Assessing the Efficiency of TiO₂-Modified Rubber Tiles for Photocatalytic Degradation of Rainwater Runoff Contaminants

Featured Application

Originally developed for passive air purification in urban environments such as playgrounds and sports fields, TiO₂-modified rubber tiles manufactured from recyclate were assessed for their potential to passively contribute to rainwater runoff treatment under outdoor exposure. This dual-functionality concept—air and water protection—offers an innovative application for multifunctional materials in real-world infrastructure. Although further research is needed to fully quantify secondary leachates and assess long-term performance, this study provides promising initial evidence that such surfaces will play a role in micropollutant load discharge from rainwater runoffs, particularly in areas where permeable infrastructure is exposed to both anthropogenic pollution and precipitation.

Abstract

Triclosan (TCS), a persistent antimicrobial and endocrine-disrupting compound, is commonly found in surface and groundwater due to incomplete removal by conventional wastewater treatment. This study evaluated its fate in authentic rainwater runoff collected from a state road using rubber tiles made from recycled tires that were either uncoated (RRT) or coated with TiO₂ via the sol–gel method (SGT). Pollutants were analyzed by a high-resolution liquid chromatography–quadrupole time-of-flight mass spectrometry system (LC/MS QTOF) before and after treatment in a flat-plate cascade reactor under UV-A irradiation. After 120 min SGT achieved >50% TCS removal, while RRT achieved ~44%. Further analysis identified degradation products (chlorocatechole, quinone, and transient dioxin-like species). ECOSAR predictions indicated moderate to high toxicity for some degradation products, but their transient and low-abundance detection suggests that photocatalysis suppresses accumulation, ultimately yielding less harmful products such as benzoic acid. These findings highlight the dual role of TiO₂-coated rubber tiles: improving material durability while enabling photocatalytic degradation.

1. Introduction

Amid ongoing climate change, increasingly frequent heavy rainfall events, and earlier seasonal snowmelts, the risk of discharging untreated wastewater into natural watercourses is rising. A major ecological concern over the past two decades has been endocrine-disrupting compounds (EDCs) in freshwater, which negatively impact reproductive health in humans and wildlife [1,2,3]. Rainwater runoff often carries chemical and microbiological contaminants, including organic and inorganic pollutants from traffic such as heavy metals and PAHs originating from tire wear, brake linings, engines, and vehicle corrosion [4,5,6].

Pesticides are widely used in agriculture and urban areas and represent a significant source of water pollution. They enter the environment through stormwater runoff, soil erosion, leaching, airborne transport, and spraying. They can move by being dissolved in water or bound to soil particles, easily ending up in surface and groundwater—the main sources of drinking water. Their persistence and toxicity pose risks to human health, as they are linked to immune disorders, cancers, and ecosystem damage [7]. Considering that rainwater runoff is one of the major water sources, it is essential to continuously monitor and improve our understanding of the contaminants it contains [8]. Most EDCs are hydrophobic and tend to adsorb onto environmental particles. Although PAHs—produced by asphalt abrasion, tire wear, and incomplete combustion of vehicle emissions—are not considered typical EDCs, their persistent presence in aquatic systems may interfere with the hormonal balance of organisms over time. Certain compounds, such as triclosan, can interact with androgen receptors, necessitating further investigation into both their agonistic and antagonistic potentials [1].

Triclosan (TCS) is an antimicrobial agent commonly used in pesticides, antiseptics, disinfectants, and various personal care products such as shampoos, soaps, toothpastes, and detergents [3,9,10,11]. Despite its widespread application, TCS has proven to be a highly persistent and toxic contaminant that cannot be effectively removed by conventional wastewater treatment processes. TCS undergoes direct phototransformation, resulting in the formation of 2,8-dichlorodibenzo-1,4-dioxin, a carcinogenic substance. Additionally, methyl triclosan, produced through biological methylation, may be more lipophilic and bioaccumulative than TCS itself [11].

Alongside triclosan, attention should also be given to octylphenol ethoxylate (OPE), which is derived from octylphenol. OPE is also an endocrine disruptor with estrogenic activity and is considered a risk, particularly for freshwater and marine animals [12]. It is present in paints, coatings, adhesives, printing inks, and rubber products, and is also found in detergents and surfactants used in households, industry, and pesticides, as well as in some personal care products such as cosmetics, body lotions, soaps, facial creams, and hair products. Once released into the environment, OPE can break down into octylphenol, further contributing to environmental exposure [13,14].

Road runoff also contains benzothiazoles like DCBS, MBTS, and 2-(4-morpholinyl) benzothiazole, which are used as vulcanization accelerators in rubber. Their transformation products (HOBT, MBT, BTSA) are linked to tire wear particles and road runoff [15,16]. Fine rubber particles enriched with HOBT accumulate on roads and are mobilized by rainwater, as confirmed in previous and current studies. Benzothiazole-2-sulfonic acid is also used in rubber as a vulcanization accelerator and enters the environment in the same manner as HOBT [15].

Benzoic acid is commonly used in pesticides, the pharmaceutical industry, and rubber product manufacturing due to its antimicrobial properties and ability to act as a preservative. It is also found in personal care products, food preservation, and as a color fixative in textiles. Additionally, it is used in construction, where it serves as a corrosion inhibitor and mold repellent [17,18]. Benzoic acid is water-soluble and has low environmental persistence and toxicity but can cause reproductive toxicity as well as skin and eye irritation [17].

