Molybdenum Telluride-Promoted BiOCl Photocatalysts for the Degradation of Sulfamethoxazole Under Solar Irradiation: Kinetics, Mechanism, and Transformation Products

Abstract

This work examines the solar photocatalytic degradation of the antibiotic sulfamethoxazole (SMX) using molybdenum telluride (MoTe₂)-promoted bismuth oxychloride (BiOCl). Different loadings of molybdenum telluride in the 0–1% range on BiOCl were synthesized and evaluated. Although the presence of MoTe₂ did not alter either the adsorption capacity or the energy gap of BiOCl, the synthesized photocatalyst demonstrated higher photocatalytic activity due to the enhanced separation of photogenerated pairs. The 0.5MoTe₂/BiOCl photocatalyst achieved a kinetic constant nearly 2.8 times higher than that of pure BiOCl, leading to the elimination of 500 μg/L SMX within 90 min. The system’s performance was enhanced under neutral to acidic conditions and lower SMX concentrations. Based on experiments with radical scavengers, photogenerated holes appeared to be the dominant species, with the contribution of reactive species following the order ${\mathrm{h}}^{+}>{\mathrm{O}}_2^{•-}/{\mathrm{e}}^{-}>$¹${\mathrm{O}}_2{>\mathrm{H}\mathrm{O}}^{•}$. Interestingly, in different water matrices, photocatalytic activity was not diminished and even increased by 20%, likely because of the action of photogenerated holes and the selectivity of secondary generated radicals. The photocatalyst retained > 90% of its activity after three sequential experiments. Finally, four transformation products from SMX photodegradation were identified via UHPLC-TOF-MS, and a degradation pathway is proposed.

1. Introduction

Pharmaceutical (PhAC) consumption has shown an increasing trend in recent decades, driven by the need to treat both age-related and chronic diseases as well as changes in clinical practice. This trend has raised concerns, considering the fact that PhACs have been detected in low concentrations (ng/L) in aqueous media and have been recognized as a potential environmental threat to living organisms [1]. To date, PhACs have been identified globally in surface waters [2] including freshwater ecosystems [3], estuaries [4], marine environments [5], groundwater [6], and drinking water [7], highlighting the inadequacy of conventional wastewater treatment methods to eliminate them and the emerging need for the development of more efficient technologies. In this context, advanced oxidation processes (AOPs) have gained great attention, showing very promising results as a sustainable approach to address water pollution challenges [8]. AOPs are characterized by the generation of highly reactive species, particularly hydroxyl radicals (^•OH), which possess strong oxidizing power and can degrade a wide variety of contaminants into harmless byproducts, such as water and carbon dioxide [9]. Among them, heterogeneous photocatalysis is free from the need to add oxidizing agents, as radical species are generated by photon absorption after the activation of the photocatalyst, usually a semiconductor, from sun irradiation [10,11]. The photogenerated radical species (electrons and holes) usually recombine in the bulk of the photocatalyst, losing their energy in the form of heat. However, under specific conditions, they remain separated and migrate to the photocatalyst surface, where they can participate in redox reactions with adsorbed species (e.g., water, oxygen, or organic pollutants) [12,13]. As a result, photocatalysis is considered a “green” technology for water remediation, resulting in an enormous number of publications on the topic in recent years [14]. In specific, a great amount of research efforts have been devoted to the removal of PhACs from aqueous media through photocatalytic reactions, showing promising results [15].

However, despite the fact that the photocatalytic mechanism seems quite simple, achieving high yields is a multifactorial function highly dependent on the semiconductor/photocatalyst characteristics. The redox potential of conduction and valence bands in correlation with the oxidation potential of the organic component to be treated, the curvature at the liquid–solid interface of the photocatalyst energy bands, and the energy band gap value are the most critical [16]. Additional challenges in photocatalytic materials development include cost reduction and environmental safety [17]. In this context, the construction of novel photocatalysts with superior activity, stability, and selectivity is probably the most crucial step to increase photocatalytic efficiency.

Among the various photocatalytic materials studied to date, bismuth-based photocatalysts such as BiVO₄ [18]_, Bi₂O₃ [19], BiOBr [20,21], and BiOCl stand out because of their desirable features, including low cost, low toxicity, high stability, and promising photocatalytic performance. More specifically, because of their distinct electronic and hierarchical structure, BiOCl and BiOCl-based photocatalysts have generated significant interest within the scientific community [22]. Specifically, BiOCl has a layered structure belonging to the tetragonal crystal system, creating an internal electric field that promotes effective charge separation of photoinduced electron–hole pairs, while the ionic bonds that prevail in its structure make it chemically stable. Different synthetic strategies have been proposed to determine the absorption properties and morphological characteristics of BiOCl, such as the hydrolysis method [23,24], the hydro-/solvothermal methods [25], sol–gel [26], electrospinning [27], the solid reaction method [28], and so on, resulting in BiOCl photocatalysts rich in oxygen vacancies and active sites with desired electronic properties. One step further, the formation of heterojunctions can facilitate charge transfer and minimize undesirable electron–hole pair recombination thus significantly enhancing photocatalytic efficiency. Yu et al. synthesized Bi-modified Nb-doped BiOCl microflowers and examined their efficiency towards tetracycline degradation under visible light irradiation [29]. They concluded that the (Nb/Bi 1:7)/BiOCl sample exhibited the highest photocatalytic activity, resulting in 100% tetracycline degradation in a very short time (15 min). Wang et al. [30] proposed a BiOCl/TiO₂ series of photocatalysts that could efficiently degrade a mix of antibiotics (tetracycline and ofloxacin) in aqueous media. In addition, PVP-induced Bi₂S₃/BiOCl microflowers with open hollow structures were proposed by Liu et al. for ciprofloxacin degradation in water, achieving more than 70% removal after 120 min under visible light irradiation [31].

