A hidden route of exposure: adsorption of endocrine disrupting compounds and chemicals of emerging concern on tire rubber

Abstract

Tire wear particles (TWPs) are a major component of non-exhaust traffic emissions and an important source of microplastics capable of retaining and transporting organic contaminants. This study investigated the sorption behaviour and adsorption kinetics of antibiotics (AAs), the endocrine-active compound E3, hydroxylated PAHs (OH-PAHs) and the biomarker cotinine using batch experiments (15–24 h) combined with LC-MS/MS analysis. Sorption was rapid and compound-specific. Cotinine showed the highest adsorption capacity (q_e = 90.91 µg g^–1), 5-hPZA was the most strongly retained among AAs (38.18 µg g^–1), and 4-OH-PHEN exhibited the highest uptake among OH-PAHs (24.63 µg g^–1). Most analytes followed the pseudo-second-order kinetic model (typically R² > 0.98), while several OH-PAHs displayed deviations, indicating diverse adsorption behaviour. Raman and ATR-FTIR analyses confirmed the heterogeneous composition of TWPs, including carbonaceous and inorganic fillers. Overall, the results demonstrate that TWPs act as effective sorbents for multiple classes of micropollutants and may influence their environmental mobility and persistence. Further research under environmentally realistic conditions - including quantitative assessment of TWP abundance and competitive sorption between co-occurring contaminants - is needed to better predict their role in air, soil and water systems.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37140-7.

Introduction

Tire wear particles (TWPs), a major component of non-exhaust particulate matter, are generated through the mechanical abrasion of tires, brakes, and road surfaces, and are subsequently released into roadside environments and aquatic ecosystems¹. Owing to their widespread occurrence and persistence, TWPs have been recognized as a significant source of microplastics (MPs) in the environment, raising increasing concern in the context of human and ecosystem health². Different studies proved the interaction between TWPs and other existing in the environment contaminants, due to their complex structure and potentially large specific surface area. These particles can not only effectively adsorb organic compounds but may also act as a secondary source of contamination by releasing previously adsorbed substances under changing environmental conditions-such as variations in pH or the presence of surfactants³.

There are thousands of known and potentially toxic compounds contained in tires. Tires can pose risks to environmental and human health across their entire life cycle⁴. Among the most concerning tire-derived chemicals are antioxidants such as N-(1,3-dimethylbutyl)-N′-phenyl-p-phenylenediamine (6PPD) and its transformation product 6PPD-quinone (6PPD-Q), which has recently been recognized as highly toxic to aquatic organisms⁵. Vulcanization agents like benzothiazoles, as well as plasticizers such as phthalates, are also released from tires and are known for their persistence, bioaccumulation potential, and endocrine-disrupting effects^5,6. In addition, leachates from tire materials have been shown to contain a wide range of micropollutants, including aniline, phenolic derivatives, polycyclic aromatic hydrocarbons (PAHs), and nitroaromatic compounds^4,6,7. These substances can induce acute and chronic toxicity in aquatic and soil organisms, disrupt microbial communities, and contribute to the overall burden of chemical pollution in the environment⁸. The chemical composition of TWP is particularly complex, typically consisting of 40–60% rubber polymers (ntural rubber, styrene–butadiene rubber, butadiene rubber), 20–45% fillers such as carbon black and silica, and 5–15% chemical additivesincluding oils, sulfur, zinc oxide, and antioxidants. Recent studies identified more than 200 different organic compounds in tires, of which nearly 70% were classified as easily leachable, underscoring their high mobility and environmental relevance⁹.

Beyond their intrinsic toxicity, tire materials interact strongly with other contaminants once released into the environment⁵. Their heterogeneous composition-combining rubber polymers, carbon black, oils, and chemical additives-provides a complex surface with multiple adsorption sites^2,5.

The complexity of interaction mechanisms further highlights the role of TWPs as efficient sorbents of diverse environmental contaminants. Due to their large surface area, oxygen-containing functional groups, and combined hydrophobic and electrostatic properties, TWPs can strongly adsorb not only metals but also a wide spectrum of non-metal pollutants, including antibiotics (AAs), pesticides, persistent organic pollutants (POPs), and pharmaceuticals.

TWPs undergo complex aging processes, including UV exposure, mechanical abrasion, and weathering during their natural functioning, which significantly modify their physicochemical properties. These alterations increase surface roughness, introduce oxygen-containing functional groups, and enhance negative surface charge, thereby promoting stronger interactions with contaminants^2,10. Aging increases oxygen-containing functional groups (e.g., hydroxyl, carboxyl), enhancing TWP polarity, hydrophilicity, and contaminant affinity¹⁰. Environmental conditions such as pH, salinity, natural organic matter, or competing ions further modulate these interactions. Collectively, these findings confirm that TWP act as dynamic carriers of both heavy metals and organic contaminants in aquatic and terrestrial systems. For instance, aged TWP show higher adsorption capacities toward heavy metals such as Cd(II) and Pb(II), increasing from 3.81 to 5.57 mg g⁻¹ for Cd and from 4.98 to 11.78 mg g⁻¹ for Pb after UV aging¹⁰. In addition to Cd and Pb, TWPs have also been shown to adsorb and leach other metals, particularly Zn, Cu, and Cr, with zinc concentrations in leachates reaching levels from hundreds of µg L^-1 to several mg L^-111,12. Such findings further confirm the strong sorption potential of TWPs toward inorganic contaminants.