In addition to pesticides and chemicals originating from tire wear particles, pharmaceuticals are also frequently present in rainwater runoff. Valproic acid (VPA) treats neurological disorders and acts as a histone deacetylase inhibitor (HDACi) with potential in cancer therapy by regulating genes involved in antitumor responses. It induces autophagic cell death in prostate and glioma cancer cells, associated with oxidative stress [19].

Consequently, with the increasing number of pollutants originating from the use of pesticides, pharmaceuticals, and tire wear particles, alternative methods for treating rainwater runoff are being increasingly researched, with photocatalysis using TiO₂ standing out as a particularly effective and environmentally friendly solution [3,11,20]. Wastewater containing these pollutants, with a special emphasis on triclosan, enters sewage systems and is subsequently redirected to wastewater treatment plants. However, conventional treatment systems are often not fully effective in removing triclosan, with removal efficiencies ranging from 17.3% to 95.6%. The remaining triclosan is then released into aquatic ecosystems [10].

It is worth noting the typical composition of passenger and truck tires by weight percentage, as recycled rubber tiles are made from end-of-life vehicle tires. Passenger car tires typically consist of natural and synthetic rubbers (47%), carbon black and silica (22.5%), metals (steel beads and belts) (14%), textiles (carcass) (5.5%), vulcanizing agents (sulfur, peroxides, and metal oxides) (2.5%), and additives such as antioxidants, antiozonants, and extenders (8.5%). Truck tires, on the other hand, require more steel reinforcement (23.5%) to withstand higher mechanical stresses during use, whereas passenger car tires contain less steel and more textiles to reduce weight [21]. Regarding the specific metals found in the steel components of tires, the following elements have been quantified in the rubber granules commonly used to produce surfaces such as those examined in our study: Al, As, Ba, Be, Cd, Co, Cr, Cu, Hg, Fe, Li, Mg, Mn, Mo, Ni, Pb, Rb, Sb, Se, Sn, Sr, Tl, V, W, and Zn. Research demonstrated that only Zn was detected in concentrations exceeding regulatory limits [22,23]. Consequently, attention should be focused on reducing Zn leaching, which has been successfully achieved by incorporating TiO₂ into the rubber flooring formulation [24]. Furthermore, metallic co-catalysts play a crucial role in enhancing photocatalytic performance [25,26,27,28,29,30,31,32]. Recycled rubber tiles may thus be particularly suitable for photocatalytic applications, as they already contain certain metals that can contribute either independently or synergistically as co-catalysts with TiO₂.

The primary aim was to identify and quantify the concentration of organic pollutants in rainwater collected from asphalt surfaces by LC/MS QTOF and to assess the water’s toxicity before and after the treatment process. The findings may serve as a basis for developing methods to purify collected rainwater and reduce pollutant discharge, particularly in agricultural areas. This research closely follows and contributes to the broader guidelines of sustainable development, emphasizing integrated approaches to environmental protection, resource efficiency, and improved quality of life. In this context, it addresses several UN Sustainable Development Goals: it supports Goal 3 (Good Health and Well-being) by reducing air and water pollution that negatively affects human health; contributes to Goal 6 (Clean Water and Sanitation) through the potential mitigation of stormwater pollution; enhances urban air and water quality in line with Goal 11 (Sustainable Cities and Communities); and promotes Goal 12 (Responsible Consumption and Production) through the use of recycled rubber substrates and circular economy practices [33].

TiO₂ is especially notable for its diverse properties, such as its ability to photocatalytically decompose organic compounds [3,34,35,36], which promotes surface self-cleaning [37,38,39] and provides antibacterial effects [40,41,42]. Also, it is commonly used as a photocatalyst due to its excellent chemical stability, non-toxicity, affordability, widespread availability, and favorable optical and electronic properties [43,44]. Due to its photocatalytic and bactericidal properties [45] and its safety for humans and animals [45], it is commonly utilized as a photocatalyst in environmental applications, including air and water purification, hazardous waste treatment, and water disinfection [44]. It has also been widely used for the removal of triclosan, which is present in wastewater from treatment plants, in receiving rivers, tap water, and even in groundwater [46]. In the study [47], it was proven that 85% of pollutants, including triclosan, were removed within 120 min using TiO₂, which was immobilized on glass beads through the sol–gel method. Furthermore, [3] also employed sol–gel immobilization on porous Tezontle stones for the removal of triclosan using a solar photocatalytic reactor with TiO₂, where Tezontle served as the immobilization medium, achieving an efficiency of up to 74% when persulfate was used as a stronger electron acceptor.

In addition to their persistence, compounds such as triclosan may generate potentially toxic transformation products, including chlorophenols, chlorocatechols, quinones, and even dioxin-type derivatives [11,48,49,50]. These intermediates have been associated with endocrine disruption, carcinogenicity, and bioaccumulation, underlining the importance of evaluating not only the removal efficiency of triclosan but also the potential ecological risks linked to its degradation pathways [51].

