Water contamination by organic and inorganic substances is a critical problem for humanity. Various chemical substances pollute water reservoirs. These chemical substances originate from human activities, such as agricultural, textile, food processing, mining, and automotive industries. Given this problem, it is necessary to develop effective, efficient technologies to remove or reduce these pollutants in aquatic environments. In the field of chemistry, advanced oxidation processes (AOPs) [1] are widely used for their effectiveness and efficiency in removing or reducing organic pollutants from aqueous media.

Advanced Oxidation Processes technology is based on the generation of positive (holes) and negative (electrons) charges at the surface of a photoactivated material, which is usually a semiconductor [2,3]. As the surface of a semiconductor material receives light, electrons e^- from the valence band are promoted to the conduction band, generating a hole with a positive charge (h⁺) that specifically reacts with the chemical groups of certain organic pollutants.

Based on this process, several studies present creative alternatives to better control the generation of the (e^-/h⁺) pair. For example, structural modification [4] (doping) of the classic semiconductor TiO₂ creates new electronic and optical properties that promote rapid degradation of pollutants. Within this same strategy, the intrinsic modification of zinc oxide with boron [5], g-C₃N₄ [6], Mg [7], and Al [8] has also been ingeniously reported as a way to enhance the photodegradation of pollutants in water.

Another widely reported strategy to increase the photodegradation of these pollutants is using materials with heterojunctions. Contact between two materials with distinct band gaps inhibits recombination of the (e^-/h⁺) pair, thereby increasing the amount of charge on the semiconductor surface for a longer time to react with the organic pollutant. Martins’ co-workers [9] reported that the yellow dye tartrazine was more readily removed from water using a semiconductor heterojunction based on SiO₂ and Fe₃O₄. The same conclusion was also reported for other heterojunctions formed between ZnO/BiOI [10], gC₃N₄/BiVO₄ [11], and gC₃N₄/BiOCl [12].

Both doping and the creation of heterojunctions in photocatalytic materials are widely used for degradation. Dyes containing the azo group (N=N) are extensively studied due to their widespread use in the food and textile industries [13]. In general, these dyes have high molecular mass and complex chemical structures, making them difficult to biodegrade. The presence of these macromolecules in the environment leads to serious problems for animals living in aquatic biomes, as they can even accumulate in the human body, causing health damage.

Critically, the development of strategies for the efficient removal of dyes from aqueous media often relies on expensive materials and time-consuming syntheses. It is desirable that a photocatalyst material has the advantages of being a low-cost precursor material, easy and rapid synthesis, being reusable, photostable, and biologically inert [14].

To generate motive power, there has been much recent discussion about the global need to shift from fossil fuels to biomass-based energy generation. Given the projected large-scale use of various biomasses, the waste generated could become a global problem. For example, when improperly disposed of, waste from sugarcane, corn, green coconut, soy, and sorghum can cause environmental contamination. The proposal of physical and chemical treatment routes to add value to this waste stands out as a possible solution. The literature [15,16,17] reports several studies highlighting the potential of certain waste products as adsorbents for decontaminating aqueous systems.

Using biomass derived from lignocellulosic constituents is an interesting strategy for the decontamination of aquatic environments because it involves an environmentally friendly adsorbent. On the other hand, there are some disadvantages, mainly due to limitations compared to adsorbents that depend on specific chemical interactions. To overcome this limitation in adsorption specificity, some studies convert the organic matrix of biomass into inorganic matrices via calcination. In this context, the generation of a silica-based material from biomass waste converges on using a material that is generally discarded into the environment, transforming it into a value-added [18] material at zero cost. For example, Fatimah and co-workers [19] studied the photocatalytic performance of the TiO₂/SiO₂ composite, in which the silica was derived from bamboo leaves. Additionally, Hajam and co-workers [20] used wood sawdust residue as biomass to produce bioactive silica during the photodegradation of the methylene blue dye. Similarly, silica from rice husks [21,22,23] is also widely used for the photodegradation of organic pollutants.

The production and consumption of green coconuts worldwide are mostly concentrated in Asia. However, Brazil is often cited as a world leader in producing this fruit. Locally, the improper disposal of green coconut waste is an environmental problem. Therefore, this biomass can be, and is, used as a precursor biomaterial for silica for the photodegradation of organic pollutants. The husks [24], endocarp [25], and mesocarp of coconuts are good biomass residues capable of generating silica with photocatalytic properties.

Given the large amount of coconut produced and the waste generated in Brazil, it becomes very obvious to answer two questions: (i) How to reuse this biomass, seen as an environmental problem, to depollute aqueous systems from complex organic molecules? (ii) Why then use the silica generated from this biomass?

