Photocatalytic removal of 17β-estradiol from water using a novel bimetallic NiCu/Nb₂O₅ catalyst

Abstract

The development of effective photocatalytic materials is essential for removing emerging pollutants from aqueous media, such as the hormone 17β-estradiol (E2). In this study, a novel photocatalyst based on niobium pentoxide (Nb₂O₅) functionalized with nickel (Ni) and copper (Cu) was synthesized for E2 removal. The NiCu/Nb₂O₅ photocatalyst was prepared using a facile wet impregnation method and characterized by various techniques. The incorporation of Ni and Cu into Nb₂O₅ reduced the band gap energy from 3.3 to 2.8 eV, enabling efficient utilization of visible light. Moreover, NiCu/Nb₂O₅ exhibited the highest E2 removal efficiency (82%) under UV-A-assisted conditions at a concentration of 1.5 g L⁻¹. The reaction kinetics were found to follow a second-order model with a rate constant of k = 0.0020 L g⁻¹ min⁻¹, and a plausible reaction mechanism was proposed. Through the study of radical elimination, it was proven that the radical oxidation reaction mechanism predominated in the reaction. The results of the toxicity assays, combined with the TOC parameter, demonstrated the efficacy of photocatalytic degradation in reducing E2. These findings demonstrate the great potential of the NiCu/Nb₂O₅ photocatalyst for removing persistent pollutants.

Introduction

The contamination of aquatic matrices by emerging pollutants is a critical and urgent issue, given the ability of these compounds to persist in the environment and their potential for negative impacts on ecosystems (Parida et al. 2021; Majumder and Gupta 2021; Nippes et al. 2022b). Furthermore, among the emerging pollutants, there are compounds known as endocrine disruptors (EDs), a class of substances capable of significantly impacting human health, increasing the risk of cancer, and causing disorders in the immune and nervous systems, even at low concentrations (Segner 2006; Orozco-Hernández et al. 2019).

Among the EDs, estrogenic steroids are highly dangerous, given the adverse effects they can cause in environmental ecosystems (Du et al. 2020). 17β-Estradiol (E2) is particularly concerning, as it is recognized as the most potent estrogen and has been widely detected in various aquatic matrices (Yu et al. 2019). E2 is a natural estrogen primarily used in hormone replacement therapy and birth control (Omar et al. 2016) and is primarily introduced into environmental waters through animal and human urine (Kabir et al. 2015; Orozco-Hernández et al. 2019). Some effects of this estrogen have been reported in the literature, such as increased cancer risk and deformities in humans, reproductive organ modifications, and sex alteration in fish (Purdom et al. 1994; Qing et al. 2022). The difficulty of conventional treatment processes in removing compounds like E2 highlights the need for alternative water treatment technologies that can complement traditional methods and eliminate this threat from effluents (Majumder and Gupta 2020).

In this context, we highlight the process of heterogeneous photocatalysis, a promising and environmentally friendly water treatment technology, due to its ease of use and ability to mineralize various water contaminants into non-hazardous compounds (Yin et al. 2010; Arora et al. 2022). This process employs semiconductors (such as BiVO₄, CeO₂, Fe₂O₃, MnO₂, TiO₂, ZnO, and Nb₂O₅) as fine solids or nanostructures (Boughelout et al. 2020). Nb₂O₅ is one of these semiconductors, which is a transition metal oxide of the n-type with a band gap energy value (Eg) of approximately 3.0–3.4 eV, making it an attractive candidate for photocatalysis (Su et al. 2021). Due to its good chemical stability, non-toxicity, and commercial availability (Yan et al. 2014), Nb₂O₅ finds extensive use in photocatalysis (Hu and Liu 2015; Souza et al. 2016; Goswami et al. 2022; Ücker et al. 2022a, 2023a, b; Yang et al. 2022; Lenzi et al. 2022; de Almeida et al. 2022; Liu et al. 2023). Additionally, Brazil is one of the main producers of niobium, making Nb2O5 a valuable material for use in water treatment applications (Nowak and Ziolek 1999).

