Photocatalytic removal and detoxification of antibiotic mixtures in real water matrices: a comparative study of BiOBr and BiOCl

Introduction The continuous release of pharmaceuticals into aquatic environments has become a growing global concern due to their persistence and biological activity even at low concentrations. Among emerging contaminants, antibiotics are particularly critical, as their presence in surface waters may contribute to the development and spread of antibiotic resistance. Sulfamethoxazole (SMX) and trimethoprim (TMP), commonly administered in combination therapy, are frequently detected in freshwater systems and have been reported in up to 62% of monitored surface water samples in Brazil, highlighting the need for effective remediation strategies [1]. Conventional wastewater treatment plants are not specifically designed to remove such micropollutants, resulting in their continuous discharge into natural water bodies. In this context, semiconductor-based photocatalysis has emerged as a promising and sustainable advanced oxidation technology. Although TiO₂ remains the most extensively investigated photocatalyst, its wide band gap restricts activation mainly to UV light, which accounts for only a small fraction of the solar spectrum [2]. Furthermore, rapid charge recombination and reduced efficiency in complex aqueous matrices limit its practical applicability under realistic environmental conditions. Bismuth oxyhalides (BiOX) have recently attracted increasing attention as visible-light-responsive photocatalysts, owing to their layered structure that promotes effective charge separation and enhanced photocatalytic performance [3]. Nevertheless, most studies are still conducted in ultrapure water and focus primarily on degradation efficiency, often overlooking the influence of real water matrices and the potential ecotoxicological effects of transformation products. In this framework, the present work evaluates the photocatalytic degradation of SMX and TMP using BiOBr and BiOCl under simulated solar irradiation, with particular emphasis on matrix-dependent performance in deionized water, simulated tap water, and real river water collected in Santa Catarina (Brazil). Beyond pollutant removal, the ecotoxicological impact of treated waters was systematically assessed using the Microtox® assay to determine residual toxicity. Material and Methods Bismuth oxyhalides (BiOBr and BiOCl) were synthesized via a co-precipitation route according to the procedure reported by Falletta et al. [3]. The obtained materials were characterized using complementary physicochemical techniques to investigate their structural, morphological, and surface properties. Photocatalytic experiments were carried out in a Pyrex glass batch reactor (V = 250 mL) containing 100 mL of antibiotic solution at the fixed concentration, with a catalyst loading of 0.25 gL-1. The suspension was irradiated using a simulated solar lamp (Osram Ultra-Vitalux 300 W) with a mean power density of 35 Wm-2. SMX and TMP photodegradations were monitored by LC–MS analysis. Mineralization was evaluated through total organic carbon (TOC) analysis, while the ecotoxicity of untreated and treated solutions was assessed using the Microtox® assay. Results and Discussion The structural characterization confirmed the successful synthesis of phase-pure BiOCl and BiOBr. SEM analysis revealed the typical layered morphology of bismuth oxyhalides, consisting of lamellar nanosheets assembled into hierarchical microstructures. BiOBr exhibited a more open and loosely packed architecture, whereas BiOCl appeared comparatively more compact. UV–Vis DRS analysis highlighted distinct optical properties, with BiOBr displaying a narrower band gap (2.6 eV) than BiOCl (3.3 eV), influencing their light absorption under simulated solar irradiation. The photocatalytic performances of BiOBr and BiOCl were comparatively evaluated for SMX and TMP degradation, focusing on pollutant concentration, matrix composition, mixture effects, catalyst stability, and regenerability. Under simulated solar irradiation, both photocatalysts achieved complete degradation of SMX and TMP (10 mgL-1) in ultrapure water, with negligible dark adsorption. BiOCl consistently exhibited faster degradation rate constant (0.08 min-1) than BiOBr (0.06 min-1), indicating higher intrinsic photocatalytic activity. The influence of the initial antibiotic concentration (10 – 50 mgL-1) revealed progressively slower degradation at higher pollutant loadings, although complete removal was still achieved within 3 h of irradiation. Mineralization was therefore evaluated at 50 mg L-¹ to assess oxidative performance under more demanding conditions. SMX showed limited TOC removal (<15%) independently of the photocatalyst used, whereas TMP reached significantly higher mineralization values, up to ca. 80% in the presence of BiOCl, indicating different susceptibilities to deep oxidation pathways. Microtox® assays further clarified the environmental relevance of these findings. Treated waters exhibited lower acute toxicity compared to the initial solutions, with BiOCl-treated samples showing a more pronounced toxicity reduction than those treated with BiOBr. This result is consistent with the higher mineralization observed for BiOCl and suggests that the transformation products formed are less toxic than the parent compounds. When extended to simulated tap water and real river water (Pilões River, Palhoça 27°35'49" S, 48°32'56" W; Brazil), degradation rates were slightly affected due to scavenging effects from inorganic ions and natural organic matter. Nevertheless, complete removal was achieved in all matrices, with BiOCl maintaining superior kinetic performance compared to BiOBr. To better simulate realistic environmental scenarios, the photocatalytic degradation of an SMX:TMP mixture (4:1 mass ratio), selected to reflect commonly detected proportions in surface waters, was investigated in river water (Figure 1). Both photocatalysts effectively degraded the mixture, although competitive interactions between the two antibiotics slightly influenced individual degradation rates. Catalyst stability was finally evaluated through reuse cycles performed on the SMX:TMP (4:1) mixture in river water (see Graphical abstract). Both BiOBr and BiOCl retained high degradation efficiencies over successive runs, with only minor activity loss. Regeneration tests performed under UV-C irradiation (4 h) demonstrated effective recovery of photocatalytic activity, with BiOCl showing slightly higher regeneration efficiency, confirming its greater structural and functional robustness. Figure 1. Photocatalytic degradation of an SMX:TMP mixture (4:1 mass ratio) at 10 mgL-1using (a) BiOBr and (b) BiOCl under simulated solar irradiation (power density = 35 Wm-2). Conclusions This study demonstrates that bismuth oxyhalides are effective photocatalysts for the removal of SMX and TMP under simulated solar irradiation, even in complex water matrices. BiOCl exhibited faster kinetics, higher mineralization, lower residual toxicity, and improved regeneration efficiency compared to BiOBr. Photocatalytic performance was influenced by pollutant concentration and matrix composition, yet complete degradation was achieved under all tested conditions. The results highlight the potential of BiOCl as a robust and reusable catalyst for antibiotic remediation.