The answer to the first question lies in the theme of coconut waste sustainability. Developing a physicochemical route that is easy to process and uses conventional procedures, such as chemical and thermal treatments, imparts new properties in the biomass or biosilica, creating a chemically active material for the decontamination of effluents. In addition to being sustainable and environmentally friendly, this approach is a much more economically viable option than using precious-metal semiconductors, which makes photocatalyst production much more expensive.

To answer the second question, this study will explain the physicochemical properties of SiO₂ from the perspective of its application in the decontamination of effluents. Silicon is the second-most-abundant chemical element on Earth, second only to oxygen [26,27]. Due to this abundance, it satisfies the need for non-exhaustible resources. Furthermore, regarding the formation of silicon oxide, it is classified as a semiconductor, exhibiting a band gap of approximately 9.0 eV [28,29]. Since the difference between the valence band (VB) and the conduction band (CB) is the critical factor for photocatalytic applications, this difference makes SiO₂ an unfavorable material for photogenerating surface charges. However, with the different synthesis routes, the literature reports that doping and the formation of heterojunctions allow for easy adjustment of the band gap value to favor the formation of the e^-/h⁺ pair.

Given the context of the two previous questions, this study reported the reuse of coconut mesocarp as a precursor biomass in the production of silica (SiO₂). This silica served as a support material for the rapid, easy co-precipitation of magnetite nanoparticles. The resulting composite nanomaterial (Np-Fe₃O₄/SiO₂) was used as a nanophotocatalyst in heterogeneous catalysis reactions to degrade tartrazine yellow dye molecules. The proposed material incorporates the principles of reusing coconut waste and magnetic nanomaterials, offering an efficient technology for the depollution of aquifer systems.

The transformation of coconut husk biomass from organic to inorganic was initially studied by infrared spectroscopy. Figure 2A shows the gradual reduction of this organic profile with the chemical (2 mol L^-1 NaOH) and thermal (calcination at 800°C) steps. In the infrared spectrum of the natural fiber (green line), we can see vibrational bands associated with the main chromophores of lignin, cellulose, and hemicellulose, as Table 1 shows [31,32]. When coconut fibers are subjected to chemical treatment (blue line) in a basic NaOH solution (2 mol L^-1), the stretching band of the C–O bond disappears [33] considerably, followed by a decrease in the intensity of the O–H bonds and in the aromatic region (800 cm^-1). These modifications in a basic medium indicate rupture of the plant cell wall [34], which is composed of the cellulosic units of coconut fiber, as similarly reported by Mulana and co-workers [35].

Complete modification of coconut fiber is achieved by subjecting the fiber treated in a basic medium to calcination at a temperature of 800°C for 1 hour in a muffle furnace (black line). Virtually all vibrational modes associated with lignin, cellulose, and hemicellulose disappear. In addition, the band associated with the silicate group (Si–O–Si) appears at 1060 cm^-1 [36], indicating the formation of silica.

It is also possible to track the morphological changes generated by chemical and thermal treatment of the fiber using FEG-SEM. The FEG-SEM image in Figure 2B shows fibers with lumens arranged in parallel, typical of fibers under natural conditions, as also discussed by Armad and co-workers [37]. However, these plant structures disappear upon treatment in a basic medium, leaving structures with flatter and smoother profiles (Figure 2C and Figure 2D). This proves that the basic medium breaks down cell walls and deconstructs the cellulosic units, as evidenced by FTIR. After calcination, the material consists of smaller micrometric particles and a random geometry with numerous scattered pores [9]. The images show silica particles with an average size ranging from 500 nm to 1 mm. These characteristics indicate the formation of silica (SiO₂), as can be seen in Figure 2E and Figure 2F, as similarly reported by Sarker and co-workers [38].

Chemical mapping of the natural fiber and silica was performed using EDS. By focusing the energy beam on point P2 in Figure 3A, corresponding to the natural fiber, the energy (in eV) of the chemical elements C, O, Al, and Si appears with higher intensities (Figure 3B). The elements Cu and K also appear with lower intensities. This composition is associated with the lignocellulosic structure and the plant’s minerals. However, when the energy beam was positioned at point P1 in Figure 3C, chemical mapping showed increased intensities for the elements O and Si, indicating the formation of silica. Other chemical elements also exhibit lower energy intensities, such as Mg, Al, and Fe, and may be associated with the formation of oxides of these metals (Figure 3D).