Given the potential for photocatalytic application, Nb₂O₅ can still have its photocatalytic activity enhanced through the use of modification techniques such as functionalization (Lin et al. 2011; Furukawa et al. 2012; Peng et al. 2021; Xia et al. 2022), heterojunctions (Yan et al. 2014; da Silva et al. 2017; Sacco et al. 2020), and addition of transition metals (such as Ag, Fe, Cu, Ni, Pt, Au, etc.) (Vivek et al. 2022). Additionally, recently, it has been discovered that bimetallic nanostructures in semiconductors are effective in various catalytic processes, due to the synergistic effect of combining two metals (Riaz et al. 2020; Vivek et al. 2022; Nippes et al. 2022a). In particular, bimetallic copper and nickel catalysts have been reported as an efficient method to increase reaction efficiency, including in heterogeneous photocatalysis (Riaz et al. 2020). However, further studies are needed to understand the effects of this methodology and its application in the development of Ni and Cu bimetallic catalysts incorporated into Nb₂O₅ that are efficient in removing EDs such as E2. It is important to note that we did not find any literature reports on the photocatalytic degradation of the 17β-estradiol molecule using Nb₂O₅, either in its pure form or functionalized form.

Therefore, the objective of this study was to evaluate the degradation of the estrogen E2 using the semiconductor Nb₂O₅ in its raw form and also to explore the design of a bimetallic copper and nickel (Cu: Ni) catalyst supported on Nb₂O₅, establishing a relationship between the physicochemical properties of this new photocatalyst and its photocatalytic performance in the degradation of E2 in an aqueous medium, in a UV-A irradiation-assisted photocatalytic system. The efficiency of the photocatalytic process was also evaluated in terms of effluent toxicity, using two bioindicators (Artemia salina and Lactuca sativa), which provide highly relevant results from a toxicological perspective.

Materials and methods

Materials

For the synthesis of the material, nickel nitrate (Ni(NO₃)₂·6H₂O, 97%) and copper nitrate (Cu(NO₃)₂·3H₂O, 98%), both obtained from Sigma-Aldrich, were used along with niobic acid (HY-340) provided by Brazilian Company of Metallurgy and Mining (BCMM). The HY-340 was calcined to obtain niobium (V) oxide (Nb₂O₅). For toxicity tests, sodium chloride (NaCl, 99%) and potassium dichromate (K₂Cr₂O₇, 99%) from Synth were used. Lactuca sativa seeds and Artemia salina cysts were obtained from an agricultural supplier. For scavengers tests, the reagents ethylenediaminetetraacetic acid (EDTA, 99%), isopropanol ((CH₃)₂CHOH, 99.9%), and potassium iodide (KI, 99%), all obtained from Sigma-Aldrich, were used. Laboratory-grade water (LGW, 18 MΩ) was prepared using a Millipore purification system. The synthetic hormone 17β-estradiol (> 98%) was obtained from Sigma-Aldrich, and Table 1 present its main characteristics.

Table 1 Properties of 17β-estradiol (E2)

Methods

Synthesis

The catalyst used in this study was NiCu/Nb₂O₅ containing 1% Cu and 5% Ni by mass, which were determined by preliminary tests. The catalyst was prepared through a wet impregnation method by adding copper and nickel nitrates to niobium pentoxide. The resulting precursor was dried at 100 °C for 24 h and then calcined at 500 °C for 4 h, yielding a brown solid material.

Characterization

The catalysts were characterized by textural analysis using nitrogen gas (N₂) physisorption studies (Quantachrome, model NOVA-1200). The morphology of the materials was observed by scanning electron microscopy (SEM) (Shimadzu SS-550 software, Superscan SS-550), atomic force microscopy (AFM) conducted on an atomic force microscope (Shimadzu SPM-9700) in the threading mode using a Silicon-Pt/Ir coated tip (spring constant of 0.5–9.5 N/m and at a frequency of 45–115 Hz), and transmission electron microscopy (TEM) using a JEOL transmission electron microscope, model JEM-1400. X-ray fluorescence spectroscopy (XRF) was performed on a Rigaku ZSX Primus II equipment using oxide standards. Atomic absorption (AA) was performed on a Thermo Scientific iCE3000 to determine the metal composition.

The structural properties of the catalysts were analyzed by an X-ray diffractometer Shimadzu (model XRD-6000, 2Ө range 2 to 95°, CoKα radiation, with speed 2°/min, voltage 40 kV, and current 50 mA) and Fourier transform infrared spectroscopy (FTIR) on a Bruker Spectrometer (model Vertex 70 v, resolution 4 cm⁻¹, 128 scans, spectral range 4000–400 cm⁻¹).

