Magnetite nanoparticles decorated on cellulose aerogel for p-nitrophenol Fenton degradation: Effects of the active phase loading, cross-linker agent and preparation method

Abstract

Magnetite nanoparticles (Fe₃O₄ NPs) are among the most effective Fenton-Like heterogeneous catalysts for degrading environmental contaminants. However, Fe₃O₄ NPs aggregate easily and have poor dispersion stability because of their magnetic properties, which seriously decrease their catalytic efficiency. In this study, a novel environmentally friendly method for synthesising Fe₃O₄@CA was proposed. Fe₃O₄ NPs were immobilized on the 3D cellulose aerogels (CAs) in order to augment the degradation efficiency of p-nitrophenol (PNP) treatment and make the separation of the catalyst accessible by vacuum filtration method. Besides, CAs were fabricated from a cellulose source extracted from water hyacinth by using different cross-linking agents, such as kymene (KM) and polyvinyl alcohol–glutaraldehyde system (PVA–GA), and other drying methods, including vacuum thermal drying and freeze drying, were evaluated in the synthesis process. As-synthesized samples were analysed by various methods, including Powder X-ray diffraction, Fourier transform infrared spectroscopy, field emission scanning electron microscopy, energy dispersive X-ray analysis and Brunauer–Emmett–Teller. Then, using ultraviolet–visible spectroscopy, the difference in the degradability of PNP of the obtained material samples was also investigated to determine their potential applications. Results highlighted that the Fe₃O₄-3@CA-KF catalyst with an Fe₃O₄ loading of 0.40 g/g_CA used KM as a cross-linker and the freeze-drying method demonstrated the highest PNP removal efficiency (92.5 %) in all Fe₃O₄@CA samples with a H₂O₂ content of 5 g/L. The degradation kinetics and well-fitted pseudo-first-order model were investigated. Notably, after five successive PNP degradation experiments, this catalyst retained ∼80 % of the ability to degrade PNP, indicating its outstanding reusability. In environmental remediation, this study provides valuable insights into the development of simply separated and high-efficiency catalysts for heterogeneous catalytic reactions.

Highlights

• •Magnetite nanoparticles (Fe₃O₄) decorated on cellulose aerogel (CA) from water hyacinth.
• •Properties, activity, stability in p-nitrophenol treatment of material were assessed.
• •Effects of Fe₃O₄ loading, crosslinker, and preparation method on the performance of material.
• •Fe₃O₄@CA enhancing Fenton degradation of p-nitrophenol.

1. Introduction

Aerogels have attracted significant attention due to their notable characteristics, including their high porosity, low density, porous structure, huge specific surface areas, and very low thermal conductivity [1,2]. Because of their abundant availability, renewable natural resources and favourable biodegradability, cellulose-based aerogels have garnered significant attention and used in several sectors, such as food [3], energy storage [4], wastewater treatment [5,6], wound dressing [7] and pharmaceutical industries [8] over the last decades. Cellulose aerogel from water hyacinth has been fabricated and published in many previous works [[9], [10], [11]]. It has a variety of uses and possesses numerous unique properties, such as extremely low density (0.003–0.500 g/cm³), high porosity (80 %–99.8 %), huge surface area (100−1600 m²/g) and high compressive strength (5.20 kPa–16.67 MPa) and has good surface activity.

The aerogels' three-dimensional networks, characterized by several open cavities, enable efficient molecular or ionic transportation, rendering them highly appropriate for developing functional materials that serve as exceptional support for various nanoparticles (NPs) [12,13]. Numerous research has been conducted to investigate the development of composite materials using functional matrices for the fabricating of organic or inorganic hybrid composites, including various constituents, such as SiO₂ [14], TiO₂ [15], CoFe₂O₄ [16] and CaCO₃ [17]. These composite materials have demonstrated prospective applications in the fields of life science, environment, and catalysis, rendering them the most promising. However, studies that have modified metallic NP catalysts on CA to create heterogeneous catalyst systems for environmental treatment applications are rare. Moreover, CA materials from water hyacinth with suitable cross-linkers and optimal synthesis conditions have high potential as materials that enable catalysts to decompose difficult organic pollutants, such as p-nitrophenol (PNP) with high concentrations.

However, the matrix of native CAs is often stiff and rapidly collapses when soaked in water, restricting their scope of applicability [18,19]. Improving their mechanical strength and water stability is tough. Some researchers have mixed cellulose with kymene [10,20], used a combination of polyvinyl alcohol (PVA) and glutaraldehyde [10,21] and a cross-linking agent to solve the aforementioned deficiency in the performance of cellulose. Different cross-linkers alter the characteristics of CA, thus affecting their usefulness as catalyst carriers and the activity of the catalytic system [22]. Therefore, investigating the influence of cross-linking agents on catalytic activity during CA synthesis is crucial. Drying procedures used in CA synthesis, such as heat drying or sublimation drying, have substantial impacts on a product's characteristics [23]. CA's porosity, density, mechanical strength and reform ability are dependent on the drying procedure. No published findings have indicated that drying conditions alter catalytic activity, stability or reusability of CA in catalyst systems.