Molybdenum ditelluride (MoTe₂) belongs to the group of two-dimensional (2D) transition metal dichalcogenides (TMDs), which are known for their unique electronic, optical, and catalytic properties [32]. In general, TMDs are described by the formula MX₂, with M symbolizing a transition metal (Mo, V, Nb, etc.) and X symbolizing a chalcogen (S, Te, Se), and have attracted great interest in photocatalysis mainly due to their tunable bandgap [33]. For example, L. Li et al. synthesized and investigated the photocatalytic properties of WS₂/MoSe₂ composite materials, finding that they had semiconducting characteristics with a direct bandgap of 1.65 eV and suitable redox potentials for water splitting [34]. MoSe₂ was also combined with BiVO₄, showing high tetracycline (91.9%) removal from water under sunlight irradiation and hydrogen production in a photoelectrocatalytic system [35]. Yao et al. prepared BiOCl modified with MoSe₂ quantum dots, showing that the composite samples exhibited higher efficiency towards rhodamine B degradation than pristine BiOCl [36]. In contrast to the above-mentioned formulations, MoTe₂ formed by stacked layers of Mo and Te has been only slightly investigated in the field of photocatalysis [37]. In each layer of MoTe₂, molybdenum atoms are covalently bonded to tellurium atoms, while the layers themselves are held together by weak van der Waals forces, enabling the exfoliation of bulk MoTe₂ into layered structures. In addition, MoTe₂ is thermodynamically stable in both the semiconducting and semimetallic phases under ambient conditions, with a minimal energy difference between the two, allowing phase transitions to occur [32]. Its fast charge transport, adjustable bandgap, excellent carrier mobility, high sensitivity, optical responsiveness, large surface-to-volume ratio, and reactive edge sites contribute to its utility in applications such as energy, catalysis, sensors, photodetectors, and transistors [38,39,40]. In contrast, in the field of photocatalysis, there are only a few studies, dealing mainly with reductive reactions such as hydrogen production through water splitting [41] and CO₂ reduction [42] rather than the decomposition of persistent micropollutants, as investigated in this work. Moreover, the performance of the catalytic system in environmentally relevant matrices remains unexplored.

Under this perspective, the present study seeks to leverage the unique properties of both components, proposing a MoTe₂/BiOCl heterostructure as an innovative material for the degradation of sulfamethoxazole (SMX), a representative antibiotic, in aqueous media under simulated solar irradiation. By integrating the complementary properties of MoTe₂ and BiOCl, the heterostructure is expected to be characterized by reduced photogenerated species recombination rate and enhanced optical response. Therefore, this work focuses on the investigation of the photodegradation efficiency and the examination of the effect of operating parameters, with particular emphasis on the effect of water matrices. Finally, the investigation of the possible generation of transformation products of SMX provides additional information for an integrated assessment of the proposed system. As far as we know, this is the first time the proposed system has been systematically examined for the photocatalytic degradation of sulfamethoxazole, with its transformation products identified. Moreover, the study comprehensively evaluates the system under a wide variety of conditions, including environmentally relevant matrices, highlighting its robustness and potential for practical applications.

2. Results

2.1. Physicochemical and Optical Characterization

The specific surface area (SSA) of the photocatalysts was calculated using the Brunauer–Emmett–Teller (BET) method, based on nitrogen adsorption isotherms recorded at liquid nitrogen temperature (77 K). Pure BiOCl, as well as molybdenum telluride-modified composites, exhibited a specific surface area of 16 ± 2 m² g⁻¹, in good accordance with values reported in previous studies [43]. The addition of MoTe₂ had no discernible effect on the SSA values because of its low content and uniform dispersion within the BiOCl structure, which preserved the material’s structural and surface characteristics.

The bulk characteristics of the photocatalysts were determined using X-ray diffraction (XRD) analysis. The XRD patterns were recorded using a Cu Ka radiation source (λ = 1.54 Å) over a diffraction angle of 20–80°, and the results are presented in Figure 1. All observed diffraction peaks corresponded to the tetragonal structure of BiOCl, as indexed JCPDS No 6-249, confirming the retention of the crystalline framework of BiOCl across all samples. The low content of MoTe₂ (≤1%) in the composites accounted for the absence of its characteristic diffraction peaks, as the signals were either below the detection limit of XRD or overlapped by the dominant BiOCl reflections. Additionally, as XRD primarily provides bulk structural information rather than surface-specific insights, the characteristic peaks of MoTe₂ were undetectable, presumably because of its highly homogeneous dispersion on the BiOCl surface. However, it was observed that with the addition of MoTe₂, these peaks gradually become sharper, indicating that the primary crystallite size increased. Indeed, the primary crystallite size estimated by the Scherrer equation using the (101) plane increased from 16 nm for pure BiOCl to 19 nm for the 0.50MoTe₂/BiOCl sample.

Figure 2 shows characteristic TEM images of 0.50MoTe₂/BiOCl. It is observed that BiOCl consists of well-shaped nanoplates with diameters ranging from 42 to 125 nm and an average thickness equal to ~13 nm. MoTe₂ was detected on the surface of BiOCl in the form of nanoparticles of irregular size characterized by lower crystallinity.

The optical properties of the pristine sample and the modified photocatalysts were thoroughly investigated using diffuse reflectance spectroscopy (DRS), and the results are summarized in Figure 3. The absorbance spectrum revealed an absorption edge at approximately 470 nm, corresponding to the onset of electronic transition in the materials.

The band gap energy (E_g) of the photocatalysts was determined, employing the Tauc method, by plotting (αhν)^1/2 vs. hν, assuming an indirect electronic transition [44]. Specifically, E_g was calculated by extrapolating both the linear region of the absorption edge and the initial offset region, identifying the intersection point as the band gap energy (inset graph).