The synthesis of magnetic nanoparticles (Fe₃O₄) was carried out based on the coprecipitation method in the presence of silica [9], as detailed in the experimental section. Figure 4A shows photos of the final Np-Fe₃O₄/SiO₂ suspension in response to an applied external magnetic field (neodymium magnet). Within 70 seconds, all the suspended material responds to the magnetic field, indicating rapid magnetization of the nanoparticles. UV-Vis spectroscopy was used to confirm the formation of nanoparticles. Thus, Figure 4B shows the UV spectrum of the solution containing a mixture of iron (II) and iron (III) chloride salts, with an intense absorption band at nm, attributed to metal-ligand transitions [39] (red line). After the synthesis of the nanoparticles (black line), the spectrum shows a baseline evaluation due to light scattering from the nanoparticle surface, without the appearance of defined bands, which is typical of magnetite in suspension at nanometric scale. UV-Vis spectroscopy for the silica suspension (blue line) also did not reveal defined bands, as expected.

In Figure 4C, the UV-Vis absorption spectra enabled us to estimate the band gap between SiO₂ and Np-Fe₃O₄/SiO₂ materials. Considering the Tauc plot (αhν × E) n, where α is the absorption coefficient, h is Planck’s constant, v is the photon frequency, ν is equal to 2 (direct transitions), and E is the energy in eV. Extrapolating the tangent line to the curves, we obtained the band gap energies of 2.7 eV and 2.4 eV for SiO₂ and Np-Fe₃O₄/SiO₂, respectively. The value obtained for the SiO₂ band gap may be attributed to impurities present in the silica, originating from the organic matrix of the calcined biomass, as seen in the EDS spectra in Figure 3. The decrease of 0.3 eV for the Np-Fe₃O₄/SiO₂ can be explained by the formation of a heterojunction generated in the mixed composite.

The analysis of the crystalline behavior of the materials was performed by X-ray diffraction. In Figure 4D, the diffractogram associated with the natural fiber reveals an amorphous profile (without fine peaks), showing only two peaks (2θ = 18° and 25°) of low intensity and broad base, which may be related to crystalline cellulose [40] (green line). However, the silica obtained at 800°C presents a diffractogram containing low-intensity peaks at 2θ = 35°, 43°, 57°, and 63° (blue line). The presence of these peaks indicates the presence of silica in the quartz phase. Upon forming the nanoparticles (black line), the diffractogram shows higher-intensity peaks at 2θ = 23°, 31°, 32°, 47°, 52°, and 58°. These peaks are associated with the crystalline phase of magnetite. The presence of two peaks associated with the quartz phase of silica is also noted. This behavior suggests that magnetite formed on the silica, potentially leading to phase overlap, as reported by Martins and co-workers [9].

SEM confirmed the geometry of the Np-Fe₃O₄/SiO₂ system. Figure 5A and Figure 5B show FEG-SEM images at different magnifications and zoom levels. In these two images, the nanoparticles are aggregated, approximately spherical, and average 8 nm in size. It is possible to calculate the crystallite size (D) from the Scherrer equation (Equation 2) using the highest intensity indexed peak in the nanoparticle diffractogram. By using the constant K = 0.91 (spherical particles with cubic symmetry), the X-ray wavelength (λ = 0.54 nm), the full-height peak width (FWHM) β = 1.31, and the Bragg angle (θ = 0.31), the value of 6.1 nm was obtained, a result close to the information provided by FEG-SEM.

The chemical mapping performed in the highlighted red region of Figure 5C revealed a chemical composition with energies associated with the chemical elements O and Fe at higher intensities, as well as energy peaks associated with the chemical elements Cu, Al, and Si at lower intensities (Figure 5D). The low peak intensity of the energy associated with silicon corroborates the information seen in the X-ray diffractogram for the Np-Fe₃O₄/SiO₂ system, in which the magnetite phase overlays the quartz phase of silica.

The suspension of nanoparticles was applied in heterogeneous photodegradation reactions of the yellow dye tartrazine. Initially, the region of greatest UV absorption for the dye was analyzed. Thus, Figure 6A shows an intense absorption band of tartrazine at 426 nm, associated with the absorption of the azo group (N=N) present in the chemical structure of the dye [9,41], as Figure 6A shows. The photocatalytic performance of the systems Ttz + H₂O₂, Ttz + SiO₂ + H₂O_2, and Ttz + Np-Fe₃O₄/SiO₂ + H₂O₂ was evaluated using artificial UV light. Figure 6B shows the absorbance as a function of exposure time of the reaction between the materials and in the presence of hydrogen peroxide. After 75 min of reaction, the absorbances decreased from 0.3 (a.u.) to 0.2, 0.1, and 0.03, respectively, for systems Ttz + H₂O₂, Ttz + SiO₂ + H₂O_2, and Ttz + Np-Fe₃O₄/SiO₂ + H₂O₂. The photodegradation of tartrazine was also investigated in the presence of the two materials, without peroxide, as detailed in Figure 6B. In the absence of H₂O₂, the catalytic performance of the materials is less favorable for dye degradation, with degradation percentages of 4.1% for TTz + light, 6.2% for Ttz + SiO₂ + light, and 18% for Ttz + Np-Fe₃O₄/SiO₂ + light. However, the dye-degradation capacity against the Np-Fe₃O₄/SiO₂ system is three times greater and can be attributed to the increased surface area of the nanocatalyst.