Numerous studies have utilized the heterogeneous Fenton procedure to handle PNP in wastewater [[24], [25], [26], [27]]. By generating *OH radicals with potent oxidizing activity, these radicals are intended to facilitate the decomposition of organic compounds utilizing Fe₃O₄ catalysts [28,29]. The Fe₃O₄ nanoparticles have attracted significant attention due to their large specific surface area, superparamagnetic properties, biocompatibility, catalytic activity, and environmentally favourable characteristics [2]. Nevertheless, these nanoparticles are prone to aggregation and the subsequent production of larger particles, resulting in a decrease in their overall activity [30]. Consequently, the incorporation of a carrier is required to immobilize Fe₃O₄ NPs.

Herein, a straightforward and ecofriendly approach for decorating CAs with Fe₃O₄ with different synthesis methods was established, and PNP degradation performance was evaluated. Simultaneously, the findings demonstrated that CA is an exceptional matrix for supporting Fe₃O₄ NPs. Besides, Fe₃O₄ modified on a CA surface (Fe₃O₄@CA) was produced from water hyacinth. For the cross-linking agent, the method employed kymene and the PVA-GA system. In addition, heat drying, and sublimation drying were used in Fe₃O₄@CA synthesis. We examined the physicochemical features of the catalyst system, and PNP degradation, recovery and reuse, to validate the product's prospective applicability.

2. Methodology

2.1. Chemicals and reagents

Water hyacinth was gathered from Sai Gon River in Ho Chi Minh City's District 12. Ferric chloride (FeCl₃.6H₂O, Xilong, 99 %) and ferrous sulphate (FeSO₄.7H₂O, Xilong, 99 %) were used in the preparation of Fe₃O₄ NPs. Kymene (Solenis 5574), glutaraldehyde (25 % in H₂O, Aldrich) and PVA (M_n ∼145.000) were used as cross-linking agent. Sodium hydroxide (NaOH, Xilong, 99 %), hydrogen peroxide (H₂O₂, Xilong, 30 %), sulphuric acid (H₂SO₄, Xilong, 95 %), sodium chlorite (NaClO₂, 80−83 %) and PNP (Tianjin Chemical Reagent 99 %). All of the reagents used in the experiment were employed without undergoing purification. All studies used water that has been filtered using a Milli-Q system.

2.2. Synthesis of samples

2.2.1. The process of extracting cellulose from water hyacinth

A combination was created by blending 2 % w/w dried WH fibres with a 2 % w/w NaOH solution for lignin and hemicellulose removal. The removal efficiency was enhanced by subjecting the combination to microwave heating for a duration of 20 min using the AEM-G2135W microwave model manufactured by AQUA. An aqueous solution of H₂O₂ with a concentration of 17 % (v/v) was introduced into the microwave-treated mixture after it had been cooled. The pH of the combination was then approximately adjusted to 11 using NaOH before subjecting the mixture to an additional 10 min of microwave heating. Following the cooling of the combination to ambient temperature, it underwent filtration and subsequent rinsing with distilled water in order to get a pH value of 7.0. The pre-treated material underwent a drying process at a temperature of 70 °C for 24 h, after which it was then kept for the purpose of aerogel manufacturing [9].

2.2.2. Preparation of Fe₃O₄@CA-K using kymene as the cross-linker

CA samples (CA-K) were prepared as previously reported [7,8]. A homogeneous mixture was prepared by mixing 100 mL of a cellulose fiber solution with a concentration of 2 % w/w in water. Subsequently, a volume of 0.5 mL of kymene was introduced into the mixture and agitated for 2 h using a magnetic stirrer operating at a speed of 800 revolutions per minute. The combination was subjected to refrigeration for a duration of 24 h prior to undergoing sublimation at a temperature of −70 °C, resulting in the acquisition of an aerogel sample. Subsequently, the acquired components were subjected to an oven at a temperature of 80 °C for a duration of 3 h, with the purpose of augmenting the cross-linking of kymene inside the resultant aerogel cellulose product [10].