For all samples, the calculated E_g was approximately 2.55 eV, indicating a minimal variation between the pristine BiOCl and the MoTe₂-modified photocatalysts. These findings suggest that BiOCl retained its intrinsic electronic properties, with the addition of MoTe₂ having a negligible effect on the bulk electronic structure, in good agreement with similar studies [45].

2.2. Photocatalytic Results

2.2.1. Photocatalytic Performance of MoTe₂/BiOCl

To evaluate the photocatalytic activity of BiOCl, and that of the enhanced MoTe₂/BiOCl samples (0.25–1% wt. MoTe₂), photocatalytic experiments for SMX oxidation were conducted under simulated solar light with a radiation intensity of 7.3 × 10⁻⁷ einstein/(L.s). The solution was stirred in the dark for 15 min to allow the SMX to reach adsorption/desorption equilibrium on the photocatalyst surface. All photocatalysts demonstrated low adsorption capacity for SMX, with adsorption not exceeding 10% of the initial SMX concentration (500 μg/L) in any case.

As shown in Figure 4, the presence of MoTe₂ as a cocatalyst with BiOCl enhanced the photocatalytic performance of the pure BiOCl for the SMX degradation, likely because of the effective separation of photogenerated electron–hole pairs (h⁺-e⁻) [46].

Notably, as the MoTe₂ loading increased from 0 to 0.25 wt.% and then to 0.5 wt.%, the apparent rate constant (kₐₚₚ) increased 1.37-fold and 2.82-fold, respectively, compared with the kₐₚₚ of pure BiOCl. Similarly, SMX removal improved from 376 μg/L to 415 μg/L and 480 μg/L after 60 min of reaction. However, further increasing the MoTe₂ loading resulted in decreases in SMX removal (441 μg/L at 60 min) and kₐₚₚ, though both remained higher than those of pure BiOCl.

An optimal loading is commonly observed in photocatalysis, beyond which degradation decreases because of the blockage of active catalyst sites and the increased recombination of photogenerated electron–hole pairs [47,48].

For instance, Ma et al. [47] reported that the optimal loading of the carbon materials (C60) on BiOCl for the photocatalytic degradation of phenol was 1.0 wt.%, while Yan et al. [49] found 15 wt.% biochar in BiOCl to be the optimal loading. In others studies, He et al. [50] and Ioannidi et al. [45] reported optimal loadings of 0.75 wt.% BiOCl in BiVO₄ and 0.5 wt.% MoB in BiOCl, respectively.

To investigate the enhanced photocatalytic activity of the MoTe₂/BiOCl heterojunctions, photoluminescence (PL) spectroscopy was performed. Figure 5 depicts the PL spectra of pure BiOCl, 0.25 MoTe₂/BiOCl, 0.50 MoTe₂/BiOCl, and 1.00 MoTe₂/BiOCl. The luminescence intensities of the samples increased in the following order: 0.50 MoTe₂/BiOCl < 0.25 MoTe₂/BiOCl < 1.00 MoTe₂/BiOCl < BiOCl. Pure BiOCl exhibited the highest PL intensity, indicating a higher recombination rate of electron–hole pairs compared with the composite materials. In contrast, the PL peak intensities of the MoTe₂/BiOCl heterojunctions were lower, suggesting that the addition of MoTe₂ effectively suppressed the recombination of electron–hole pairs, thus enhancing their photocatalytic efficiency. The fact that 0.50 MoTe₂/BiOCl exhibited the lowest PL intensity was consistent with its status as the best-performing sample.

2.2.2. Effect of Initial Concentration of 0.5 MoTe₂/BiOCl and SMX

Figure 6A illustrates the computed apparent rate constants (k_app) and SMX removal after 45 min of reaction for various initial concentrations of 0.5 MoTe₂/BiOCl. Interestingly, although k_app increased with rising catalyst concentration from 0 to 1000 mg/L, this increase was not significant for 250 and 500 mg/L of 0.5 MoTe₂/BiOCl, while for 1000 mg/L of 0.5MoTe₂/BiOCl, the k_app increased only 1.15-fold compared with that for 500 mg/L 0.5 MoTe₂/BiOCl.

Regarding SMX photolysis, it achieved only 21% SMX removal at 45 min (Figure 6A and Figure S2). However, with the addition of 100 mg/L of 0.5 MoTe₂/BiOCl, SMX removal significantly improved, reaching 83% after 45 min photocatalytic reaction (Figure 6A and Figure S2). Further increases in 0.5MoTe₂/BiOCl concentration resulted in smaller improvements, achieving 91%, 92%, and 96% for 250 mg/L, 500 mg/L, and 1000 mg/L 0.5MoTe₂/BiOCl, respectively (Figure 6A and Figure S2).

It is widely recognized that increasing the catalyst concentration boosts the degradation rate because of the higher availability of active sites for the reaction. This trend persists until the degradation rate reaches its peak. At this point, the catalyst concentration provides enough active sites to absorb all available photons. Beyond this level, the degradation rate stabilizes [51,52]. This optimal concentration appears to be beyond the range of catalyst concentrations tested in this study but is likely close, given the minimal increase in SMX degradation observed after 250 mg/L of 0.5 MoTe₂/BiOCl.

Considering that nearly complete SMX removal was achieved within 45 min using 500 mg/L 0.5 MoTe₂/BiOCl, along with the cost of the catalyst and the impact of real water matrices on process efficiency, all subsequent experiments were conducted with 500 mg/L of the photocatalyst.