To the best of our knowledge, this is the first study to investigate the performance of TiO₂-coated recycled rubber tiles under authentic rainwater runoff conditions. The novelty of this work lies in demonstrating the proof-of-concept potential of these materials as multifunctional systems: on one hand, TiO₂ enhances the mechanical durability and UV stability of recycled rubber substrates, and on the other, it provides photocatalytic activity towards micropollutant degradation. This dual functionality positions TiO₂-coated tiles as promising sustainable construction materials with added environmental value.

2. Experimental Section

2.1. Materials

2.1.1. Water Samples

Rainwater was collected in July 2023, during which the total monthly precipitation was 208.2 mm, while the total annual precipitation was 1264.1 mm [52]. Rainwater was collected in the village Remetinec, located near Novi Marof (Croatia); its GPS coordinates are 46.174902° N, 16.331490° E, and the distance from the main road is approximately 100 m. It is a suburban area through which a significant number of cars and other motor vehicles pass on a daily basis. At the same time, almost every household is engaged in agriculture—about 80% of the population owns small private gardens, while around 20% cultivate larger plots of land where they grow corn, potatoes, beans, and similar crops. According to the 2021 national census, Remetinec had 1372 inhabitants comprising 406 households and is part of the Municipality of Novi Marof [53,54].

2.1.2. Recycled Rubber Tiles (Reference and Modified with TiO₂)

The reference rubber tiles (RRTs) used in this work were produced from recycled waste tires obtained from Gumiimpex-GRP Ltd. (Varaždin, Croatia). The waste tires were mechanically ground into rubber granulates of sizes 0.0–0.5 mm, 0.5–2.0 mm, and 2.0–3.5 mm. The rubber tiles used in this paper measured 1000 × 1000 × 10 mm and were made using 9 kg of rubber granulate, 380 g of binder (polyurethane Stobicoll^® R352.00, STOCKMEIER Urethanes, Cernay, France), and 5 g of catalyst (DABCO K 2097, Air Products, Allentown, PA, USA). The prepared rubber tiles were subsequently modified to introduce photocatalytic properties. The recycled rubber tiles were first etched with a sodium hydroxide solution (NaOH, 1:10, w:V) to roughen the surface as much as possible and to form –OH groups, thereby enabling the binding of the TiO₂ photocatalyst to the rubber surface using the sol–gel method. Prior to soaking, the rubber tiles were washed with ethanol (96%, Gram-Mol) and air-dried for 10 s. The tiles were then immersed in the NaOH solution for 40 min, followed by rinsing with deionized water and drying for 24 h at 60 °C. Next, the sol–gel solution was prepared by mixing 2 g of TiO₂, 200 mL of deionized water, and 200 mL of ethanol (Gram-Mol, 96%), followed by 10 min of stirring and a 3 min treatment in an ultrasonic bath. Once the solution was well-stirred, 70 mL of acetic acid (99–100% p.a.) was added to adjust the pH to an acidic level of approximately 3–4. To improve adhesion, 5 mL of TEOS was added to the solution. The mixture was then stirred at 50 °C for 1 h. The rubber tiles were soaked in the prepared solution for 10 min and subsequently dried at 80 °C for 20 min. This soaking and drying procedure was repeated four consecutive times. After the final cycle, the tiles were left for one week at room temperature. Since the modification involved sol–gel immobilization of the TiO₂ photocatalyst, it resulted in the formation of sol–gel tiles (SGTs), and more detailed procedures are described in a previous published paper [55].

2.2. Reactor Setup

The experiment was conducted using a modular flat plate cascade reactor (FPCR), shown here in Figure 1, consisting of the following components: (1) cascades made from chemically resistant poly(methyl methacrylate) (PMMA, plexiglass), each with surface dimensions of 250 × 500 mm; (2) a collecting tank for water accumulation and recirculation; (3) a peristaltic pump (Rotarus Smart 30, Hirschmann); and (4) a lamp panel equipped with three full-spectrum UVB fluorescent lamps (Terra Exotica Sunray, 36 W, 120 cm, UVB 6.0, Terraristik Groß- und Einzelhandel, Alfeld, Germany).

In the experimental setup, a single cascade was used to host either four reference rubber tiles or the TiO₂-coated rubber tiles. The distance between the cascade surface and the lamp centerline was 10 cm, to ensure uniform irradiation across the cascade length matching the effective lamp emission zone. To characterize the UV radiation intensity, radiometric measurements were performed using a UVX radiometer (UVP Products, Cambridge, UK) fitted with appropriate sensors: UVX-31 for midrange UVB (280–340 nm) and UVX-36 for longwave UVA (335–385 nm).

Rainwater runoff was continuously recirculated from the collecting tank over the cascade surface using the peristaltic pump at a constant flow rate of Q = 26.5 cm³/s. This value was chosen as a representative low-flow condition to simulate a moderate runoff scenario where photocatalytic activity is possible [56]. The outlet tubing was perforated to provide uniform surface distribution, maintaining a thin water film not exceeding 0.5 ± 1 mm in thickness over the rubber tiles.

The first 15 min of the experiment were carried out in the absence of UV radiation (dark phase), followed by 120 min of UVB irradiation. Water samples were collected at 15, 30, 60, 90, and 120 min after the initiation of UV exposure. Each sample was drawn using a syringe fitted with a 0.22 µm PET membrane filter (Chromafil, Macherey-Nagel, Düren, Germany) to remove particulate matter and ensure sample clarity for subsequent analysis. The filtered samples were stored in vials and refrigerated until analysis.