The fabrication process of the Fe₃O₄@CA-K samples involved the coprecipitation method. Firstly, 1.6 g of CA-K was dispersed in a mixture of 100 mL containing an aqueous solution of FeSO₄.7H₂O and FeCl₃.6H₂O in a 1:1 M ratio. Proceeding to agitate while maintaining the inert medium for the reaction at 75 ± 5 °C, 200 mL of distilled water was subsequently added to the mixture in a 500 mL beaker equipped with a magnetic agitator. The reaction was continued for an additional 40 min after the pH was adjusted to 11.5 with a 1 M NaOH solution. The acquired substance was rinsed with ethanol and distilled water until pH 7 was reached. The sample was dried with two methods: i) vacuum drying for 12 h at 60 °C to obtain Fe₃O₄-X@CA-KV (X was varied [[1], [2], [3], [4]] corresponding to the content of Fe₃O₄ in samples [0.17, 0.25, 0.40 and 0.49 g/g_CA, respectively]; Table S1) and ii) freeze drying within 24 h to obtain Fe₃O₄-X@CA-KF.

2.2.3. Preparation of Fe₃O₄@CA-P by using PVA and GA as the cross-linker

CA samples (CA-P) were generated in the identical manner as CA-K. A PVA (5 % w/w in water) solution was introduced to 100 mL of a 2 % (w/w in water) cellulose fiber combination at a PVA-to-cellulose ratio of 4:3. After incorporating 0.1 mL of 1 % H₂SO₄ and 0.25 mL of GA (25 % w/w in water), the mixture was agitated for an additional 1 h at 800 rpm using a magnetic stirrer. CA-P was obtained through the freeze-drying process of cellulose hydrogel at −70 °C for 24 h. To enhance the PVA-GA's cross-linking, the aerogel material was cured for an additional 3 h at 80 °C [10].

Method of synthesising Fe₃O₄@CA-P via coprecipitation: 1.6 g of CA-P was immersed in 100 mL of an aqueous FeSO₄.7H₂O and FeCl₃.6H₂O solution. The best ratio and molar number were selected according to the previous experiments of Fe₃O₄-x@CA-K. Then, a volume of 200 mL of distilled water was introduced into a 500 mL beaker with a magnetic stirrer, keeping the reaction's inert medium at 75 ± 5 °C. The sample was dried with two methods: i) vacuum drying for 12 h at 60 °C to obtain Fe₃O₄-X@CA-PV; ii) freeze-dried within 24 h to obtain Fe₃O₄-X@CA-PF. The fabrication of the synthesis of magnetite NPs decorated on CA is illustrated in Fig. 1.

Fig. 1.

A schematic representation of the experimental technique of manufacture of Fe₃O₄ NPs decorated on CA.

2.3. Characterisation

The crystal structure of pure CA and Fe₃O₄@CA catalysts were examined using X-ray diffraction (XRD, D2-PHASER) with Cu Kα radiation and recorded at 2θ = 10°–80° and Fourier transform infrared (FTIR) system (Tensor 27 Bruker Instrument) at a range of 400–4000 cm⁻¹. A scanning electron microscope (Hitachi, S-4800) and high-resolution transmission electron microscopy (JEOL JEM 1400 equipment) were utilized in observing surface morphology, pore structure and Fe₃O₄ particles on the surfaces of the materials. Additionally, the composition of the catalyst materials was evaluated, and energy-dispersive X-ray spectroscopy was employed to analyze the elemental distribution on the surfaces of the catalysts (EDS, JEOL JST-IT 200).

The experimental procedure for synthesising Fe₃O₄ NPs that are decorated on CA is depicted in the schematic. The Nova Station A was utilized to conduct nitrogen physical adsorption measurements, and Quantachrome was employed to estimate the surface area, pore size, and pore volume of the materials. The Brunauer–Emmett–Teller (BET) equation and Barret–Joyner–Halenda (BJH) techniques were used to automatically derive these parameters from the equipment software. The point of zero charge (PZC) of the samples was determined with the salt addition technique through previously described method [31]. The magnetic characteristics of the Fe₃O₄@CA composites were examined at room temperature with a vibrating sample magnetometer (VSM) across a field range spanning from −11,000 to +11,000 Oe (ADE Technology–DMS 5000).

2.4. Fenton-like degradation experiments

The PNP solution (2000 ppm) utilized in the experiment comprised of 500 mL distilled water and 1.00 g of PNP. The initial pH of the procedure was adjusted to 3 by adding a 95 % H₂SO₄ solution. H₂O₂ was added to the solutions along with Fe₃O₄@CA-K and Fe₃O₄@CA-P samples. Samples were collected at 20-min intervals throughout the reaction, and the degradation of PNP was observed using ultraviolet–visible (UV–Vis) spectroscopy. The spectrum of wavelengths was recorded in the range of 190–800 nm, and the total reaction time was 180 min. The maximum absorption in the spectrum was detected at a wavelength of 317 nm. The removal of PNP was determined by employing Equation (1).

$$H\%=\frac{\left(C_0-C_t\right)}{C_0}\times 100\%\text{,}$$ (1)

Where, the variables C₀ and C_t represent the concentrations of PNP at the initial time (0) and at an additional point (t) during the reaction, respectively.