Next, several experiments were conducted with different initial concentrations of SMX under simulated solar illumination. As Figure 6B demonstrates, as the SMX concentration increased, the k_app value declined, indicating that the reaction did not follow true first-order kinetics but rather pseudo-first-order kinetics. This phenomenon is well-known in AOPs and is explained by the fact that, under specific experimental conditions, reactive species are produced at a constant rate. For low concentrations of pollutants, these reactive species are in excess. However, as the pollutant concentration increases, the reactive species likely become the limiting factor.

Additionally, higher concentrations of pollutants result in increased production of transformation products, leading to competition between the transformation products and pollutant molecules for the active sites of the catalysts and the reactive species [43,53,54]. Nevertheless, for the proposed photocatalytic system, as the SMX concentration increased from 250 μg/L to 500 μg/L and 1000 μg/L, the k_app value was reduced only 1.17-fold and 1.59-fold, respectively, while remaining within the same order of magnitude. This is an important finding, since, as shown in Figure 6C, the proposed photocatalytic system is capable of removing nearly all the added SMX. Specifically, 250 μg/L, 479 μg/L, and 919 μg/L of SMX were removed within 60 min when the initial concentrations were 250 μg/L, 500 μg/L, and 1000 μg/L, respectively, which is highly satisfactory considering the low concentrations at which SMX is often detected in environmentally relevant matrices.

2.2.3. Effect of Initial pH and Evaluation of Reactive Species Contribution Using Scavengers

Continuously, the impact of solution pH on SMX degradation and adsorption was investigated, and the results are shown in Figure 7A. Acidic and alkaline conditions were adjusted by adding H₂SO₄ and NaOH, respectively. The pH values were monitored during both adsorption and photodegradation of SMX. The final pHs of the adsorption experiments were found to be 3.2, 5, and 8 for initial pH values of 3.2, 5.5, and 9.1, respectively, while the final pHs for the photocatalytic degradation of SMX were 3, 4.5, and 5 for the same initial pH values. The pH reduction observed when the initial pH was 9.1 was likely due to the formation of acidic transformation products rather than the acidic character of BiOCl (pzc ≈ 2) [55].

It is worth noting that pH did not practically affect SMX adsorption, which was negligible in all cases. Furthermore, the solution pH had only a minor impact on SMX photodegradation, with a slight reduction observed under alkaline conditions. This was likely due to the weaker activity of the photogenerated reactive species as the solution pH increased [56]. However, in all cases, more than 85% of 500 μg/L SMX was successfully removed after 45 min of photodegradation.

Interestingly, the performance of treated BiOCl varied depending on the cocatalyst used. For instance, pure BiOCl showed low pH sensitivity, as in this study, but exhibited lower photocatalytic efficiency for SMX oxidation [55]. In contrast, the photocatalyst MoB/BiOCl demonstrated strong pH sensitivity for losartan photooxidation under solar light, achieving very high efficiency in acidic environments but low efficiency in alkaline conditions, as its k_app decreased from 0.9529 min⁻¹ to 0.0541 min⁻¹ [45].

In subsequent experiments, the roles of ${\mathrm{H}\mathrm{O}}^{•}$, ${\mathrm{h}}^{+}$, and ¹${\mathrm{O}}_2$ and ${\mathrm{O}}_2^{•-}$ were examined by using 2 mM tert-butanol (t-BuOH) [57], potassium iodide (KI) [58], and sodium azide (NaN₃) [59], respectively, as radical scavengers, and the results are depicted in Figure 7B. The addition of NaN₃ and t-BuOH slightly reduced the degradation rate of SMX, while the presence of KI caused a higher inhibition of SMX removal but did not completely inhibit it. This suggests that, besides the photogenerated holes, other reactive species such as ${\mathrm{O}}_2^{•-}$ contributed to the photodegradation of SMX. To confirm the critical role of photogenerated electrons in the in the examined photocatalytic system, and thus the indirect formation of ${\mathrm{O}}_2^{•-}$, an experiment was performed in the presence of N₂ gas [60]. Additionally, an experiment in the presence of 2 mM p-benzoquinone (p-BQ) as an ${\mathrm{O}}_2^{•-}$ scavenger was carried out [61]. The results of these experiments are demonstrated in Figure 7B. It was observed that both N₂ gas and p-BQ inhibited the SMX degradation, with the latter causing even greater inhibition than KI. This could be attributed to the fact that BQ reacts not only with superoxide radicals but with conduction band electrons in the semiconductor, leading to competition between these processes and the formation of 1,4-hydroquinone [61]. Consequently, BQ competes with ${\mathrm{H}\mathrm{O}}^{•}$, ${\mathrm{O}}_2^{•-}$, and electrons for reactions, which can notably hinder the photodegradation of the SMX even more.

Specifically, the k_app values were 0.0417 min⁻¹, 0.0335 min⁻¹, 0.0169 min⁻¹, 0.0063 min⁻¹, and 0.0050 min⁻¹ in the presence of t-BuOH, NaN₃, N₂ gas, KI, and p-BQ, respectively, while the k_app value without any scavenger was 0.0586 min⁻¹. To further elucidate the role of ${\mathrm{H}\mathrm{O}}^{•}$, an experiment was conducted at 4 mM t-BuOH (Figure S3). Although a stronger obstruction of SMX photocatalytic degradation was observed (80% SMX removal at 90 min) compared with 2 mM t-BuOH (96% SMX removal at 90 min), it was still less than the inhibition observed with N₂ gas, KI, and p-BQ.