2.3. Water Leachate Testing (LC/MS QTOF)

Three rainwater samples were prepared for analysis. The first sample consisted of untreated rainwater, whereas the second and third samples were rainwater tested in the reactor setup described in Section 2.2, using the reference rubber tile (RRT) and the TiO₂-coated rubber tile (SGT), respectively. All samples were filtered (0.22 μm, PET, Chromafil, Macherey-Nagel, Düren, Germany) and prepared for high-resolution liquid chromatography–quadrupole time-of-flight mass spectrometry (LC/MS QTOF, Agilent 6530 C Accurate Mass Q-TOF LC/MS with an Agilent 1260 Infinity II LC system, Santa Clara, CA, USA).

The samples were analyzed in accordance with the modified LC/MS method 5991-6627EN [57], which was created for acquiring the Agilent MassHunter Water Screening Personal Compound Database and Library (PCDL), which contains more than 1400 environmental contaminants (Agilent Water Screening PCDL B.07.00Water). Chromatographic separation was performed on an Agilent Zorbax SB-Aq column (4.6 × 150 mm, 3.5 μm, Agilent Technologies, Santa Clara, CA, USA) at 40 °C, with a total run time of 20 min and a post-run time of 5 min. Prior to each analysis, the column was conditioned and equilibrated with the intended mobile phase by running a sample method without sample injection. The flow rate was set to 0.40 mL/min with an injection volume of 5 μL. All solvents used were of LC/MS-grade purity: ultrapure water (LC/MS, VWR Chemicals BDH, Radnor, PA, USA), acetonitrile (LC/MS, Honeywell, Seelze, Germany), acetic acid (LC/MS, VWR Chemicals BDH, Radnor, PA, USA), and ammonium acetate (LC/MS, Carlo Erba Reagents, Milan, Lombardy, Italy). The mobile phase consisted of solvent A (MilliQ water with 0.1% acetic acid and 1 mM ammonium acetate (v/v/v)) and solvent B (0.1% acetic acid in acetonitrile). The gradient elution program began with an isocratic step of 100% A for 2.00 min, followed by a 12.00 min linear gradient, and a 2.00 min isocratic elution at 2% A. The gradient was then linearly returned to 100% A over 3.00 min and maintained until the end of the run.

Mass spectrometric conditions were as follows: drying gas temperature of 160 °C at a flow rate of 12 L/min, sheath gas temperature of 350 °C (12 L/min), and nebulizer pressure of 30 psi. The fragmentor voltage was set to 100 V, with nozzle and capillary voltages of 1000 V and 3500 V, respectively, in negative ionization mode.

Preliminary identification of suspect compounds was conducted using Agilent MassHunter Qualitative Analysis Software version 10.1, in combination with the Agilent Water Screening Master Accurate Mass Compound Database, which includes 1451 organic contaminants with corresponding high-resolution MS/MS spectra by applying a target/suspect screening algorithm and compound discovery algorithm.

Subsequent data processing and feature extraction were performed using Agilent MassHunter Profinder Software version B.10.00, employing a targeted feature extraction algorithm. Compounds with a score above 95 were considered for further interpretation. In negative ionization mode, [M − H]⁻ and [M + CH₃COO]⁻ ion species were prioritized based on their predominant presence in the Water Screening PCDL B.07.00 database.

Among the detected compounds, triclosan (Sigma-Aldrich, St. Louis, MO, USA) was identified as a confirmed target. It was detected in negative mode at a retention time of 16.9 min, with a precursor ion m/z = 286.9439. Additional fragment ions, m/z = 288.941 and 290.9383, were observed, corresponding to isotopic patterns predicted using the Isotope Distribution Calculator module (Agilent MassHunter Workstation Data Analysis Core 10.1). More details are provided in Table S1 and Figure S1 in the Supplementary Materials.

To enable quantification, three calibration solutions of triclosan were prepared at concentrations of 1 ppm, 0.1 ppm, and 0.001 ppm, and analyzed under the same LC/MS QTOF conditions. Calibration curve construction and quantification were performed using Agilent MassHunter Quantitative Analysis Software version 10.0, yielding a coefficient of determination R² = 0.9928, indicating high linearity.

Pseudo-first-order rate constants were estimated from linear fits of ln(C/C₀) versus irradiation time [11,49,50]. Because TCS concentrations did not decrease during the first 90 min, fits were restricted to the 90–120 min interval where a monotonic decrease was observed. The resulting values are therefore considered interval-specific indicators rather than full-period constants.

2.4. Environmental Impact Estimation

Based on the literature, possible degradation products of triclosan were listed (see Supplementary Materials Table S12), and all samples were screened by MassHunter Profinder Software version B.10.00, employing a targeted feature extraction algorithm—the Find by Formula, where [M − H]⁻ and [M + CH₃COO]⁻ ion species were prioritized.

Tentatively identified degradation products were further screened for aquatic toxicity using the ECOSAR module within the EPI Suite™ Web (beta version 1.0, US EPA, 2024) [58], since the standalone ECOSAR v2.2 is currently not functional due to technical issues. Predicted endpoints included acute LC₅₀ (mg/L) values for fish (96 h) and daphnia (48 h) and EC₅₀ (mg/L) for green algae (96 h) as well bioaccumulation factors (log BCF, log BAF). The predictions are intended as preliminary indicators and were used to qualitatively assess the potential ecological risks of identified triclosan degradation products as well as for the identified semi-quantified compounds.