When plotting the C/C₀ ratio versus time (Figure 6C), now for photodegradation in the presence of H₂O₂, the percentage degradation values for system Ttz + H₂O₂ are 22%, for Ttz + SiO₂ + H₂O₂ are 43%, and for Ttz + Np-Fe₃O₄/SiO₂ + H₂O₂ are 91%, after 75 minutes of reaction.

In a similar study, Martins and co-authors [9] evaluated the degradation yield of tartrazine using a mixed nanomaterial composed of Fe₃O₄/SiO₂. The authors produced polymorphic nanostructures measuring 200 nm, achieving 45% dye degradation in 120 min. In contrast, this work advances the synthesis and application of the same system for the removal of tartrazine. Specifically, the nanostructures formed are approximately spherical, with a monomorphic morphology and a diameter of 8 nm. Furthermore, nearly 100% of the tartrazine is degraded at a faster rate (75 min).

The Langmuir adsorption isotherm model was used to describe the adsorption equilibrium at the interface of the photocatalysts SiO₂ and Np-Fe₃O₄/SiO₂ and the tartrazine yellow dye. Considering that photodegradation occurs at the solid-liquid interface, the rate of tartrazine degradation depends on its adsorption/photodegradation on the surface of photocatalysts and the materials’ capacity to generate e^-/h⁺ pairs. Thus, the linearized Langmuir Equation 1 was used to determine these parameters. The graph in Figure 6D was obtained to depict the linearized isotherm for the reaction systems SiO₂, Np-Fe₃O₄/SiO₂, and the tartrazine yellow dye. As the experimental data (blue and black balls) follow a linear trend in the fit, this indicates that the adsorption capacities of systems SiO₂ and Np-Fe₃O₄/SiO₂ increase with increasing dye concentration. However, this increase in the adsorption capacity of tartrazine molecules is much greater when the system consisting of silica and magnetite reacts with hydrogen peroxide, which is known as the Fenton reaction. In this reaction, the magnetite surface reduces hydrogen peroxide to hydroxyl radicals, as given by Equation 3. The excess of these radicals, via photo-Fenton, accelerates the degradation of tartrazine molecules.

The q_e values were obtained using Equation 1, and the q_m, K_L, and R_L parameters were extracted from the graph in Figure 6D and are presented in Table 2. For the Np-Fe₃O₄/SiO₂ system, these parameters are practically double when compared to the adsorption of the dye at the SiO₂ interface alone.

The Langmuir R_L parameter is a dimensionless constant, known as the separation factor, which can be used to determine whether the dye adsorption/photodegradation on the surface of adsorbents is favorable (0 < R_L < 1). This parameter is calculated by Equation 4.

Based on the analysis of the kinetic parameters and photocatalytic performance, it is possible to propose a reaction mechanism for the two systems studied for the photodegradation of tartrazine in the presence of H₂O₂. When photodegradation occurs in the presence of silica, only the e^-h⁺ charges generated on this surface can generate radicals (HO^∙ and O2⋅−), leading to the degradation of 43% of the dye present in aqueous solution. However, in the presence of Np-Fe₃O₄/SiO₂, two factors may be at play: first, the generation of excess HO^∙ radicals via the Fenton reaction, which leads to the degradation of more dye molecules. Second, the structure of the material generated between silica and magnetite, which likely forms heterojunctions. The presence of heterojunctions [9] inhibits recombination between electrons and holes, favoring the degradation of more dye molecules, reaching up to 91%. Figure 7 is a schematic representation of this proposed photodegradation process. The advantage of this process lies in the short time it takes to virtually degrade all of the dye, the use of a magnetic photocatalyst, which allows it to be reused, and the use of hydrogen peroxide, as it is a non-toxic molecule that degrades easily in water in the presence of magnetite.

The synthetic engineering of new materials with photocatalytic properties addresses a global need to develop and improve technologies for removing organic molecules considered pollutants in aqueous media. This study highlighted the obtaining of biosilica from coconut mesocarp, which served as a support material in the synthesis of magnetic nanoparticles. The produced nanophotocatalyst showed 18% photodegradation of the tartrazine yellow dye, which increased to almost 100% removal when the photo-Fenton reaction was used. This material is very promising because it combines decontamination characteristics, such as being removable, inexpensive, and easy to produce.