Titanium dioxide (TiO₂, Êxodo Científica) was employed as the inorganic support for the preparation of the investigated Ag-containing materials. Silver nitrate (AgNO₃, Sigma-Aldrich) was used as the silver precursor for the incorporation of silver-containing species into the TiO₂-based materials. Cassava starch was selected as a renewable biopolymeric gelling agent, while propolis extract was employed as a natural reducing and bioactive medium, in accordance with a green synthesis approach. All reagents were analytical grade, and ultrapure Milli-Q water was used throughout the experiments.

Prior to synthesis, TiO₂ was dried at 120 ^°C for 21 h to remove adsorbed moisture and improve dispersibility in aqueous media. Ag-containing TiO₂ materials were prepared via an impregnation route using AgNO₃ solution corresponding to a nominal elemental Ag loading of 10 wt% in the final TiO₂–Ag composition (adapted from Colpini et al. 2014). In a typical procedure, 30 g of pre-dried TiO₂ was dispersed in 30 mL of ultrapure Milli-Q water under mechanical stirring until a homogeneous suspension was obtained. Subsequently, an aqueous AgNO₃ solution, prepared by dissolving 5.2477 g of AgNO₃ in 12.4 mL of ultrapure Milli-Q water, was added under continuous agitation to promote uniform distribution of Ag⁺ species over the TiO₂ surface. The resulting suspension was maintained under rotational stirring (120 rpm) for 17 h at room temperature to ensure effective metal–support interaction. Thereafter, the excess solvent was removed by vacuum evaporation using a rotary evaporator under heating at approximately 80 ^°C. The obtained solid was dried at 120 ^°C for 21 h and subsequently calcined with a heating ramp at 400 ^°C for 5 h. The sample was designated as I.

Titanium dioxide, used as the support material for the preparation of Ag-containing TiO₂ material, was initially dried in an oven at 120 ^°C for 21 h. The material was synthesized using a modified sol–gel route employing cassava starch as a gelling agent, adapted from the methodologies described by Almeida et al. (2020), Ferreira et al. (2016), and Zarzzeka et al. (2026). Briefly, 5 g of tapioca starch were dispersed in 150 mL of distilled water under continuous stirring. Subsequently, a previously prepared suspension containing TiO₂ (30 g) and silver nitrate (AgNO₃, 5.2477 g) in 50 mL of distilled water was slowly added to the starch solution. The resulting mixture was heated at 80 ^°C for 1 h to obtain a homogeneous and viscous gel. The final suspension was concentrated using a rotary evaporator maintained at approximately 80 ^°C. To obtain the xerogel, the material was dried in an oven at 100 ^°C for 24 h and subsequently calcined with a heating ramp at 400 ^°C for 5 h. The material was prepared using a nominal elemental Ag loading of 10 wt% in the final TiO₂–Ag composition.

Raw propolis was supplied by a local producer in northern Paraná State, Brazil. Prior to extraction, the material was dried at 35 ^°C for 24 h and stored under dark and dry conditions until further use. The extraction process was performed using a semicontinuous pressurized liquid extraction (PLE) system (Figure 1). The extraction reactor had a capacity of 110 mL and was coupled to a high-performance liquid chromatography (HPLC) pump (Teledyne SSI, model SSI 12 × 6, State College, PA, USA). The extraction solvent was preheated before entering the reactor through a heat exchanger, ensuring that the desired extraction temperature was reached prior to contact with the sample. During the extraction process, pressure and temperature were continuously monitored using pressure gauges (0–7500 psi, accuracy ±0.1%) and K-type thermocouples (Wika, Klingenberg am Main, Germany), respectively. PLE was carried out at a constant pressure of 200 bar, a solvent flow rate of 5 mL min^-1, and an extraction time of 30 min. Extract aliquots were collected at 5 min intervals throughout the extraction. The solvent-to-feed (S/F) ratio was 60 g of solvent per gram of propolis. A pressure of 200 bar was selected to maintain the extraction solvent in the pressurized liquid state throughout the investigated temperature range while ensuring stable and safe operating conditions. The extraction conditions used in this study were 65% (v/v) ethanol in water, an extraction temperature of 120 °C, and a solvent pH of 2.

The Ag-containing material was prepared using a nominal elemental Ag loading of 10 wt% in the final TiO₂–Ag composition. For the synthesis, 15 g of pre-dried TiO₂ were dispersed in 30 mL of propolis extract, whose pH was adjusted to 10 using 0.1 mol L^-1 NaOH, forming a homogeneous paste. Next, 2.6240 g of the metal precursor silver nitrate (AgNO₃, Sigma-Aldrich), previously dissolved in the same extract, was added. The resulting suspension was mechanically stirred for 30 minutes to ensure complete homogenization. Subsequently, the excess solvent was removed by evaporation at 50 ^°C until a solid material was obtained (Castro et al. 2022).