According to the literature, MoTe₂ has a direct energy gap equal to 1.01–1.5 eV [41,62,63], and its valence band (VB) is located around 0.7–0.8 eV [64]. Gagné et al. [65] reported that the redox potential of SMX was 0.9 eV. Additionally, the bandgap of BiOCl depended on its preparation method, and for the preparation method used in this work, the band gap value was approximately 2.6 eV, with a VB of $~$2.5 [66]. Although we did not have access to UPS to experimentally determine the positions of the valence and conduction bands, the E_VB and E_CB were estimated using the Mulliken electronegativity theory. The E_CB was determined to be −0.35 and 0.736 eV, and the E_VB, 1.15 and 3.33, for MoTe₂ and BiOCl, respectively (as shown in Scheme 1).

Taking into account the above information, along with the experimental results in the presence of scavengers, a possible Z scheme photocatalytic mechanism for SMX degradation using 0.5MoTe₂/BiOCl is depicted in Scheme 1. Specifically, MoTe₂ acts as an electron trap, enhancing the separation of photoinduced charge carriers through electron cascading from BiOCl. The photogenerated ${\mathrm{h}}^{+}$ may react with ${\mathrm{H}\mathrm{O}}^{-}$ or ${\mathrm{H}}_2\mathrm{O}$ to produce ${\mathrm{H}\mathrm{O}}^{•}$ or directly oxidize SMX, while the electrons in the CB of BiOCl transfer to the VB of MoTe₂. The role of MoTe₂ as an electron trap/acceptor has also been stated by Ali et al. [62], who studied the efficiency of Te−MoTe₂−MoS₂/ZnO photocatalysts for H₂ production.

The more negative conduction band (CB) potential of MoTe₂ relative to ${\mathrm{O}}_2$/${\mathrm{O}}_2^{•-}$(−0.33 eV, pH = 7) allows electrons to reduce ${\mathrm{O}}_2$, producing ${\mathrm{O}}_2^{•-}$. Then, ${\mathrm{O}}_2^{•-}$ may react directly with SMX or/and with ${\mathrm{h}}^{+}$ in the VB of BiOCl to form ¹${\mathrm{O}}_2$. The potential contribution of reactive oxygen species follows the order ${\mathrm{h}}^{+}>{\mathrm{O}}_2^{•-}/{\mathrm{e}}^{-}>$ ¹${\mathrm{O}}_2{>\mathrm{H}\mathrm{O}}^{•}$.

All the degradation pathways mentioned above appear to play a significant role in the decomposition of SMX, as suggested by the transformation pathway outlined in Section 2.3.

It is interesting that the dominant reactive species vary depending on the modification of BiOCl. For instance, in our previous work [45], which examined the photocatalytic degradation of losartan using the photocatalyst 0.5 MoB/BiOCl, it was found that the dominant reactive species was singlet oxygen, followed by ${\mathrm{O}}_2^{•-}$ and ${\mathrm{h}}^{+}$. In contrast, Shen et al. [67], highlighted the significant role of e⁻ in the photocatalytic removal of NO using the Mn₃O₄/BiOCl photocatalyst. In another study, Pan et al. [15] reported that for the photocatalytic degradation of phenol under UV light, the reactive species contribution followed the order ${\mathrm{h}}^{+}>{\mathrm{O}}_2^{•-}$>${\mathrm{H}\mathrm{O}}^{•}$ in the presence of CdS/BiOCl and ${\mathrm{O}}_2^{•-}>{\mathrm{h}}^{+}>{\mathrm{H}\mathrm{O}}^{•}$ with CdS/CQDs/BiOCl.

2.2.4. Effect of Water Matrix

Subsequently, the photocatalytic performance of 0.5MoTe₂/BiOCl for SMX degradation in various water matrices was investigated, and the results are presented in Figure 8A,B.

Table 1. Physicochemical characteristics of the water matrices used in this work.

Parameter	Wastewater (WW)	Bottled Water (BW)
pH	8.5	7.4
Conductivity (20 °C), [μS/cm]	934	513
Total dissolved solids (TDS), [mg/L]	654	312
Total suspended solids (TSS), [mg/L]	22	-
Total hardness (CaCO3), [mg/L]	287	260
Chemical oxygen demand, [mg/L]	48.5	-
Total organic carbon, [mg/L]	4.7	-
Chlorides (Cl−), [mg/L]	262	9.9
Bicarbonates (HCO3−), [mg/L]	278	263.1
Sulfates (SO42−), [mg/L]	62.3	29
Phosphates (PO4−) [mg/L]	14.9	-
Nitrates (NO3−), [mg/L]	2.3	10.1
Bromides (Br−), [mg/L]	165.6	-
Ca+2, [mg/L]	112	92.7
K+, [mg/L]	15.4	0.65
Na+, [mg/L]	76.3	5.5
Mg2+, [mg/L]	-	7.1

provides the physicochemical characteristics of the bottled water (BW) and wastewater used in this study.

Surprisingly, the degradation rate of SMX remained constant or increased in all synthetic and real water matrices, achieving complete removal within less than 60 min. WW unexpectedly exhibited the highest k_app. This behavior can possibly be attributed to the presence of bicarbonate (${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}$) and ${\mathrm{C}\mathrm{l}}^{-}$, which generally have diverse roles. In some cases, these inorganic ions negatively impact performance processes [45,48,68], while in others, their presence enhances the degradation process [45,48,52,69,70,71,72].