3. Results and Discussion

3.1. Detection of Micropollutants in Rainwater Samples

Preliminary identification of suspect compounds in untreated rainwater runoff using a target/suspect screening algorithm resulted in 51 qualified candidates, of which only two compounds achieved a confidence score above 95: 2-octylphenol (99.20) and triclosan (97.21). Further confirmation using the compound discovery algorithm identified 14 compounds overall, again confirming the presence of 2-octylphenol and triclosan with high confidence. A summary of these findings is provided here in

Table 1. Preliminary identification of suspect compounds in untreated rainwater runoff and rainwater runoff after treatment (RRT, SGT).

		Initial Rainwater Runoff	After Treatment over RRT	After Treatment over SGT
Target/Suspect Screening	Qualified—overall	51	106	164
	Qualified—score ≥ 95	2	4	4
Compound Discovery	Identified—overall	14	26	56
	Identified—score ≥ 95	2	2	4

, while more details can be found in the Supplementary Materials (Tables S2 and S3).

Both triclosan and 2-octylphenol can enter rainwater runoff through multiple pathways, including leaching from urban surfaces such as PVC, rubber, and concrete, surface runoff from contaminated soils and paved areas, and overflow of water bodies receiving effluents from wastewater treatment plants (WWTPs). While triclosan is primarily introduced through agricultural and biosolid-related activities [59], (2-octylphenol contamination, although also reported from agricultural sources [60], is more strongly associated with urban infrastructure, particularly asphalt surfaces, due to its presence in tire wear particles—a significant non-point source pollutant [61]. Their environmental persistence, photoreactivity, and tendency to bioaccumulate classify both compounds as priority contaminants. Furthermore, their recognized roles as endocrine-disrupting chemicals (EDCs) raise significant concerns for ecosystem stability and public health.

The results highlight the need for advanced treatment strategies to mitigate the environmental impact of micropollutants in rainwater runoff. Therefore, the effect of rainwater runoff circulating over photocatalytically active pavement materials was further investigated. Specifically, rainwater runoff was tested after contact with reference rubber tiles (RRT) and rubber tiles coated with TiO₂ (SGT), both of which were fabricated from recycled rubber. As a result, along with triclosan and 2-octylphenol, 2-hydroxybenzothiazole (HOBT) and benzothiazole-2-sulfonic acid were qualified through target/suspect screening, while valproic acid and benzoic acid were identified using the compound discovery algorithm following 120 min of circulation over both untreated (RRT) and treated (SGT) tiles (Tables S2 and S3). The occurrence and potential origin of these compounds will be discussed in more detail in the subsequent sections. For the purposes of this study, however, the primary focus remains on the quantification of triclosan as a representative micropollutant of concern.

3.2. Identification and Quantification of Triclosan

Triclosan was identified and quantified in both calibration and rainwater runoff samples at a retention time (RT) of 16.9 ± 0.018 min. The identification was based on characteristic fragment ions, as determined by a preliminary database search in the initial runoff sample (Supplementary Materials—Table S1, Figure S1). Using this confirmation, all samples were subsequently screened for triclosan, and its presence was verified throughout the dataset (Supplementary Materials—Figures S14 and S15), clearly indicating a reduction in concentration after 120 min of recirculation over both tested surfaces.

Quantification was performed using a three-point calibration curve (0.01, 0.1, and 1 ppm) derivedfrom EIC and MS data (Supplementary Materials—Figures S16 and S17). The quantification of triclosan at each time-point sample is summarized in Table S11 (Supplementary Materials), with supporting chromatographic data provided in Figure S18 (Supplementary Materials). The average identification score (Tgt) was 96.99 ± 0.7, with a mass accuracy deviation (Diff, Tgt, ppm) of 2.037 ± 0.5. A minimum of six fragment ions were detected in each sample, confirming compound identity.

The initial concentration of triclosan in the runoff sample was 14.5 ppb. Data in the literature indicate that triclosan concentrations in river water range from 18 ng/L to 2.7 µg/L, while in runoff from land-applied biosolids, levels between <51 ng/L and 309 ng/L have been observed [59,62]. This comparison highlights the importance of source origin and land use in determining environmental triclosan levels. For comparison, 2-octylphenol concentrations in rainfall runoff from biosolid-amended fields have been reported to be between <4.9 ng/L and 203 ng/L, and up to 755 ng/L in surface water [63]. These values further underscore the significance of both urban infrastructure and agricultural practices in contributing to micropollutant loads in rainwater runoff.

3.3. Triclosan Degradation Efficiency over Photocatalytic (SGT) and Non-Photocatalytic (RRT) Rubber Tiles

The comparative analysis of triclosan concentrations revealed a clear difference in degradation efficiency between reference rubber tiles (RRT) and TiO₂-coated rubber tiles (SGT) (Figure 2, Supplementary Materials Table S11). Although both surfaces initially retained triclosan with minimal loss over the first 90 min, a notable decrease was observed after 120 min of circulation. Specifically, triclosan concentration on RRT was reduced from 15.3 ppb to 8.5 ppb during UV exposure, corresponding to a C/C₀ ratio of 0.5571. In contrast, the SGT demonstrated superior photocatalytic activity, with triclosan levels decreasing to 6.5 ppb and a C/C₀ ratio of 0.4690, representing an over 53% reduction compared to the initial state. This aligns with studies showing that TiO₂ enables up to 90% degradation of TCS under UV irradiation due to reactive oxygen species (ROS) generation [50].