The optical properties of the photocatalysts were examined using photoacoustic spectroscopy, performed on a spectroscopy module operated with 21 Hz frequency modulation in the range 200 to 800 nm, and normalized with ultra-pure carbon spectrum. The band gap energy was calculated using Tauc’s direct method. Photoluminescence spectrometer was performed on a PerkinElmer luminescence spectrometer, LS-50B, in the spectral region of 350 to 600 nm.

Reactional tests

The reaction system used in the experiments was isolated from the external environment by an aluminum metal box, equipped with two side fans for cooling (Fig. 1). The system consists of a batch-operated beaker equipped with a magnetic stirrer and five UV-A tubular lamps. The lamps measure 26 mm × 450 mm with a power of 45 W, providing a light intensity of 0.064 W m⁻². Each test used an initial concentration (C₀) of 10 mg L⁻¹ of E2 in 250 mL of solution, with the photocatalyst mass ranging from 1 to 2 g L⁻¹. The reactions were carried out at room temperature and at the natural pH of the solution. Samples were collected at specific time intervals (0, 15, 30, 60, 90, 120, 150, and 180 min) using a plastic syringe connected to a hose and passed through 0.22-μm Millipore membranes supported by stainless steel. The efficiency of E2 removal by adsorption (absence of light) and by photolysis (absence of catalyst) was also evaluated. The main active species of the photocatalytic reaction and degradation process were investigated by adding radical scavengers such as 1 mM EDTA (h⁺ scavenger), 10 mM isopropanol (${HO}^{\cdot }$ scavenger), and 10 mM potassium iodide (${HO}^{\cdot }$ and h⁺ scavenger) during the photocatalytic reaction experiments (Upreti et al. 2016). The residual E2 concentration in the treated solutions was quantified via HPLC (Shimadzu HPLC, 10A VP) using a C18 column (250 × 4.6 mm, 5 μm, Hichrom) (Isecke et al. 2018). An isocratic mobile phase composed of a mixture of acetonitrile/water 80: 20 (v/v) at a flow rate of 1.0 mL min⁻¹ was used. The UV detector was set at 210 and 280 nm. Prior to the analysis, a calibration curve was generated using E2 solutions with known concentrations (1–15 mg L⁻¹). The retention time for E2 was approximately 3.75 min. Analysis of the treated effluent generated after treatment included quantification of total organic carbon (TOC) following the methodology described in standard methods (APHA 1999). TOC was determined using a Shimadzu TOCL analyzer, and all analyses were performed in triplicate.

Fig. 1

Reaction module for photocatalytic tests with NiCu/Nb₂O₅ for E2 degradation in aqueous media

Kinetics study

To choose the best model that describes the best result obtained by this study, a mathematical method was proposed in which the reaction order was varied: zero order, ½ order, first order, and second order, according to Eq. 1, where n is the respective reaction order. The fitting of the models to the experimental data was performed by the numerical method in MATLAB® software, starting from the differential equation below. The differential equation was solved using the 4th-order Runge–Kutta.

$$\frac{d{C}_{A}}{dt}=-k{C}_{A}^{n}$$

(1)

Catalyst reuse

The NiCu/Nb₂O₅ photocatalyst was evaluated for its regeneration and subsequent reuse in new E2 degradation cycles. After each execution, the photocatalyst was removed from the reaction vessel by filtration and then washed thoroughly with deionized water and finally dried in an 80 °C oven overnight. Atomic absorption (AA) was performed on a Thermo Scientific iCE3000 to evaluate the metal lixiviation.

Toxicity test

Toxicity tests were conducted to determine the optimal degradation condition using Artemia salina and Lactuca sativa (lettuce), a species of microcrustacean and plant, respectively. These bioindicators possess fundamental characteristics to respond to environmental changes, even at low levels of contaminant concentration (Silveira et al. 2017).

The methodologies employed were based on the germination of Lactuca sativa seeds to calculate the relative germination rate (RGR) and the relative root length rate (RLR) and on the hatching of Artemia salina cysts to determine the lethal concentration (LC₅₀) after exposure to the treated effluent. Guidelines for these methodologies can be found in the work of Nippes et al. (2021b).