A modified sol–gel route was implemented using cassava starch as a biopolymeric structuring agent and propolis extract as the liquid medium for material preparation (Figure 2). Cassava starch (5 g) was dispersed in 30 mL of propolis extract previously adjusted to pH 10 using 0.1 mol L^-1 NaOH to enhance colloidal stability and precursor reactivity. Under continuous mechanical stirring, TiO₂ (15 g) and the AgNO₃ (2.6240 g) precursor were sequentially introduced into the system. Stirring was maintained for 30 min until a homogeneous viscous sol was obtained, resulting in a viscous mixture before drying. The resulting sol was dried at approximately 50 ^°C to induce controlled gelation and solvent removal, leading to the formation of a xerogel precursor (adapted from Almeida et al. 2020 and Ferreira et al. 2016).

The strain used in the antimicrobial activity tests was Escherichia coli (ATCC 25922), cryopreserved in glycerin. To activate the cryopreserved strain, 200 μL of bacterial suspension was inoculated into test tubes containing 5 mL of Brain Heart Infusion (BHI) broth. The tubes were incubated at 35 ^°C for 24 h. After incubation, the bacterial suspension was diluted in saline solution (NaCl 9 g L^-1) to adjust the turbidity to 0.5 on the McFarland scale, equivalent to a concentration of 1 to 2 × 10⁸ CFU·mL^-1. From this adjusted suspension, successive dilutions (1:10) were performed in saline solution (9 g L^-1 NaCl) until the bacterial suspension used as an inoculum in the antimicrobial activity test was obtained, with a concentration of 10⁴ CFU mL^-1. All materials used in the analyses, as well as culture media and solutions, were prepared and sterilized in advance.

Antimicrobial activity was evaluated against Escherichia coli ATCC 25922 as a representative Gram-negative bacterial strain (Figure 3). The bacterial inoculum was prepared as described in Section 2.7 and adjusted to a final concentration of 10⁴ CFU mL^-1. Material suspensions were prepared in BHI broth at a concentration of 0.5% (w/v), corresponding to 0.5 g of material per 100 mL of culture medium, and inoculated with the standardized bacterial suspension at 10⁴ CFU mL^-1. Sample preparation and inoculation were performed under ambient laboratory light. The inoculated samples were incubated at 35 ^°C for 24 h in a closed incubator under dark conditions, without an intentional or controlled irradiation source. Following incubation, aliquots were serially diluted in sterile saline solution and plated onto Plate Count Agar (PCA). Plates were incubated at 35 ^°C for an additional 24 h. Colonies capable of forming visible colonies under the specified incubation conditions were counted and expressed as CFU mL^-1. The detection limit of the plate-count procedure under the experimental conditions was 1 CFU mL^-1. Antimicrobial activity was assessed based on the reduction in bacterial viability relative to the untreated control. The assay was conducted according to established microbiological procedures for broth-based antimicrobial evaluation, and viable cell enumeration (Balouiri et al. 2016). Three experimental repetitions were performed for each material using a homogeneous sample preparation. For each experimental repetition, bacterial enumeration was performed in duplicate, resulting in two technical plate counts per experimental repetition. Thus, the three experimental repetitions constituted the experimental replicates, whereas the duplicate plate counts were considered technical measurements.

Three experimental repetitions were used for each material, and differences among groups were evaluated using one-way analysis of variance (ANOVA), followed by Tukey's post hoc test. Differences with a p-value < 0.05 were considered statistically significant. Data are presented as mean values with the corresponding uncertainty shown in Figure 4. Statistical analyses were performed using Statistica 7.0 software (StatSoft Inc., Tulsa, OK, USA).

Escherichia coli ATCC 25922 was handled in accordance with the biosafety and laboratory safety procedures established by the institution. No institutional biosafety approval was required for this work.

The antimicrobial activity of the investigated Ag-containing TiO₂ materials against Escherichia coli ATCC 25922 is presented in Figure 4. Significant differences (p < 0.05) were observed among the evaluated samples. Under the experimental conditions used in this study, pristine TiO₂ did not significantly reduce bacterial viability compared with the untreated control. In contrast, the Ag-containing formulations showed marked reductions in bacterial counts for both synthesis routes. The incorporation of propolis was associated with a further reduction in bacterial viability, with the S-P material showing the lowest bacterial recovery among the investigated formulations.