This difference underscores the photocatalytic contribution of the TiO₂ coating, particularly in the final stage of exposure. The absence of substantial degradation within the first 90 min for both treatments suggests that longer contact time and cumulative photon exposure are critical for effective transformation of triclosan under the tested conditions. Furthermore, the slight increase in triclosan concentrations in SGT samples at early time points (30–90 min) may reflect desorption or redistribution dynamics on the photocatalytic surface prior to net degradation [64]. Adsorption during the initial dark phase (15 min) was minimal, further underscoring that surface reactivity under irradiation, rather than passive sorption, drove the observed removal. These trends emphasize the importance of contact duration and surface reactivity in optimizing treatment efficiency.

3.4. Indicative Kinetics of Triclosan Degradation

Because triclosan concentrations remained stable during the first 90 min (Figure 2), likely due to competitive processes in the authentic rainwater matrix, pseudo-first-order rate constants were estimated only for the active 90–120 min interval. The resulting apparent rate constants were k(TCS, RRT) = 0.01716 min⁻¹ (R² = 1.00; two-point estimate) and k(TCS, SGT) = 0.02565 min⁻¹ (R² = 1.00; two-point estimate). Given the ppb-level starting concentrations and the chemical complexity of rainwater, these values should be regarded as interval-specific indicators rather than full-period kinetic constants.

The lack of an initial decrease during the first 90 min is consistent with radical scavenging and matrix effects, which are absent in simplified model solutions. Consequently, kinetic parameters derived only from the 90–120 min window are not directly comparable to full-period fits reported in model systems. Studies examining ppm-level concentrations in model matrices typically report faster and more uniform pseudo-first-order behavior, achieving 70–90% removal within similar irradiation times [11,50].

3.5. Degradation Products and Mechanistic Insights

Re-analysis of LC/MS QTOF data enabled the tentative identification of several TCS degradation products. Detected compounds included 4-chlorocatechol (m/z 143.9978), dichlorobenzene diols (m/z 177.9586), TCS-quinone (m/z 303.9461), and low-abundance signals consistent with 2,8-dichlorodibenzo-dioxin (m/z 253.9904) and dibenzo-p-dioxin (m/z 183.045). These findings are in line with previous reports where ·OH radical attacks on the aromatic rings generated chlorophenols and catechols [11] and dioxin-type products were only transient or suppressed under efficient photocatalysis [50].

The detection of chlorinated catechols and quinone-type intermediates confirms the central role of ·OH-driven pathways, as reported by [11,50]. The transient presence of dioxin-type products (dibenzo-p-dioxin, DCDD) aligns with evidence in the literature that these species form under radical-deficient conditions but are subsequently degraded by ·OH radicals during active photocatalysis [50]. Thus, the present results reinforce the view that TiO₂-driven processes can mitigate accumulation of highly toxic dioxin-type by-products compared to direct photolysis.

Recent studies further emphasize the importance of surface and structural properties in shaping photocatalytic efficiency. For example, a recent paper [65] demonstrated that variations in TiO₂ surface area and pore structure significantly influence oxygen activation pathways and reactive oxygen species generation during VOC degradation, directly linking textural properties to catalytic performance. Although BET/N₂ adsorption data were not included in this study, our previous work [24] confirmed that TiO₂ incorporation enhances the microstructural durability and UV stability of recycled rubber tiles. Together, these insights highlight that both chemical pathways (radical-mediated degradation of TCS) and material properties (surface area, durability, stability) jointly determine the performance of TiO₂-coated substrates, which we identify as an important avenue for future research.

3.6. Environmental Relevance of Co-Detected Compounds: A Qualitative Insight

While the primary focus of this study was on triclosan, qualitative screening also revealed several additional organic compounds in rainwater runoff samples. Using Agilent MassHunter Profinder Software version B.10.00 and a target/suspect screening workflow, five compounds, 2-octylphenol, 2-hydroxybenzothiazole, benzothiazole-2-sulfonic acid, valproic acid, and benzoic acid, were additionally identified with high confidence.

These compounds, detected in both RRT and SGT-treated samples after 120 min of recirculation, are either known emerging pollutants, rubber-related leachates, or photo-transformation by-products. Although not quantified with external standards, their presence and trend behavior were assessed using A/A₀ ratios and EIC peak area comparisons (Supplementary Materials—Figures S1–S12; Tables S4–S9), providing valuable insight into their environmental fate during treatment, as shown here in

Table 2. Indication of semiquantitative A/A₀ analysis.