2.5. Recyclability measurements

In assessing the practical application potential of catalysts, reusability is one of the essential characteristics. The recyclability of the optimal catalyst for degrading PNP via heterogeneous Fenton was tested under optimised reaction conditions for five cycles. After each degrading cycle, Fe₃O₄-3@CA-KF was removed from the solution through vacuum filtering for reusability testing, and the supernatant was collected. The filtered Fe₃O₄-3@CA-KF was then rinsed with deionised water to eliminate contaminants before it was employed in the subsequent degradation experiment. The degradation rates of PNP solutions and their accompanying UV–Vis absorption spectra were determined.

3. Results and discussion

3.1. Catalytic activity

3.1.1. Effect of Fe₃O₄ loading

In a Fenton-like system, PNP was chosen as the model organic pollutant to assess the catalytic activity of Fe₃O₄@CA-KV. The pH value plays a crucial role in the Fenton process because it governs the generation of hydroxyl radicals [32]. The literature reports that an acidic pH is useful in removing pollutants because H₂O₂ has high stability at low pH levels and does not dissociate into O₂ and H₂O [33]. Based on prior research, the investigation of the Fenton decomposition process has shown that a pH level of 3 is considered to be the most [[34], [35], [36]]. Therefore, for the purpose of this investigation, a fixed pH value of 3 was used in the examination of additional variables in the degrading response of PNP.

The PNP degradation process was carried out on Fe₃O₄@CA-KV samples containing different amounts of Fe₃O₄. The initial solution pH was set at 3 and the H₂O₂ concentration was set at 4.0 g/L. The stirring speed during the process was 300 rpm. Fig. 2 depicts the process of eliminating PNP from a solution using an initial concentration of 2000 ppm, following a reaction time of 180 min. The results showed that all the samples can be used to degrade PNP and showed different levels of efficiency. After 100 min of reaction, the PNP conversion efficiency were 57.6 % for Fe₃O₄-1@CA-KV, 79.1 % for Fe₃O₄-2@CA-KV, 85.5 % for Fe₃O₄-3@CA-KV and 86.2 % for Fe₃O₄-4@CA-KV. When the Fe₃O₄ content was increased, the reaction rate increased significantly because the introduction of Fe₃O₄ NPs might have increased the number of active sites for *OH production. However, increase in the Fe₃O₄ content in Fe₃O₄-3@CA-KV and Fe₃O₄-4@CA-KV promoted PNP conversion but not significantly. After 100 min of reaction, the conversion yields of Fe₃O₄-3@CA-KV and Fe₃O₄-4@CA-KV were practically stable. At low concentrations, the inefficiency of Fe³⁺ removal was attributable to the low level of *OH radical formation required for oxidation. Conversely, the surplus Fe³⁺ ions effectively eliminated the formed *OH, thereby potentially impeding the degradation of PNP [37]. Thus, the Fe₃O₄-3@CA-KV sample with 0.40 g/g_CA of Fe₃O₄ achieved the highest PNP conversion efficiency, although Fe₃O₄-4@CA-KV had the highest amount of loaded Fe₃O₄ NPs (0.49 g/g_CA). This result indicated that the PNP degradation efficiency is not proportional to the loaded amount of Fe₃O₄ at high content. The findings were consistent with the existing literature, which posited that the degradation rate of organic pollutants rises in proportion to the concentration of Fe₃O₄, up to a certain threshold beyond which additional Fe₃O₄ incorporation has minimal impact [38]. Thus, Fe₃O₄-3@CA-KV was selected for all subsequent experiments.

Fig. 2.

The investigation of the activity of samples in the heterogeneous Fenton degradation of PNP; 1) PNP + H₂O₂; 2) PNP + Fe₃O₄-1@CA-KV; 3) PNP + H₂O₂ + Fe₃O₄-1@CA-KV, 4) PNP + H₂O₂ + Fe₃O₄-2@CA-KV; 5) PNP + H₂O₂ + Fe₃O₄-3@CA-KV; 6) PNP + H₂O₂ + Fe₃O₄-4@CA-KV.

3.1.2. Effect of H₂O₂ dosage

The presence of iron plays a critical role in the function of H₂O₂ as an oxidizing agent. To augment the competitiveness of the Fenton process in comparison to alternative processes, it is crucial for the applications to demonstrate a cost-effective operation, as well as enhanced control over the dosage of H₂O₂. The primary aim of the experiment was to optimize the Fenton degradation process to attain optimal performance levels, while simultaneously minimizing the consumption of H₂O₂. The effects of H₂O₂ content on Fe₃O₄-3@CA-KV are shown in Fig. 3. H₂O₂ concentration had a substantial influence on PNP elimination over time. After 80 min, the elimination effectiveness of PNP with varying H₂O₂ dosages was almost stable. When the dosage of H₂O₂ was raised from 3 g/L to 5 g/L, the conversion efficiency of PNP significantly increased from 77.1 % to 92.0 %. The experimental results suggested that when the concentration of Fe₃O₄ had a specified value, raising the H₂O₂ dosage caused the production of •OH radicals, which considerably enhanced the oxidation rate and degree of mineralisation [39].

