević S. S. Matović L. L. Tehnika. 2014;69:367–372. [Google Scholar]
  37. Yao P. Cen J. Fang M. Wang T. Wang Q. RSC Adv. 2018;8:17558–17568. doi: 10.1039/C7RA13344H. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Alam S. Khan M. S. Bibi W. Zekker I. Burlakovs J. Ghangrekar M. M. Bhowmick G. D. Kallistova A. Pimenov N. Zahoor M. Water. 2021;13:1453. doi: 10.3390/w13111453. [DOI] [Google Scholar]
  39. Rameshbabu R. Ramesh R. Kanagesan S. Karthigeyan A. Ponnusamy S. J. Mater. Sci.: Mater. Electron. 2014;25:2583–2588. doi: 10.1007/s10854-014-1914-2. [DOI] [Google Scholar]
  40. Guo P. Lv M. Han G. Wen C. Wang Q. Li H. Zhao X. Materials. 2016;9:806. doi: 10.3390/ma9100806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Baker F. S., Miller C. E., Repik A. J. and Tolles E. D., in Kirk-Othmer Encyclopedia of Chemical Technology, Wiley, 2000 [Google Scholar]
  42. Machado F. M. Carmalin S. A. Lima E. C. Dias S. L. P. Prola L. D. T. Saucier C. Jauris I. M. Zanella I. Fagan S. B. J. Phys. Chem. C. 2016;120:18296–18306. doi: 10.1021/acs.jpcc.6b03884. [DOI] [Google Scholar]
  43. Insuwan W. Srihongthong S. Permngam N. Tongkasee P. Agric. Nat. Resour. 2023;57:163–172. [Google Scholar]
  44. Zhang R. Zheng X. Chen B. Ma J. Niu X. Zhang D. Lin Z. Fu M. Zhou S. J. Cleaner Prod. 2020;256:120662. doi: 10.1016/j.jclepro.2020.120662. [DOI] [Google Scholar]
  45. Rajapaksha P. Orrell-Trigg R. Truong Y. B. Cozzolino D. Truong V. K. Chapman J. Environ. Sci. Adv. 2022;1:456–469. [Google Scholar]
  46. Budhiraja V. Urh A. Horvat P. Krzan A. Polymers. 2022;14:2674. doi: 10.3390/polym14132674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Fayazi M. Ghanei-Motlagh M. Taher M. A. Mater. Sci. Semicond. Process. 2015;40:35–43. doi: 10.1016/j.mssp.2015.06.044. [DOI] [Google Scholar]
  48. Pamphile N. Xuejiao L. Guangwei Y. Yin W. J. Hazard. Mater. 2019;368:602–612. doi: 10.1016/j.jhazmat.2019.01.093. [DOI] [PubMed] [Google Scholar]
  49. Fu F. Gao Z. Gao L. Li D. Ind. Eng. Chem. Res. 2011;50:9712–9717. doi: 10.1021/ie200524b. [DOI] [Google Scholar]
  50. Rápó E. Tonk S. Molecules. 2021;26:5419. doi: 10.3390/molecules26175419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Khamis M. I. Ibrahim T. H. Jumean F. H. Sara Z. A. Atallah B. A. Processes. 2020;8:556. doi: 10.3390/pr8050556. [DOI] [Google Scholar]
  52. Delpiano G. R. Tocco D. Medda L. Magner E. Salis A. Int. J. Mol. Sci. 2021;22:1–16. doi: 10.3390/ijms22020788. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zeng S. Kan E. Chemosphere. 2022;306:135554. doi: 10.1016/j.chemosphere.2022.135554. [DOI] [PubMed] [Google Scholar]
  54. Zhao M. Shao M. Ma X. E3S Web Conf. 2021;251:02036. doi: 10.1051/e3sconf/202125102036. [DOI] [Google Scholar]
  55. Padmavathy K. S. Madhu G. Haseena P. V. Procedia Technol. 2016;24:585–594. doi: 10.1016/j.protcy.2016.05.127. [DOI] [Google Scholar]
  56. Wagh P. J. Sci. Res. Rep. 2014;3:2197–2215. doi: 10.9734/JSRR/2014/8005. [DOI] [Google Scholar]
  57. Shalaby A. A. Mohamed A. A. RSC Adv. 2020;10:11311–11316. doi: 10.1039/C9RA10568A. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Ahmed M. B. Zhou J. L. Ngo H. H. Johir M. A. H. Sun L. Asadullah M. Belhaj D. J. Hazard. Mater. 2018;360:270–278. doi: 10.1016/j.jhazmat.2018.08.005. [DOI] [PubMed] [Google Scholar]
  59. Yu H. Zhang Y. Wang L. Tuo Y. Yan S. Ma J. Zhang X. Shen Y. Guo H. Han L. RSC Adv. 2024;14:1854–1865. doi: 10.1039/D3RA05964B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Palma-Goyes R. E. Sosa-Rodríguez F. S. Rivera F. F. Vazquez-Arenas J. Environ. Sci. Pollut. Res. 2022;29:42201–42214. doi: 10.1007/s11356-021-16154-w. [DOI] [PubMed] [Google Scholar]