		−15	0	15	30	60	90	120
Triclosan	SGT	1	0.96	0.96	0.98	0.99	0.99	0.45
	RRT	1	1.06	1.03	1.01	1.00	0.98	0.59
2-Octylphenol	SGT	1	0.94	0.94	0.90	0.90	0.94	0.44
	RRT	1	1.03	1.01	1.01	0.97	0.96	0.50
2-Hydroxy -benzothiazole	SGT	/	1	1.11	1.17	1.31	1.33	1.13
	RRT	/	1	1.25	1.36	1.61	1.77	1.71
Benzothiazole-2-sulfonic acid	SGT	/	1	1.12	/	/	/	/
	RRT	/	1	1.30	1.51	1.89	2.19	5.68
Valproic acid	SGT	/	/	1	1.10	1.35	/	1.63
	RRT	/	/	1	0.72	1.41	1.77	1.90
Benzoic acid	SGT	/	1	1.09	1.15	1.40	1.52	/
	RRT	/	/	/	/	/	detected value	/

.

Among the detected compounds, 2-octylphenol displayed a reduction trend nearly the same to that of triclosan. A/A₀ ratios decreased steadily in both setups, reaching 0.44 in SGT and 0.50 in RRT at 120 min. The initial dark phase and first 15 min under UV revealed slightly greater adsorption in the SGT setup, followed by mild desorption at 60 min, possibly linked to surface rearrangement or photolysis-triggered release. Conversely, the RRT surface exhibited earlier desorption upon UV exposure, suggesting weaker retention and passive surface interaction.

On the other hand, 2-hydroxybenzothiazole (HOBT), benzothiazole-2-sulfonic acid, valproic acid, and benzoic acid have been previously reported as leachates from rubber or degradation products under UV exposure [66,67].

Valproic acid showed a markedly different trend than triclosan, with A/A₀ increasing over time to 1.63 (SGT) and 1.90 (RRT). Valproic acid is not typically reported in runoff but has known degradation pathways under oxidative conditions [68]. Benzoic acid may emerge from aromatic compound breakdown, or even indirectly from benzoyl-peroxide-derived substances within rubber [69]. The same trend was observed for benzothiazole-2-sulfonic acid, particularly in RRT samples, where A/A₀ rose over fivefold, reinforcing the idea of UV-enhanced mobilization of rubber additives. 2-hydroxybenzothiazole demonstrated a consistent increase across both surfaces up to 90 min, followed by a slight drop in the SGT setup at 120 min. This could suggest partial degradation under photocatalytic conditions, while the compound continued to accumulate on the untreated rubber. The following is consistent with 2-hydroxybenzothiazole and benzothiazole-2-sulfonic acid originating from tire wear particles [66,70,71].

These results suggest that. while TiO₂-modified tiles effectively degrade priority micropollutants like triclosan, they may also alter the profile of released or transformed by-products. The integration of chemical monitoring with photocatalytic efficiency tests offers a robust framework for evaluating such multifunctional urban materials. Additionally, the potential leaching of TiO₂ from the surface of the SGT, as well as from the RRT and its presence in “clean” rainwater, was examined. The analysis showed that for all three samples had a Ti concentration below 0.35 µg/L (detection limit). Based on these results, we can conclude that the Ti concentrations obtained in our study are not expected to have any negative impact on the environment [72].

3.7. Environmental Impact Results

Protection of human health and the environment requires reliable evaluation of chemical hazards, but traditional ecotoxicity testing is costly, time-consuming, and relies on extensive animal experiments. To overcome these limitations, in silico approaches such as Quantitative Structure–Activity Relationships (QSARs) have been developed, and their regulatory acceptance has promoted their use in environmental risk assessment. QSAR-based tools, including the Toxicity Estimation Software Tool (T.E.S.T.) and the Ecological Structure Activity Relationships (ECOSAR) Predictive Model developed by the US EPA, estimate ecotoxicity endpoints by combining molecular descriptors with training sets derived from experimental data [73,74]. In this study, we applied the ECOSAR module within the EPI Suite™ Web (beta version 1.0, US EPA, 2024) to triclosan, co-detected compounds, and tentatively identified transformation products in order to obtain preliminary insights into their aquatic toxicity and bioaccumulation potential.

The results (Figure 3; more details in Supplementary Materials Tables S14 and S15) indicated that triclosan and 2-octylphenol exhibit high acute aquatic toxicity and bioaccumulation potential, consistent with their classification as priority pollutants. Transformation products such as chlorinated catechols showed moderate toxicity, while quinone-type intermediates displayed higher toxicity with potential for bioaccumulation. Importantly, dioxin-type products (e.g., 2,8-dichlorodibenzo-1,4-dioxine) were predicted to be extremely toxic and bioaccumulative, even at very low concentrations, although their presence in our samples was only transient and at low abundance. In contrast, further oxidized products such as benzoic acid and carboxylic acids were predicted to have low toxicity [50].

Taken together, these results provide a first indication of the ecological risks associated with triclosan transformation products. While catechols and diols appear less problematic, quinones and dioxin-type derivatives remain of concern due to their higher predicted toxicity and persistence. However, their low abundance and transient detection suggest that efficient photocatalysis suppresses their accumulation, in agreement with previous reports. The eventual formation of less toxic products supports the environmental relevance of TiO₂-coated tiles as a proof-of-concept system, provided that future work includes experimental ecotoxicity assays to validate these predictions.

4. Limitations and Practical Implications

The present study demonstrates the potential of TiO₂-coated rubber tiles (SGT) for reducing concentrations of micropollutants such as triclosan and 2-octylphenol in rainwater runoff. However, several limitations should be considered when interpreting the results and assessing their practical applicability.