Fig. 3.

The impact of H₂O₂ dosage on the heterogeneous Fenton degradation of PNP occurring on a catalyst composed of Fe₃O₄-3@CA-KV.

The further enhancement from 5 g/L to 6 g/L initiated no notable change in PNP degradation. After 80 min of reaction, the PNP conversion efficiency reached 85.5 % with 6 g/L of H₂O₂ and 92 % with 5 g/L of H₂O₂. This result can be explained by the high H₂O₂ dosage, at which *OH is scavenged. The aforementioned phenomenon led to the generation of hydroxyl radicals, which exhibit lower reactivity compared to *OH and consequently impeded the conversion of PNP [40].

Under acidic conditions, the activation of H₂O₂ proceeds through a process wherein a complex of ≡ FeIII and H₂O₂ forms [24,41,42]. This complex is denoted as ≡ FeIIIH₂O₂ (Eq. (2)), where ≡ FeIII represents Fe(III) sites present on the surface of a catalyst. The first formed ≡ FeIIIH₂O₂ species undergo a conversion to ≡ FeII species and *HO₂ (Eq. (3)). Additionally, the created *HO₂ may react further with ≡ FeIII to yield ≡ FeII species (Eq. (4)). The various ≡ FeII species (Eq. (5)) can undergo a reaction with H₂O₂, resulting in the production of *OH radicals. These radicals are extremely effective in degrading and mineralising PNP (Eq. (6)). This process aligns with the findings in the literature, which have demonstrated the ability of iron oxide to catalyse the activation of H₂O₂ and to promote the generation of *OH radicals.

≡FeIII + H₂O₂ → ≡FeIIIH₂O₂ (2)

≡FeIIIH₂O₂ → ≡FeII + *O₂H + H⁺ (3)

≡FeIII + *O₂H → ≡FeII + O₂ + H⁺ (4)

≡FeII + H₂O₂ → ≡FeIII + *OH + OH⁻ (5)

*OH + PNP →low molecular substances + CO₂ + H₂O (6)

3.1.3. Effects of cross linkers and drying methods

The effects of different cross-linkers and drying methods on PNP conversion efficiency were investigated (Fig. 4). The results showed that the degradation performance of PNP was reduced in the order Fe₃O₄-3@CA-KF > Fe₃O₄-3@CA-PF > Fe₃O₄-3@CA-KV > Fe₃O₄-3@CA-PV (Fig. 4a). CA samples using kymene as a cross-linker had significantly better PNP degradation efficiency. The aforementioned outcome has been elucidated in a prior investigation [8], wherein it was demonstrated that kymene exhibits aptitude as a cross-linking agent due to its ability to form cross-links via its epichlorohydrin moieties. The presence of cross-links improves the interaction between kymene and cellulose, thereby improving the mechanical strength, especially in wet conditions, of aerogels. PVA and GA can promote the formation of three-dimensional interconnected continuous network in a matrix material through hydrogen bonding and facilitate aerogel synthesis through PVA and GA cross-linking. Nevertheless, the products exhibit limited mechanical strength and lack high compressibility. Consequently, the CA samples incorporating kymene as a cross-linker exhibit greater porosity and reduced density compared to the samples utilizing PVA and GA as cross-linkers as cross-linkers. Moreover, the specific surface areas of heterogeneous catalysts were identified as the most significant parameter affecting the degradation activity of PNP [43,44]. As the specific surface area of a catalyst increases, the sites where degradation reactions occur can be increased.

Fig. 4.

Effects of cross-linkers and drying methods on the heterogeneous Fenton degradation of PNP.

Fig. 4 shows that the freeze-drying method has a higher degradation efficiency for PNP than the vacuum-drying method. The freeze-drying samples of CA had higher porosity and lower densities than the samples obtained through vacuum drying. These results can be explained by the low pressure and low water evaporation temperature in vacuum drying, which can cause the pores to shrink or collapse. The resulting samples would have low porosity and specific surface area. Meanwhile, freeze drying can create a sheet-like cellulose network with large and interconnected pores of a few micrometers in diameter because pore sites were frozen before vaporisation creates new pores, resulting in products with high porosity and large specific surface areas [45].