Results and discussion

NiCu/Nb ₂ O ₅ characterization

Figure 2a shows the diffractograms of both the Nb₂O₅ support and the NiCu/Nb₂O₅ catalyst. The diffractogram of Nb₂O₅ (PDF #27–1003) indicated an orthorhombic crystal structure. Wet impregnation did not alter the structure of the support, as the same Nb₂O₅ peaks were observed in the matrix. However, incorporation of the metals led to the formation of dispersed oxides on the surface. The catalyst displayed diffraction peaks for NiO (PDF #73–1523) which exhibited a cubic crystalline structure at 43.1° and 62.8°. Copper diffraction peaks were not detected due to its low content in the catalyst, as previously reported in the literature (Dancini-Pontes et al. 2015; Da Silva et al. 2016). Nonetheless, the presence of copper can be identified in the X-ray fluorescence and atomic absorption analyses presented in Table 2, which show the chemical composition of the samples.

Fig. 2

Diffractograms (a) and infrared spectroscopy profiles (b) of the Nb₂O₅ and NiCu/Nb₂O₅ samples

Table 2 Chemical composition of catalysts by XRF and AA techniques

Figure 2b shows the Fourier transform infrared spectroscopy (FTIR) analysis profiles of the support and catalyst. The samples showed very similar spectra, both in frequency and intensity. Both analyses point out bands around 560 cm⁻¹ and 870 cm⁻¹ characteristics of Nb₂O₅ of orthorhombic structure (Graça et al. 2013; Ramanjaneya Reddy and Chennakesavulu 2014), as observed in the X-ray diffractograms.

The support and the catalyst showed bands around 1630 cm⁻¹ and 3400 cm⁻¹, characteristic of the OH functional group (De Oliveira-Cantão et al. 2010), being the intensity of the bands in the catalyst lower, indicating the elimination of part of the OH functional groups from the surface by the incorporation of copper and nickel.

Textural analysis results were obtained from N₂ adsorption/desorption isotherms and are presented in Table 3 and Fig. 3. Comparing the support and the catalyst, it can be seen that impregnation results in a decrease in the specific area compared to that of the support, possibly due to the accommodation of oxides in internal regions to the pores, which justifies the reduction of pore volume (Furtado et al. 2011; Da Silva et al. 2016; Chan et al. 2017). Despite this, both remained as mesoporous material.

Table 3 Textural analysis by N₂ adsorption/desorption

Fig. 3

N₂ adsorption/desorption isotherms of the samples

It is observed in the isotherms that the impregnation of the copper and nickel oxides did not significantly alter the adsorption isotherms. The isotherms are of type IV, and the hysteresis is of type H2, indicating that the pore shapes are poorly defined (Steele 1983).

Considering the proportion of nickel and copper in the oxides observed in the XRF analysis, the mass content of nickel is 4.87%, and copper is 0.95%, close to the values obtained by atomic absorption. Thus, the results show that the impregnation of the metals of the active phase to the supports was quite satisfactory, considering the proximity between the theoretical and real values.

The morphology of the materials was observed through scanning electron microscopy (SEM) and atomic force microscopy (AFM), and the result is presented in Fig. 4a and b. Both images showed a heterogeneous surface formed by irregular particles, with some agglomerates but similar structures. This same configuration is observed in the MFA images for the samples, with a similar tendency for agglomeration of the particles. The result confirms that the incorporation of the metals did not result in significant changes in the morphology of Nb₂O₅.

Fig. 4

SEM at × 10,000 magnification and MFA of (a) Nb₂O₅ and (b) NiCu/Nb₂O₅ and MET of (c) Nb₂O₅ and (d) NiCu/Nb₂O₅ and respective particle size distributions

Figure 4c and d show the TEM images and particle distribution of Nb₂O₅ and NiCu/Nb₂O₅. Both samples showed similar morphology, confirming no changes in the support structure occurred after impregnation. A clustering of round-shaped particles with different contrasts is also observed in the catalyst sample. The particles observed in the support range of 10 to 70 nm with an average size of 40 nm, while the catalyst is in size range of 10 to 60 nm with an average size of 31 nm. This reduction in average particle size suggests that incorporating the metals may suppress the aggregation of the Nb₂O₅ particles. This effect may provide greater availability of the active sites for OH radical generation and enhance the photocatalysis process.