This positive effect is likely due to the formation of more selective reactive species (e.g., ${\mathrm{C}\mathrm{O}}_3^{•-}$, ${\mathrm{C}\mathrm{l}}^{•}$), even though these species have lower redox potentials than ${\mathrm{H}\mathrm{O}}^{•}$ (${\mathrm{C}\mathrm{O}}_3^{•-}/{\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}$E° = 1.78 eV vs. NHE and ${\mathrm{C}\mathrm{l}}^{-}$/${\mathrm{C}\mathrm{l}}^{•}$ E° = 2.43 eV vs. NHE) (reactions 1–7) [73,74,75,76].

$${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}+{\mathrm{h}}^{+}\to {\mathrm{C}\mathrm{O}}_3^{•-}+{\mathrm{H}}^{+}$$

(1)

$${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}+{\mathrm{H}\mathrm{O}}^{•}\to {\mathrm{C}\mathrm{O}}_3^{•-}+{\mathrm{H}}_2\mathrm{O}$$

(2)

$${\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}+{\mathrm{O}}_2^{•-}\to {\mathrm{C}\mathrm{O}}_3^{2-}+{\mathrm{H}\mathrm{O}}_2^{•}$$

(3)

$${\mathrm{C}\mathrm{l}}^{-}+{\mathrm{h}}^{+}\to {\mathrm{C}\mathrm{l}}^{•}$$

(4)

$${\mathrm{C}\mathrm{l}}^{-}+{\mathrm{H}\mathrm{O}}^{•}\to {\mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}}^{•-}$$

(5)

$${\mathrm{C}\mathrm{l}}^{-}+{\mathrm{H}\mathrm{O}}^{•}\to {\mathrm{H}\mathrm{O}}^{-}+{\mathrm{C}\mathrm{l}}^{•}$$

(6)

$${\mathrm{C}\mathrm{l}}^{-}+{\mathrm{C}\mathrm{l}}^{•}\to {\mathrm{C}\mathrm{l}}_2^{•-}$$

(7)

The beneficial effects of HA are primarily linked to its role as a sensitizer through charge transfer interactions [77]. Additional factors highlighted in the literature include the production of reactive oxygen species such as ${\mathrm{O}}_2^{•-}$ via HA photolysis [78,79] and the sequestration phenomenon, which reduces the escape of pollutant molecules into the bulk solution, increasing their chances of reacting with the active species generated on the photocatalyst’s surface under solar light because of HA adsorption [80].

Thus, the high efficiency observed in BW and WW resulted from a synergistic effect of various reactive species formed by inorganic and/or organic loadings, in combination with the photogenerated holes and ${\mathrm{O}}_2^{•-}$, which, according to Figure 7B, were the dominant reactive species.

Table 2. Photocatalytic removal of SMX by different photocatalysts.

Photocatalysts	[Catalyst], mg/L	[SMX], μg/L	Type of Irradiation	Removal	Ref.
BiOCl/g-C3N4/Cu2O/Fe3O4	200	25,325	Visible light (800 W Xe lamp)	100% at 60 min in UPW	[81]
Cu2O/BiOBr	1000	20,000	Solar light (250 W Xe lamp)	90.7% at 30 min in UPW	[82]
LaFeO3/Ag3PO4@GO	-	20,000	300 W Xe lamp	80.4% at 60 min in UPW	[83]
Bi/Bi2WO6/TiO2	300	20,000	300 W Xe lamp (λ > 300 nm)	96% at 60 min in UPW 86% at 60 min in 10 mg/L HA	[84]
ZnIn2S4/g-C3N4	200	15,000	Visible light	89.4% at 120 min in UPW	[85]
2.4% wt. Pd/BiVO4	500	10,000	Visible light (300 W Xe lamp)	100% at 200 min in UPW	[86]
8% Cd doped γ-Bi2MoO6	1000	5000	500 W long-arc xenon lamp	100% at 210 min in UPW	[87]
11% wt. Fe2O3/g-C3N4	300	10,000	Visible light (350 W Xe lamp)	40% at 40 min in UPW at pH 9	[88]
0.5% wt. MoTe2/BiOCl	500	500	Solar light (100 W Xe lamp)	96% at 90 min in UPW and 100% at 60 min in WW	This study

presents a comparison of SMX photocatalytic degradation with other related studies, while

Table 3. Efficiency comparison of several photocatalysts based on BiOCl.

Photocatalysts	[Photocatalysts], mg/L	Compound, μg/L	Type of Irradiation	Removal	Ref.
BiOCl/MoSe2−30% wt.	1000	1 SD, 20,000	Solar light (300 W Xe lamp)	100% at 120 min in UPW	[89]
1% wt. Bi2WO6–BiOCl	1000	Phenol, 20,000	Solar light (500 W Xe lamp)	93% at 5 h in UPW	[90]
0.5% wt. BiOCl/SnO2	500	2 RhB, 4790	Visible light (500 W Xe lamp)	80% at 10 h in UPW	[91]
MoB/BiOCl	500	3 LOS 500	Solar light (100 W xe lamp)	100% at 7.5 min in UPW 20% at 90 min in WW	[45]
0.25% wt. MoS2/BiOCl	1000	4 VLS, 500	Solar light (100 W Xe lamp)	10% at 120 min in WW	[92]
Bi/BiOI/BiOCl	1000	5 TC, 30,000	Visible light (300 W Xe lamp)	80% at 60 min in UPW	[93]
AgI/BiOCl/biochar	500	6 EE2, 3000	Visible light (500 W Xe lamp)	98.6% at 12 min in UPW	[94]
10% wt. Ni-MOF/BiOCl	2500	5 TC, 10,000	32 W UV lamp	60% at 180 min in UPW	[95]
0.5 MoTe2/BiOCl	500	SMX, 500	Solar light (100 W Xe lamp)	96% at 90 min in UPW and 100% at 60 min in WW	This study

¹ Sulfadiazine: SD, ² rhodamine B: RhB, ³ losartan: LOS, ⁴ valsartan: VLS, ⁵ tetracycline: TC, ⁶ 17α-ethinyl estradiol: EE2.

evaluates the efficiency of 0.5MoTe₂/BiOCl compared with other BiOCl-based photocatalysts for the photocatalytic degradation of various pollutants. As shown in Table 2 and Table 3, although many photocatalysts demonstrated high photocatalytic efficiency in SMX removal and the degradation of other pollutants in UPW, their performance was comparable to that of 0.5MoTe₂/BiOCl. This is noteworthy considering that the Xe lamps used in those studies typically operate at three to five times higher power than the 100 W Xe lamp used in this work, resulting in significantly greater irradiation intensity. Additionally, as highlighted in Table 2 and Table 3, many studies have fallen behind in addressing the effect of aqueous matrices. Moreover, studies that have examined photocatalytic performance in real water matrices often report significant inhibition.