The effects of cross-linkers and drying methods on the degradation kinetics curves of PNP were determined. The compatibility between the collected data and the pseudo-first kinetic model was assessed through experimentation. The model can be expressed as Eq. (7):

$$\mathit{ln}\left(\frac{C_t}{C_0}\right)=-k_{1}t\text{,}$$ (7)

where C₀ is the initial concentration of PNP and C_t is the concentration at time t. Given that all R² values exceeded 0.95, it can be concluded that this model is highly appropriate for the specific reaction under consideration. The reaction rate constants of Fe₃O₄-3@CA-KF, Fe₃O₄-3@CA-PF, Fe₃O₄-3@CA-KV and Fe₃O₄-3@CA-PV were 0.054, 0.045, 0.024 and 0.011 min⁻¹, respectively, after linear fitting (Fig. 4b). These results were in close agreement with previous research [46,47]. The reaction rate constants were estimated, and the highest rate was found in Fe₃O₄-3@CA-KF.

3.2. Material characterisation

The XRD patterns of the Fe₃O₄@CA samples (Fig. 5a) exhibit characteristic peaks at specific angles, namely 2θ = 30.88°, 36.09°, 43.86°, 53.64°, 57.79°, and 63.43°. These peaks correspond to the diffraction lattice surfaces (220), (311), (400), (422), (511), and (440), respectively, as indicated by the JCPDS card 01-075-0449. These findings align with previous studies on the subject matter [48,49]. In addition, the observed peaks at 2θ = 16.51° and 23.14°, which correspond to the diffraction lattice surfaces (110) and (200) respectively, are indicative of the distinctive peaks associated with crystalline cellulose [50]. The obtained results have substantiated the existence of Fe₃O₄ crystals within the CA structure. The intensity of the cellulose diffraction peaks in the Fe₃O₄@CA samples was noticeably lower compared to those of the pure CA samples. This suggests a decrease in the degree of crystallinity and suggests that there was some form of interaction between the CA and Fe₃O₄ NPs. These interactions likely influenced the crystallization process of cellulose within the Fe₃O₄-decorated aerogels.

Fig. 5.

XRD patterns (a) and FTIR spectra (b) of samples.

The average crystallite size of each sample was determined by using the Scherrer equation [51] at characteristic peak (2θ) of 36.09°. Hence, the average crystallite sizes of Fe₃O₄ NPs in Fe₃O₄-1@CA-KV, Fe₃O₄-2@CA-KV, Fe₃O₄-3@CA-KV, Fe₃O₄-4@CA-KV, Fe₃O₄-3@CA-KF, Fe₃O₄-3@CA-PV and Fe₃O₄-3@CA-PF were 6.5, 6.9, 7.4, 7.7, 7.1, 7.6 and 7.4 nm, respectively. When the Fe₃O₄ content increased, the average crystallite size increased with the same drying method. At the same content, the average crystallite size of Fe₃O₄ did not change significantly. These results can be explained by the agglomeration of Fe₃O₄ on the surface of the CA samples.

In order to determine the functional group of the CA and Fe₃O₄@CA samples, an FTIR analysis was conducted within the absorption region of 4000−400 cm⁻¹ (Fig. 5b). All samples exhibited consistent absorption peaks at 2900, 1625, 894 and 554 cm⁻¹, although their intensities displayed variation. This observation suggests an interaction between Fe₃O₄ nanoparticles and cellulose molecules. Furthermore, the introduction of Fe₃O₄ NPs led to the association of oxygen atoms with the OH and COOH groups. This association resulted in the expansion of the O–H stretching peak and a shift from 3425 cm⁻¹ (pure CA) to 3320 cm⁻¹ (Fe₃O₄@CA sample). In particular, anionic groups have the ability to establish coordination-type interactions with metal ions, thereby promoting the creation of ultrafine nanoparticles that exhibit a remarkably dispersion on the surface of the CA material [52]. The peak observed at 554 cm⁻¹ was found to correspond to the vibration of the Fe–O functional group. This vibration exhibited a strong similarity to the characteristic peak of Fe₃O₄, as reported in previous studies [53,54]. The observed peak at 2900 cm⁻¹ can be attributed to the vibrational mode of the –CH group in cellulose, while the peak at 1605 cm⁻¹ corresponds to the bending vibration of the –OH group in water that is naturally absorbed [55]. In addition, the observed signals at 1098 and 898 cm⁻¹ can be attributed to the symmetrical oscillations of C–O–C in pristine cellulose, indicating the presence of β-glycosidic bonds connecting glucose units [56]. The presence of two additional peaks at 1431 and 1231 cm⁻¹ could potentially be attributed to the oscillation of the C–O bond in hemicellulose and the oscillation of the aromatic ring in lignin, as suggested by previous studies [57,58]. Following the application of chemical pretreatment, the peaks observed at 1736, 1430 and 1230 cm⁻¹, which are attributed to lignin and hemicellulose, were entirely absent from the FTIR spectrum. This absence serves as a clear indication of the substantial efficiency achieved in the removal process [59].