Figure 5a shows the direct Tauc plotting method for the photoacoustic spectra of the Nb₂O₅ and NiCu/Nb₂O₅ photocatalysts, along with the obtained band gap values for each catalyst in eV. The value found for the pure Nb₂O₅ was 3.3 eV, a value consistent with what is expected for this material (Silva et al. 2020). However, the band gap value was reduced to 2.8 eV for the modified photocatalyst, extending its photon absorption spectrum to the visible region (> 400 nm). This effect results from the addition of nickel and copper, which can improve the absorption ability of optical radiation and reduce the photocatalyst gap due to the synergistic effect of combining the two metals (Jin and Zhang 2020; Vivek et al. 2022).

Fig. 5

Photoacoustic (a) and photoluminescence (b) spectra of Nb₂O₅ and NiCu/Nb₂O₅ photocatalysts

The presence of the metals also caused optical changes in the material, as can be observed in the photoluminescence (PL) spectrum, shown in Fig. 5b. For the pure Nb₂O₅, it was visualized that a more robust and intense emission peak present in the UV region at 382 nm is due to the emission peak of Nb₂O₅. On the other hand, for the bimetallic material, it was noticed that a reduction in the intensity of the spectrum indicates that the addition of Ni and Cu inhibits the charge recombination process, which favors the performance of the photocatalyst, improving the photocatalysis process (Nippes et al. 2021a; Ücker et al. 2022b). Furthermore, a shift of the main emission peak to the more extended wavelength region was also detected, reinforcing the possible activation of the photocatalyst in the near visible region.

Reactional tests

Figure 6 shows the results obtained for the photocatalytic tests. To begin with, it is essential to evaluate the contribution of the photolysis and adsorption processes in removing E2. In percentage, photolysis was responsible for 9% of E2 removal, while adsorption reached values close to 14 and 20% for Nb₂O₅ and NiCu/Nb₂O₅, respectively. The result confirms that the addition of catalyst plus radiation is necessary to degrade the E2 hormone effectively.

Fig. 6

Results of the E2 hormone removal tests in aqueous media through adsorption, photolysis (a), and photocatalysis (b)

With the catalysts activated with UV-A radiation, it was observed that pure Nb₂O₅ could remove approximately 33% of E2. On the other hand, the catalyst modified with Ni and Cu at the same mass concentration reached 43% degradation. This proves that the functionalization of Nb₂O₅ with the metals efficiently improved the photocatalytic activity, evidenced by the reduction of its band gap energy, which allowed its activation at wavelengths in the UV-A region.

Comparing the catalyst loadings employed (g L⁻¹), it is evident that the increase from 1.0 to 1.5 g L⁻¹ causes an increase in E2 degradation (82%) due to the greater availability of active sites for the photocatalytic reaction. However, the increase to 2.0 g L⁻¹ reduces the catalytic activity. This indicates interference in the amount of catalyst applied due to the excess turbidity of many particles in the aqueous medium. This hinders the insertion of radiation into the system. With the result, we conclude that 1.5 g L⁻¹ is the most appropriate concentration for the system.

The behavior observed in the photocatalytic reaction was proven mathematically by the kinetic study. The values of the speed constant (k) and the fit (R²) are presented in Table 4. The model that best fitted the data was the second-order model (n = 2). This indicates that the initial speed of micropollutant consumption is higher in the first minutes of the reaction (Fig. 7). Over time, the changes in compound concentration are less significant, indicating some resistance, either in the production of hydroxyl radicals or in the access of E2 to these radicals. The highest value for k is for the reaction with the NiCu/Nb₂O₅ catalyst at a concentration of 1.5 g L⁻¹.