Thus, the superior performance of 0.5MoTe₂/BiOCl compared with other photocatalysts lies in its high efficiency in real water matrices. This advantage is attributed to the effective separation of holes and electrons (Figure 5), as well as the dominance of holes and superoxide oxygen that are more selective than hydroxyl radicals, resulting in enhanced photocatalytic efficiency.

2.2.5. Reuse of Photocatalysts

Τhe reusability of 0.5MoTe₂/BiOCl was evaluated for the photodegradation of SMX in UPW through a series of three consecutive tests. In each experiment, 500 mg/L of 0.5 MoTe₂/BiOCl was added to 60 mL of a 500 μg/L SMX solution, and the mixture was irradiated for 90 min. SMX conversion was monitored at 15, 30, 60, and 90 min intervals. After each test, the catalyst was recovered through vacuum filtration and dried at 50 °C overnight to remove any residual moisture. It was observed that, following a slight decrease in SMX removal after the first cycle, the photocatalytic performance remained stable in the second and third cycles (with 90% SMX removal), as shown in Figure 8C. This demonstrates the high stability of the catalyst and supports its potential for use in practical applications under realistic conditions.

2.3. Proposed Transformation Pathways

High-resolution mass spectrometry (HRMS) with a TOF analyzer operated in negative ionization mode was used for the identification of the TBPs. The structural assignment of the TBPs was based on the high-resolution accurate mass data (ion molecular formula and m/z [Μ-H]⁻) depicted in

Table 4. High-resolution accurate mass data of SMX and its TBPs.

SMX/TBP Code	Ion Molecular Formula	m/z [Μ-H]−	Δ (ppm)	RDBE
SMX	C10H10N3O3S−	252.0455	−2.6	7.5
TBP 1	C10H8N3O5S−	282.0187	1.2	8.5
TBP 2	C10H8N3O4S−	266.0241	−0.6	8.5
TBP 3	C7H6N3O2S−	196.0184	1.4	6.5
TBP 4	C7H4N3O4S−	225.9931	−1.4	7.5

. Based on the four identified TBPs, the photocatalytic transformation of SMX was found to include two main routes, as depicted in Figure 9.

The first route proceeds through the oxidation of the amino group of and the formation of the nitro-derivative of SMX (TBP1). Its formation can relate to the formation of ¹O₂ and the direct oxidation of the N atom of the amino group [96]. Nitro-SMX is a common TBP that has been generated during the degradation of SMX using various AOPs [96,97,98,99,100]. The second route includes the oxidation of the methyl attached to the isoxazole ring to a carbonyl group giving rise to the formation of TBP2. This TBP was also identified during the photocatalytic degradation of SMX using tungsten-modified TiO₂ as a photocatalyst [100]. This route potentially involves the initial hydroxyl radical attack on the methyl group and the subsequent formation of the aldehyde structure. Further oxidation of the TBP2 followed by decarboxylation reactions can also lead to possible dealkylation [101]. Thereafter, and through various reactions by hydroxyl radicals and/or photogenerated holes, the opening of the isoxazole ring takes place, with the subsequent formation of two TBPs with lower molecular weights (TBP3 and TBP4). At 240 min, all the identified TBPs are completely eliminated. Subsequent application of heterogeneous photocatalysis can lead to the formation of aliphatic compounds such as organic acids (carboxylic or sulfonic acids), ketones, amines, and aldehydes before complete mineralization.

3. Materials and Methods

3.1. Chemical Reagents

For the photocatalyst synthesis: bismuth ΙΙΙ nitrate pentahydrate (Bi(NO₃)₃ ∙ H₂O CAS: 10035-06-0), potassium chloride (KCl CAS: 7447-40-7), thiourea powder (NH₂CSNH₂ CAS: 62-56-6), acetic acid solution ≥ 99% (CH₃CO₂H CAS: 64-19-7), and crystal molybdenum telluride (MoTe₂ CAS: 12058-20-7) were purchased from Sigma-Aldrich (Saint Louis, MO, USA). Sulfamethoxazole (C₁₀H₁₁N₃O₃S CAS: 723-46-6) was also obtained from Sigma-Aldrich.

For the model water systems: sodium bicarbonate (NaHCO₃ CAS: 144-55-8), humic acid (C₁₈₇H₁₈₆O₈₉N₉S, CAS: 68131-04-4), and sodium chloride (NaCl CAS: 7647-14-5) were also supplied from Sigma-Aldrich and used without further purification.

For the trapping experiments: sodium azide (NaN₃ CAS: 26628-22-8), potassium iodide (KI CAS: 7681-11-0), and tert-Butyl alcohol ((CH₃)₃COH CAS: 75-65-0) were sourced from Sigma-Aldrich.

3.2. Photocatalysts Preparation Procedure and Characterization

Pure BiOCl was synthesized following a specific procedure detailed elsewhere [102]. For the preparation of the modified photocatalysts, appropriate amounts of commercially obtained MoTe₂ were incorporated into the BiOCl precursor solution.