Scanning electron micrographs were utilized to observe the surface morphology of CA samples cross-linked with kymene and a mixture of PVA and GA, both before and after being doped with Fe₃O₄ NPs (Fig. 6). The investigation revealed an interconnected three-dimensional matrix with a great deal of branching and an abundance of porous structure in the inner section, and the structure of the CA material that was generated is more orderly organised and homogeneous and has a high porosity after being cross-linked by kymene and PVA-GA system (Fig. 6c–f). In addition, the results shown in Fig. 6a and b confirm that CA with kymene as cross-linker has a higher porosity than that of PVA-GA mixture. Fig. 6c–f exhibit the presence of diminutive Fe₃O₄ NPs, displaying a relatively homogeneous dispersion across the material's surface. Huixia Zhu et al. similarly showed comparable scanning electron microscopy (SEM) findings for Fe₃O₄@BC materials [60].

Fig. 6.

SEM images of samples; a) CA-P, b) CA-K, c) Fe₃O₄-3@CA-KV, d) Fe₃O₄-3@CA-PV, e) Fe₃O₄-3@CA-KF and f) Fe₃O₄-3@CA-PF

The composition of the synthesized samples, meanwhile, was ascertained via EDS and mapping techniques (Fig. 7). Fig. 7a–d shows that all Fe₃O₄@CA catalysts involved Fe, O and C elements, and the EDS results of all samples indicated no statistically significant difference between the actual results and calculated values (Table S2). The observed outcome can be attributed to the significant porosity exhibited by the aerogel samples, facilitating the efficient loading of Fe₃O₄ NPs onto their surfaces [61]. The CA samples from WH in this study had a high porosity (98.2 %), which is comparable to the results obtained from a previous study concerning aerogels [11]. On the surface of the catalyst, the distribution of Fe and O was rather uniform during synthesis. The Fe₃O₄ NPs were effectively produced and incorporated into the surfaces of cellulose fibres.

Fig. 7.

EDS mapping of elements of samples; a) Fe₃O₄-3@CA-KV, b) Fe₃O₄@CA-PV, c) Fe₃O₄-3@CA-KF and d) Fe₃O₄@CA-PF.

TEM analysis (Fig. 8a) was used to investigate the microstructural characteristics of Fe₃O₄-3@CA-KF samples containing dark-field represented Fe₃O₄ NPs. By utilizing ImageJ software, the size distribution of Fe₃O₄ NPs extracted from TEM images was ascertained to be 5–15 nm for the spherical NPs (Fig. 8b). This result was entirely consistent with the XRD finding and SEM image of the material.

Fig. 8.

The size distribution histogram (b) and TEM image (a) of magnetite nanoparticles adorning the Fe₃O₄-3@CA-KF sample.

Fig. 9a and b illustrate the N₂ adsorption-desorption plots and BJH pore size distribution plots of the Fe₃O₄@CA hybrid aerogels fabricated using various methodologies. All the as-prepared samples’ physical parameters, including specific surface areas, pore diameters and pore volumes, were calculated through BET and BJH analyses. The N₂ adsorption-desorption data presented in Fig. 9a demonstrates that the samples synthesized in this study display a type IV isotherm, as classified by the IUPAC. The pore size distribution plot (Fig. 9b) of the BJH analysis revealed that the majority of pores were distributed within the mesoporous range of 15–60 Å, with an average pore diameter of 26.6 Å. Additionally, the BET surface area and total pore volume decreased in the following order: Fe₃O₄-3@CA-KF (42.9 m²/g, 0.113 cm³/g) > Fe₃O₄@CA-PF (41.2 m²/g, 0.100 cm³/g) > Fe₃O₄-3@CA-KV (40.4 m²/g, 0.083 cm³/g) > Fe₃O₄@CA-PV (21.4 m²/g, 0.113 cm³/g). The sample with kymene as the cross-linker was superior to the PVA-GA system in terms of results. In addition, freeze drying strengthened the structure of the material, enhancing PNP degradation activity.

Fig. 9.

N₂ adsorption/desorption isotherms (a), the BJH pore diameter distribution (b), point of zero charge (c) and VSM curve (d) of samples.

Fig. 9c and d shows that the PZC value and VSM curve of the samples had negligible differences. The PZC values of the samples fluctuated in a range of 6.55–6.70 in Fe₃O₄-3@CA-KF (6.55), Fe₃O₄@CA-PF (6.70), Fe₃O₄-3@CA-KV (6.66) and Fe₃O₄@CA-PV (6.63). The superparamagnetic nature of all Fe₃O₄@CA composites was evident from the magnetic hysteresis loops shown in Fig. 9d. The composites displayed zero remanence and coercivity [62]. The magnetisation saturation values of the Fe₃O₄@CA samples ranged from 20 emu/g to 30 emu/g. Besides, the Fe₃O₄@CA composites possess superparamagnetic and robust magnetic characteristics, which facilitate their separation from aqueous solutions in the presence of an external magnetic field. These results showed that the drying process and method only affect surface structure measurements, including specific surface area, pore diameter and pore volume, and do not considerably affect the magnetic properties and surface charge of Fe₃O₄-3@CA materials possibly because of the approximately equal Fe₃O₄ content.