Table 4 Kinetic fitting results using numerical interaction method

Fig. 7

Kinetic fit for second-order model of the NiCu/Nb₂O₅ reaction test

Experiments with radical scavengers were conducted to identify the main active species in the photocatalytic reaction and degradation process of the E2 contaminant. The EDTA was used as an h⁺ scavenger, isopropanol as an ${HO}^{\cdot }$ scavenger, and potassium iodide (KI) as a scavenger for both ${HO}^{\cdot }$ and h⁺ (Fig. 8). The addition of scavengers followed the order of potassium iodide > isopropanol > EDTA > no scavenger, over a 180-min reaction with [NiCu/Nb₂O₅] = 1.5 g L⁻¹ and [E2] = 10 mg L⁻¹. The iodide ion is a scavenger that reacts with both positive holes (h⁺) and surface hydroxyl radicals (${HO}^{\cdot }$) (Martin et al. 1995), whereas isopropanol is known to eliminate only surface hydroxyl radicals (${HO}^{\cdot }$) (Van Doorslaer et al. 2012). Both scavengers reduce the number of available oxidizing species on the catalyst’s surface, thereby decreasing the degradation kinetics of E2. The addition of EDTA, which acts as an h⁺ scavenger (Upreti et al. 2016), showed the least effect on the E2 photodegradation rate, suggesting that surface hydroxyl radicals (${HO}^{\cdot }$) act as the main oxidizing agents in this photocatalytic process, and the radical oxidation reaction mechanism predominates.

Fig. 8

Photocatalytic degradation experiments in the presence of various scavengers

Reaction mechanism

Based on the results obtained in this study and the conducted literature review, the pure Nb₂O₅ semiconductor is activated under ultraviolet light (Silva et al. 2020). By introducing dopant ions into Nb₂O₅, the optical properties are improved by reducing the band gap value and shifting the semiconductor’s activation to the visible region. This optical enhancement of the semiconductor ensures a longer lifetime for electron–hole pairs, providing greater opportunities for charge carriers to reach the catalyst surface and interact with dissolved organic compounds (Vasu et al. 2022). Additionally, the incorporation of Cu²⁺ and Ni²⁺ ions into the Nb₂O₅ network generates an oxygen defect that acts as an energy site for dissociating organic pollutants on the catalyst surface.

Thus, when visible light is irradiated on the NiCu/Nb₂O₅ catalyst, electrons are excited to the conduction band, leaving a hole in the valence band (Eq. 2). These holes in the valence band form hydroxyl radicals through the decomposition of water or by reacting with OH⁻ (Eqs. 3 and 4). The electrons transferred to the surface of the conduction band react with dissolved oxygen, forming ${O}_{2}^{\cdot -}$ ions (Eq. 5) which further react with water to produce ${HO}^{\cdot }$ radicals. Ultimately, both ${HO}^{\cdot }$ and ${O}_{2}^{\cdot -}$ ions promote the mineralization of the E2 molecule (Eq. 6), transforming the pollutant into CO₂ and H₂O. Figure 9 depicts the suggested possible mechanism for the degradation of E2 molecules.

$${\mathrm{NiCuNb}}_{2}{\mathrm{O}}_{5}+\mathrm{hv}\to {\mathrm{NiCuNb}}_{2}{\mathrm{O}}_{5}({e}_{\mathrm{CB}}^{-}+ {h}_{\mathrm{VB}}^{+})$$

(2)

$${h}_{\mathrm{VB}}^{+} +{\mathrm{H}}_{2}\mathrm{O}\to {H}^{+}+ {HO}^{\cdot }$$

(3)

$${h}_{\mathrm{VB}}^{+} + {\mathrm{OH}}^{-}\to {HO}^{\cdot }$$

(4)

$${e}_{\mathrm{CB}}^{-}+ {O}_{2}\to {O}_{2}^{\cdot -}$$

(5)

$${HO}^{\cdot }+{O}_{2}^{\cdot }+E2\to {\mathrm{CO}}_{2}+{\mathrm{H}}_{2}^{\cdot }$$

(6)

where h⁺ is the photogenerated hole, e⁻ is the photogenerated electron, VB is the valence band, and CB is the conduction band.

Fig. 9

Possible electron transfer reaction mechanism

In most cases, the oxidation of organic pollutants using a combination of semiconductors and light involves the generation of the hydroxyl radical, which exhibits high oxidation potential (2.80 eV). Generally, the hydroxyl radical can oxidize organic compounds through three mechanisms: electron transfer, hydrogen abstraction, and electrophilic addition (Legrini et al. 1993). Although this article does not present the degradation pathways of the pollutant, the total organic carbon (TOC) parameter indicated that the E2 molecule degraded, forming possible inorganic substances such as CO₂ and H₂O, resulting in approximately 81% reduction in organic carbon during the pollutant degradation process.