  61. Moral-Rodríguez A. I. Leyva-Ramos R. Ocampo-Pérez R. Mendoza-Barron J. Serratos-Alvarez I. N. Salazar-Rabago J. J. Adsorption. 2016;22:89–103. doi: 10.1007/s10450-016-9758-0. [DOI] [Google Scholar]
  62. Zhou Y. Chen L. Lu P. Tang X. Lu J. Sep. Purif. Technol. 2011;81:184–190. doi: 10.1016/j.seppur.2011.07.026. [DOI] [Google Scholar]
  63. Yin G. Chen X. Sarkar B. Bolan N. S. Wei T. Zhou H. Wang H. Chem. Eng. J. 2023;466:143199. doi: 10.1016/j.cej.2023.143199. [DOI] [Google Scholar]
  64. Lun L. Su Y. Gong X. Zhang L. Meng P. Peng D. Zhou Q. Zeng H. Zheng L. Ind. Crops Prod. 2023;199:116761. doi: 10.1016/j.indcrop.2023.116761. [DOI] [Google Scholar]
  65. Lu L. Gao M. Gu Z. Yang S. Liu Y. J. Environ. Sci. 2014;26:2535–2545. doi: 10.1016/j.jes.2014.04.007. [DOI] [PubMed] [Google Scholar]
  66. Ho Y. McKay G. Process Biochem. 1999;34:451–465. doi: 10.1016/S0032-9592(98)00112-5. [DOI] [Google Scholar]
  67. Thamer B. M. Al-aizari F. A. Abdo H. S. Molecules. 2023;28:7712. doi: 10.3390/molecules28237712. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Gholivand M. B. Yamini Y. Dayeni M. Seidi S. Tahmasebi E. J. Environ. Chem. Eng. 2015;3:529–540. doi: 10.1016/j.jece.2015.01.011. [DOI] [Google Scholar]
  69. Han X. Liang C. Li T. Wang K. Huang H. Yang X. J. Zhejiang Univ., Sci., B. 2013;14:640–649. doi: 10.1631/jzus.B1200353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. de Moura Oliveira C. H. de Melo C. C. Doriguetto A. C. New J. Chem. 2019;43:10250–10258. doi: 10.1039/C9NJ00586B. [DOI] [Google Scholar]
  71. Dahlan R. Mcdonald C. Sunderland V. B. J. Pharm. Pharmacol. 1987;39:246–251. doi: 10.1111/j.2042-7158.1987.tb06261.x. [DOI] [PubMed] [Google Scholar]
  72. Ijaz I. Bukhari A. Nazir A. Gilani E. Zain H. Shaheen A. Shaik M. R. Khan M. Shaik J. P. RSC Adv. 2025;15:5042–5059. doi: 10.1039/D4RA08766F. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Sun Y. Zheng L. Zheng X. Xiao D. Yang Y. Zhang Z. Ai B. Sheng Z. Front. Chem. 2022;9:814647. doi: 10.3389/fchem.2021.814647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Pauletto P. S. Lütke S. F. Dotto G. L. Salau N. P. G. J. Mol. Liq. 2021;336:116203. doi: 10.1016/j.molliq.2021.116203. [DOI] [Google Scholar]
  75. Abdulhameed A. S. Al Omari R. H. Althahban S. Jazaa Y. Abualhaija M. Algburi S. Biomass Bioenergy. 2025;193:107528. doi: 10.1016/j.biombioe.2024.107528. [DOI] [Google Scholar]
  76. Li Y. Wang B. Shang H. Cao Y. Yang C. Hu W. Feng Y. Yu Y. Chemosphere. 2023;326:138408. doi: 10.1016/j.chemosphere.2023.138408. [DOI] [PubMed] [Google Scholar]
  77. Lin R. Liang Z. Yang C. Zhao Z. Cui F. J. Colloid Interface Sci. 2020;573:21–30. doi: 10.1016/j.jcis.2020.03.112. [DOI] [PubMed] [Google Scholar]
  78. Chen X. Li M. He J. Wu Y. Sun J. Wen X. Chemosphere. 2025;376:144292. doi: 10.1016/j.chemosphere.2025.144292. [DOI] [PubMed] [Google Scholar]
  79. Diao Y. Shan R. Li M. Gu J. Yuan H. Chen Y. ACS Omega. 2023;8:879–892. doi: 10.1021/acsomega.2c06234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Qiu X. Zhao Y. Zhao C. Jin R. Li C. Mutabazi E. Front. Environ. Sci. 2023;11:1275087. doi: 10.3389/fenvs.2023.1275087. [DOI] [Google Scholar]

Associated Data

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

Supplementary Materials

RA-016-D6RA01341D-s001
RA-016-D6RA01341D-s001.pdf (557.6KB, pdf)

Data Availability Statement

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d6ra01341d.

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

RESOURCES