Limitations:

• Semi-Quantitative Detection of Co-Compounds: While Agilent MassHunter Profinder Software version B.10.00 allowed for trend analysis of several additional organic compounds (HOBT, benzothiazole-2-sulfonic acid, valproic acid, and benzoic acid), their absolute concentrations remain unknown due to lack of calibration standards. As such, their environmental relevance can be only qualitatively inferred.
• Leaching and By-Product Formation: Although SGT showed greater photocatalytic degradation efficiency for triclosan and 2-octylphenol, some compounds appeared or increased in concentration during the UV treatment phase. This suggests that the rubber material itself may contribute to the chemical load through leaching or photochemical transformation, especially under UV irradiation. Without additional control experiments, attribution remains speculative.
• Matrix Effects and Environmental Conditions: The rainwater matrix used in this study may not represent all environmental conditions (e.g., variation in pH, dissolved organic matter, or ionic strength), which can significantly influence adsorption and photocatalytic processes. Further testing under diverse environmental scenarios is necessary.
• Indicative Kinetics: Kinetic estimates are based on a two-point fit in the 90–120 min interval because no net decrease was observed in the first 90 min. These values are therefore indicative and specific to the active window. Working with authentic rainwater inherently limits reproducibility and comparability with model systems; still, it provides decision-relevant insights for real-world scenarios.
• Byproduct Toxicity: While the degradation of primary contaminants is desirable, the formation of photoproducts such as chlorinated phenols or benzothiazole derivatives may pose additional ecological risks. Future studies should address the toxicity of by-products using appropriate bioassays.
• Flow Rate: Flow variability is an important factor under real-world conditions. However, due to the practical and logistical constraints of the current experimental setup, this study was limited to testing a single flow rate of 26.5 cm³/s. This value was selected as a representative condition to simulate a moderate runoff scenario within the capabilities of the laboratory system. While the flow rate may constitute a limitation of this study, emphasis is placed on the need for future research to investigate a wider range of flow rates in order to provide a more comprehensive assessment of practical applicability.
• Cost Estimation: A preliminary consideration of material costs suggests that recycled rubber tiles are ~EUR 20 per m², while TiO₂ is ~EUR 2 per kg. Although these are indicative values, comprehensive life-cycle cost analyses factoring in durability and maintenance are required to evaluate economic feasibility.

Practical Implications:

• Urban Water Management: The integration of photocatalytically active surfaces, especially in stormwater infrastructure (e.g., pavements, tiles, drainage systems, etc.), could help reduce the burden of emerging contaminants at the source.
• Material Design: Recycled rubber tiles modified with TiO₂ represent a circular economy approach. However, attention must be paid to material stability and potential additive leaching under real-world stressors such as sunlight, abrasion, and variable temperatures.
• Monitoring Strategy: The use of high-resolution mass spectrometry with suspect and non-target screening proves valuable for environmental monitoring. A combination of targeted quantification and broader qualitative screening can better capture the chemical complexity of runoff.
• Future Development: Future research should aim to optimize photocatalyst immobilization, assess long-term performance and reusability, and integrate life-cycle assessments to evaluate the sustainability of such systems in real-world applications.

This study provides a promising starting point for innovative pollutant mitigation strategies but also highlights the need for thorough material evaluation and long-term field validation.

5. Conclusions

This study investigated the presence and photocatalytic degradation of triclosan in authentic rainwater runoff from road surfaces using recycled rubber tiles coated with TiO₂ (SGT) and uncoated reference rubber tiles (RRT). While both tile types initially adsorbed triclosan, only the TiO₂-coated tiles facilitated significant photocatalytic removal, achieving >50% degradation within 120 min of UV exposure. A comparable degradation pattern was observed for 2-octylphenol, further supporting the responsiveness of the coated tiles SGT toward phenolic pollutants.

In addition to triclosan and 2-octylphenol, several other compounds (2-hydroxybenzothiazole, benzothiazole-2-sulfonic acid, valproic acid, and benzoic acid) were detected, indicating both potential leaching from rubber substrates and the formation of degradation products. LC/MS QTOF re-analysis further enabled the tentative identification of triclosan degradation products, including chlorocatechols, dichlorobenzene diols, quinone-type intermediates, and transient signals of dioxin-like species. These findings confirm the central role of hydroxyl radical pathways and suggest that TiO₂-driven processes can mitigate the accumulation of highly toxic dioxin-type by-products compared to direct photolysis.

Complementary in silico toxicity predictions (ECOSAR, EPI Suite Web) indicated that, while triclosan and 2-octylphenol exhibit high aquatic toxicity and bioaccumulation potential, most chlorinated catechols showed only moderate toxicity. In contrast, quinone-type intermediates and dioxin-like products were predicted to be extremely toxic and bioaccumulative, even at very low concentrations. Their transient and low-abundance detection in this study, combined with the eventual formation of less toxic products such as benzoic acid and carboxylic acids, suggests that efficient photocatalysis suppresses the persistence of the most hazardous by-products.

Overall, the results highlight the dual role of TiO₂-modified rubber tiles as treatment media and potential pollutant sources, while reinforcing their proof-of-concept value as multifunctional passive systems for urban and peri-urban runoff. Future research should focus on experimental validation of degradation product toxicity, optimization of catalyst formulations, and life-cycle assessment to ensure long-term sustainability and environmental safety.