Table 5 presents the results obtained by other studies on the degradation of the hormone 17β-estradiol. The results obtained in this work are similar to those of other studies; however, in some cases, the performance of the photocatalyst is inferior. Nevertheless, it is worth noting that, besides achieving a photocatalytic degradation of E2 higher than 80%, our study observed a significant removal of TOC, strongly indicating compound mineralization. Additionally, the reduction of E2 toxicity in water, as presented below, represents a significant advancement for the application of the photocatalysis process using NiCu/Nb₂O₅. The next steps involve studying the reaction variables and proposing an optimization to completely degrade E2 with a significant reaction kinetics.

Table 5 Catalytic degradation of E2 with different nanoparticles

Catalyst reuse

At this stage, we investigated the stability of our photocatalyst and its application in consecutive reaction cycles. This is a highly relevant step to verify the stability and recyclability of the produced material, as well as a crucial parameter for the techno-economic feasibility of the process. In this case, the NiCu/Nb₂O₅ sample was used for seven consecutive reaction cycles in the removal of the hormone E2 from an aqueous solution, with the results presented in Fig. 10. The results showed that the photocatalytic material exhibited good stability, maintaining a removal percentage of approximately 70%. Additionally, the resulting solution after the reaction was analyzed using atomic absorption technique, and no presence of the metals incorporated into Nb₂O₅ was detected. This confirms that the impregnation technique was efficient in fixing the metallic species onto the semiconductor, and the photocatalyst will not cause pollution of the solution with nickel or copper, thus can be safely applied in an aqueous medium.

Fig. 10

Recycling tests for the photocatalytic capacity of NiCu/Nb₂O₅

Toxicity test

The photocatalytic degradation process can generate subproducts with concerning toxicity, which must be taken into consideration to ensure its effective application. In order to evaluate the toxicity of the resulting effluent, we used two bioindicators, Lactuca sativa and Artemia salina.

Tests with Lactuca sativa were conducted in triplicate using negative control, positive control, untreated synthetic effluent samples, and synthetic effluent samples after treatment. The results presented in Table 6 indicate that 100% relative germination was achieved for the treated samples, while for the untreated sample, it was 60%. In terms of relative root growth, it was 91% for the treated sample and 70% for the untreated sample, demonstrating that the application of the photocatalytic process using NiCu/Nb₂O₅ as a catalyst was effective in reducing the toxicity of the effluent.

Table 6 Toxicity assessment against Lactuca sativa

The Artemia salina tests treatment. This is because the LC₅₀ parameter (Table 7), which corresponds also confirmed that the toxicity of the solution containing the hormone E2 was reduced after to the median lethal concentration, showed a value greater than 200 mg L⁻¹ and a mortality rate of 3%. Thus, it is possible to assert that the final effluent was non-toxic after the photocatalytic treatment. These results are of utmost importance because, as reported in the literature, the estrogen E2 has negative effects on aquatic organisms even at low concentrations (Ahmad et al. 2009; Orozco-Hernández et al. 2019). Therefore, the development of effective technologies to reduce the danger of this type of compound in water is of fundamental importance.

Table 7 Toxicity assessment against Artemia salina

Conclusion

The NiCu/Nb₂O₅ photocatalyst was synthesized via the wet impregnation method and exhibited irregular shapes with a surface area of 44.7 m² g⁻¹. The synthesis effectively incorporated the metals into the niobium surface, as confirmed by atomic absorption and FRX techniques, without altering the structure of the support. The diffractograms of the Nb₂O₅ support and the NiCu/Nb₂O₅ catalyst revealed the same orthorhombic crystal structure of Nb₂O₅, with the incorporation of dispersed oxides on the surface. The catalyst showed a higher rate of E2 degradation than the crude Nb₂O₅ under the same reaction conditions, with a maximum photocatalytic degradation of 82% for 10 mg L⁻¹ E2 at a catalyst loading of 1.5 g. Incorporating Ni and Cu into Nb₂O₅ led to an improvement in its photocatalytic performance and activation at wavelengths in the visible region (< 400 nm). The degradation reaction of E2 followed a second-order model. Toxicity tests have demonstrated a substantial decrease in the negative effects associated with E2. The TOC result corroborates these findings, providing evidence for the effectiveness of the NiCu/Nb₂O₅ catalyst in achieving this outcome. Finally, the results obtained in this study are crucial for future application of this photocatalytic system on a large scale, such as in municipal wastewater treatment plants.