3.3. Recyclability

The reusability of Fe₃O₄-3@CA-KF sample was tested (Fig. 10a). The Fenton reaction system can sustain approximately 80 % of PNP degradation rates throughout all five degradation phases, indicating that Fe₃O₄-3@CA-KF is highly recyclable and Fe₃O₄-3@CA-KF has excellent reusability. In addition, the Fe₃O₄-3@CA-KF composite was characterised by XRD, EDS and TEM after five cycles (Fig. 10b–d). There was no discernible alteration in the diffraction pattern of the Fe₃O₄ NPs, suggesting that the crystal structure of the catalyst remained unchanged after the treatment of PNP. In addition, the utilization of EDS mapping and mass percent elemental composition analysis (as depicted in Fig. 10b) indicated that there was no discernible disparity of statistical significance between the acquired data prior to and subsequent to the reaction. After five cycles of reuse, the composition and extensive dispersion of Fe₃O₄ NPs on the CA surface had a substantial impact on the activity of the catalyst. This finding implies that employing straightforward synthetic magnetic nanoparticles adorned with CA could serve as a viable approach to degrade PNP efficiently. Moreover, Fe₃O₄@CA catalysts are reusable following numerous reactions, satisfying the sustainable criteria. In contrast to previously reported magnetic catalysts, the present composite was synthesized by using a straightforward approach that is economically advantageous and exhibits remarkable performance and catalytic activity. Besides, F₃O₄ NPs can adhere to CA fibres, hence facilitating the straightforward retrieval and subsequent reuse of the catalyst through conventional filtering techniques.

Fig. 10.

Catalytic stability of the Fe₃O₄-3@CA-KF sample in the heterogeneous Fenton degradation of PNP over five consecutive cycles (a); the comparison of fresh and spent XRD patterns Fe₃O₄-3@CA-KF samples (b) and elemental mapping (c) and TEM image of spent Fe₃O₄-3@CA-KF sample.

4. Conclusion

Several approaches were used to fabricate of Fe₃O₄@CA catalytic samples, whereby the kymene and PVA-GA system were utilized as cross-linking agents. The influences of heat drying and sublimation drying techniques on the PNP treatment procedure were investigated. The experiment analysis revealed that the most active catalysts were the Fe₃O₄ NPs decorated on CA with kymene as a cross-linker and the synthetic sample freeze-dried twice. The sample Fe₃O₄-3@CA-KF catalyst was confirmed as the best material under the optimal conditions (H₂O₂ dosage of 5.0 g/L, pH 3, and the amount used of the Fe₃O₄ catalyst with a content of 0.4 g/g CA) after 100 min of reaction in the Fenton system. The degradation rate of PNP at a high concentration of 2000 ppm was 92.5 %. Furthermore, it is worth noting that Fe₃O₄@CA exhibited a sustained and noteworthy level of catalytic efficiency in the Fenton degradation process of PNP, even after undergoing five successive iterations. This observation highlights the favourable reusability of Fe₃O₄@CA, a crucial aspect to consider for its practical implementation in various applications. In addition, the magnetic properties of Fe₃O₄@CA were advantageous in enabling the efficient separation and subsequent reuse of Fe₃O₄@CA from wastewater. Consequently, this finding will renew interest in future dyestuff remediation strategies based on stable and robust catalytic activity in repetitive reaction cycles and straightforward, inexpensive, and environmentally friendly synthesis procedures.

Funding statement

We acknowledge the support of time and facilities from Ho Chi Minh University of Technology (HCMUT), VNU-HCM for this study.

Data availability statement

Data included in article/supplementary material/referenced in article.

Additional information

No additional information is available for this paper.

CRediT authorship contribution statement

Thi Thuy Van Nguyen: Writing – original draft, Visualization, Validation, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Quang Khai Nguyen: Software, Resources, Methodology, Investigation, Formal analysis. Ngoc Quan Thieu: Software, Resources, Methodology, Investigation, Formal analysis. Hoang Diem Trinh Nguyen: Writing – original draft, Visualization, Validation, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Thanh Gia Thien Ho: Writing – original draft, Visualization, Validation, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Ba Long Do: Software, Resources, Methodology, Investigation, Formal analysis. Thi Thuy Phuong Pham: Software, Resources, Methodology, Investigation, Formal analysis. Tri Nguyen: Writing – review & editing, Visualization, Validation, Supervision, Methodology, Investigation, Data curation, Conceptualization. Huynh Ky Phuong Ha: Writing – review & editing, Visualization, Validation, Supervision, Methodology, Investigation, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.