The specific surface area of the photocatalysts was evaluated through nitrogen adsorption/desorption isοtherms using the Brunauer–Emmett–Teller (BET) method. The measurements were performed using a Micromeritics Gemini III 2375 analyzer (Norcross, GA, USA). The XRD analysis was conducted using a Brucker D8 advance diffractοmeter (Billerica, ΜA, USA) that utilized a Cu Ka radiation source (λ = 1.5406 Å). Phase identification was carried out using the Crystallographica software (Crystallographica Search-Match 3.1) in comparison with the standard JCPDS card data. The primary crystallite size was estimated with the Scherrer equation:

$$\mathrm{d}={\displaystyle \frac{0.9\mathsf{\lambda }}{\mathrm{B}\cos \mathsf{\theta }}}$$

(8)

where λ: X-ray wavelength corresponding to Cu Ka radiation (0.15406 nm), θ: diffraction angle, and B: the line broadening (in radians) at half of its maximum. To evaluate the optical properties of the photocatalysts, diffuse reflectance spectroscopy (DRS) was performed using a Varian Cary 3E spectrophotometer (Ρalο Altο, CA, USA) equipped with an integrating sphere, with BaSO₄ serving as the reference material. A Gatan model 782 ES500 W Erlangshen CCD camera (Gatan Model 782 ES500 W, Pleasanton, CA, USA) was used for TEM analysis. To assess the increased photocatalytic activity of the as-prepared photocatalysts, photoluminescence (PL) spectroscopy was utilized using a Cary Eclipse Fluorescence Spectrometer (Santa Clara, CA, USA).

3.3. Analytical Determination

For the determination of SMX concentration, high-performance liquid chromatography was employed with a photodiode array (PDA) detector (Waters Alliance 2695, Miliford, MA, USA). The analysis was performed using a Kinetex XB-C18 column (dimension: 2.6 mm internal diameter × 50 mm length, Milford, MA, USA) maintained at a controlled temperature of 45 °C using an oven. The mobile phase was composed of a 70:30 ratio of 0.1% (v/v) H₃PO₄ and acetonitrile (ACN) flowing at a rate of 0.150 mL/min under isocratic conditions. All samples were filtered prior to injection.

3.4. Photocatalytic Tests

In a typical photocatalytic test, a preweighed amount of the photocatalyst (typically 60 mg) was added to 120 mL of 0.5 mg/L SMX solution (unless stated otherwise) and the suspension was kept in the dark for 15 min to achieve adsorption/desorption equilibrium and minimize any mass transfer limitations. Subsequently, the solution was exposed to simulated solar radiation applied by a solar simulator (Oriel, mode LCS-100) equipped with a 100 W xenon lamp. According to chemical actinometry, the intensity of the incident radiation to the reactor was 7.3 × 10⁻⁷ einstein/(L.s). At predetermined time intervals, 1.2 mL samples were withdrawn and filtered using PVDF membranes housed in polypropylene overmolds (0.22 μm, Whatman, Maidstone, UK).

3.5. Identification of Transformation By-Products (TBPs)

The TBPs generated during the process were identified using a high-resolution mass spectrometer (Bruker micrOTOF Focus II) coupled to a Dionex (Thermo Scientific, Waltham, MA, USA) Ultimate 3000 UHPLC. The TOF analyzer was operated in negative ionization mode with the following ESI-source parameters: dry gas flow rate at 8 L min⁻¹ (nitrogen), nebulizer pressure at 2.4 bar, capillary voltage at 4200 V, and dry temperature at 200 °C. To ensure mass accuracy, ±5 ppm sodium formate was used as calibrant of the TOF analyzer. SMX and its TBPs were separated on an AcclaimTM RSLC 120 C18 (2.2 μm, 2.1 × 100 mm) column (Thermo Scientific) protected by an AcclaimTM 120 C18 guard cartridge (5 μm, 3.0 × 10 mm) from Thermo Scientific, and the injection volume was 10 μL.

A gradient elution (90/10 (0 min), 10/90 (15 min), 90/10 (17 min), and 90/10 (18 min)) was performed using water with 0.01% formic acid (A) and acetonitrile (B), and the flow rate was set at 0.15 mL min⁻¹.

4. Conclusions

The development of catalytic materials with high potential to harness abundant solar light for the treatment of persistent micropollutants has long been an important goal for the research community. This work focused on examining the possible enhancement of the photocatalytic activity of BiOCl in the presence of MoTe₂, a well-known 2D material that remains largely unexplored in studies related to photocatalytic degradation of persistent pollutants. Particular attention was given to evaluating the system under a wide variety of operating conditions, with an emphasis on environmentally relevant matrices, where the true challenge of micropollutant removal lies. Additionally, the study aimed to identify the possible transformation products derived from the photodegradation of sulfamethoxazole, providing deeper insights into the degradation mechanisms and potential environmental implications.

It was revealed that the presence of the 2D material did not significantly affect either the specific surface area or the energy gap, or, therefore, the absorbance of the synthesized material in the solar spectrum. A significant enhancement in photocatalytic activity (2.8 times greater than the pure BiOCl) was observed for the 0.5MoTe₂/BiOCl sample, possibly due to enhanced separation of the photogenerated carriers. The system demonstrated satisfactory performance even at the high μg/L range of SMX concentration, and the degradation was enhanced at neutral and acidic pH. Unlike most photocatalytic studies, the presence of inorganic ions such as chlorides and bicarbonates or even humic acid did not decrease the degradation rate, while encouraging results were also obtained in the experiments performed in real matrices (bottled water and secondary effluent). This behavior was associated with the specific nature and selectivity of the reactive species that are involved in the SMX photodecomposition. This observation highlights the need for future research regarding the evaluation of activity observed in real matrices, where different interactions between reactive species, water matrix constituents, and target pollutants can occur. Finally, four transformation products were produced during SMX photodegradation, and future work must also examine their toxicity, as well as the toxicity of the treated effluent.