The untreated control exhibited bacterial counts close to 8 log CFU mL^-1, indicating substantial bacterial growth in the absence of the tested materials. Similar values were obtained for pristine TiO₂, with no statistically significant difference between these two groups. Although TiO₂ is one of the most extensively investigated photocatalytic materials due to its chemical stability, low toxicity, and oxidative properties, its photocatalytic antimicrobial activity is strongly associated with light-induced generation of electron–hole pairs (Foster et al., 2011; Ran et al., 2023). However, no intentional or controlled irradiation was applied during the antimicrobial assay in the present study. Therefore, under the experimental conditions employed, the antimicrobial activity of pristine TiO₂ was insufficient to significantly reduce bacterial viability. Under illuminated conditions, the recombination of photogenerated charge carriers can limit ROS formation and consequently reduce the antibacterial efficiency of TiO₂ (Foster et al., 2011; Ran et al., 2023).

The Ag-containing formulations exhibited substantially lower bacterial recovery than the untreated control and pristine TiO₂. This behavior is consistent with the well-established antimicrobial activity of silver-containing materials reported in the literature. Previous studies have associated the antibacterial effects of silver-containing systems with mechanisms including Ag⁺ release, interactions with bacterial membranes, protein modification, interference with intracellular processes, and oxidative stress (Dakal et al., 2016; Franci et al., 2015; Burdușel et al., 2018). These mechanisms provide possible explanations for the reduced bacterial recovery observed in the present study. However, the specific contribution of these processes was not directly investigated.

Interestingly, no statistically significant difference was observed between the conventional impregnation and the starch-assisted sol–gel synthesis in the absence of propolis. This observation suggests that both preparation methods successfully incorporated biologically active silver species onto the TiO₂ support, producing comparable antimicrobial performance. Nevertheless, despite exhibiting similar bacterial inhibition, the two synthesis routes differ substantially regarding sustainability and processing characteristics (Doolan et al., 2022; Radulescu et al., 2023). The impregnation method relies primarily on the deposition of silver species onto the oxide surface, whereas the modified sol–gel approach promotes molecular-level mixing of the precursors within a biodegradable polymeric network formed by cassava starch, as schematically illustrated in Figure 5. Such structural differences become particularly relevant after the incorporation of propolis.

The most remarkable finding of this study was the substantial enhancement in antimicrobial activity associated with propolis incorporation. The addition of this natural bee-derived product was associated with increased bacterial inhibition regardless of the synthesis route, suggesting that propolis may contribute not only to synthesis-related processes but also to the antimicrobial performance of the resulting materials.

The impregnation sample containing propolis (I-P) exhibited significantly lower bacterial counts than the corresponding material prepared without propolis. This improvement suggests that bioactive compounds naturally present in propolis may contribute to bacterial inactivation, in addition to their possible participation in the synthesis process. However, the highest antimicrobial activity was achieved by the modified sol–gel sample synthesized in the presence of both cassava starch and propolis (S-P). No visible colonies were detected for this material under the conditions of the plate-count assay, considering a detection limit of 1 CFU mL^-1.

The enhanced antimicrobial response observed for S-P may be related to the combined presence of the inorganic components and the bioactive compounds present in propolis. Unlike conventional chemical reducing agents, propolis contains a highly complex mixture of flavonoids, phenolic acids, aromatic compounds, terpenoids, and other polyphenolic constituents that have been reported to participate in metal-ion reduction and stabilization processes and to exhibit antimicrobial activity (Huang et al., 2014; Zulhendri et al., 2021; Mançano et al., 2025). Consequently, propolis may play a multifunctional role during material preparation, with possible contributions to both the synthesis process and the antimicrobial response.

The enhanced activity observed for S-P may also be associated with differences in silver distribution resulting from the modified sol–gel process. During gel formation, cassava starch may act as a biodegradable structuring agent capable of interacting with precursor species and influencing nucleation and growth (Almeida et al., 2020; Ferreira et al., 2016). Such effects could potentially influence the distribution of silver-containing species within the TiO₂-based material; however, this hypothesis could not be directly verified in the present study because comprehensive physicochemical characterization was not performed.

Another important aspect concerns the intrinsic antimicrobial properties of propolis itself. Numerous studies have demonstrated that flavonoids such as pinocembrin, galangin, chrysin, and caffeic acid phenethyl ester are associated with antibacterial activity and may affect bacterial membranes, metabolic enzymes, and other essential cellular processes (Zulhendri et al., 2021). In addition, phenolic compounds have been associated with oxidative stress-related effects in microbial cells (Huang et al., 2014; Zulhendri et al., 2021). Therefore, the enhanced antimicrobial response observed for S-P may reflect the contribution of multiple components, although the specific mechanisms responsible for this response were not directly investigated in the present study.

The present findings suggest that propolis may contribute not only to the synthesis of the Ag-containing materials but also to their antimicrobial performance. Rather than being considered solely as a biomass-derived synthesis component, propolis may have a multifunctional role during material preparation and may contribute to the biological functionality of the resulting material.