In particular, TWPs demonstrate stronger adsorption capacity for AAs compared to conventional microplastics such as polyethylene (PE)⁵. Studies have clearly shown that TWPs are more likely to adsorb chlortetracycline (CTC) and amoxicillin (AMX)^13,14. Moreover, after aging - for instance under UV irradiation - the adsorption capacity of TWPs toward AAs can increase up to 23-fold compared to the pristine state. Aged TWPs also show higher adsorption of CTC and AMX than aged PE, highlighting the unique nature of their complex composition^2,5. Beyond AAs, TWPs are also capable of adsorbing other compounds, including sulfamethoxazole (SMX), sulfamethazine (SMT), cephalosporin C (CEP-C), tylosin (TYL), and ofloxacin (OFL)⁴.

Beyond that, tire-derived sorbents modified with chitosan have been investigated for pharmaceutical removal, showing high efficiency in adsorbing non-steroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen, diclofenac, and naproxen from aqueous solutions¹⁵. These findings highlight not only the intrinsic sorption capacity of raw TWPs, but also the potential of tire-based materials to be engineered into effective adsorbents for micropollutants.

In addition to pharmaceuticals, TWPs have been shown to sorb a wide spectrum of organic contaminants, particularly hydrophobic organic compounds (HOCs). Studies reported that volatile organic compounds (VOCs) such as toluene, xylene, and chlorinated aliphatics can readily interact with tire-derived rubber materials¹⁶. Similarly, polycyclic aromatic hydrocarbons (PAHs), including naphthalene, fluorene, and phenanthrene, are efficiently sorbed, with sorption strength correlating positively with their hydrophobicity (logKow values)¹⁶. Competition effects were also observed, where organochlorine pesticides such as dichlorodiphenyltrichloroethane (DDT) displaced PAHs from sorption sites on tire materials, indicating complex multi-contaminant interactions in environmental matrices. Beyond these, triclosan-a widely used antimicrobial-was reported to sorb strongly onto tire crumb rubber (TCR), further underlining the affinity of tire-derived particles toward micropollutants of emerging concern¹⁶. Additionally, phenolic derivatives, nitroaromatic compounds, and phthalates have been identified in leachates from tire materials, highlighting both their intrinsic chemical complexity and their role as active sorbents for external pollutants¹⁶. Taken together, these findings suggest that TWPs may serve as multipollutant vectors in the environment, capable of both concentrating and redistributing organic contaminants across air, soil, and aquatic systems.

In this study, a range of micropollutants representing different chemical groups was examined. The first group included OH-PAHs, such as 1-hydroxynaphthalene (1-OH-NAP), 2-hydroxynaphthalene (2-OH-NAP), 2-hydroxyfluorene (2-OH-FLU), 1-hydroxyphenanthrene (1-OH-PHEN), 2-hydroxyphenanthrene (2-OH-PHEN), 4-hydroxyphenanthrene (4-OH-PHEN), 9-hydroxyphenanthrene (9-OH-PHEN), 1-hydroxypyrene (1-OH-PYR), and 3-hydroxybenzo[a]pyrene (3-OH-BaP). Another representative analyte was cotinine (COT), a nicotine metabolite, reflecting simpler hydroxylated structures of environmental relevance. Furthermore, additional compounds, including 5-hydroxypyrazinoic acid (5-hPZA), trimethoprim (TMP), erythromycin (ERY), phenoxymethylpenicillin (PmPen) and estriol (E3) were also investigated. The choice of analytes was driven by their ecotoxicological importance, structural diversity, and physicochemical characteristics, which may govern adsorption onto microplastics and their subsequent transport in surface waters.

Materials and methods

Materials

The tire material used in this study was obtained from a summer tire (Goodyear Efficient Grip, 205/60R16 92 H, manufactured in July 2019). Cryogenic grinding was performed using a Retsch MM400 ball mill mixer operating at a frequency of 30 Hz. Both the steel grinding jars and the tire fragments were pre-cooled in liquid nitrogen. The initial particle size of the tire fragments was up to 8 mm, while the final particle size achieved was as fine as 50 μm.

The analytes investigated included 5-hPZA, TMP, ERY, PmPen, E3, COT, 1-OH-NAP, 2-OH-NAP, fluorene derivatives, 1-OH-PHEN, 2-, 3-OH-PHEN, 4-OH-PHEN, 9-OH-PHEN, 1-OH-PYR, and 3-OH-BaP. All standards were purchased from LGC Standards and Sigma-Aldrich.

Characterization of tire sample

The tire sample was characterized using a confocal Raman microscope (alpha300 R, WITec) equipped with a 532 nm laser source. Measurement conditions included a laser power of 10.164 mW, 10 accumulations, and an integration time of 0.5 s.

In addition, attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectroscopy was employed. Infrared spectra were recorded using an INVENIO R spectrometer (Bruker), equipped with a DLaTGS detector and a diamond monolithic ATR crystal embedded in a pressure clamp. This setup ensured precise control of sample pressure and high-quality spectra acquisition. Spectra were collected in the range of 4000–400 cm^–1, with a resolution of 4 cm^–1 and 32 scans per sample.

Sorption measurements

The effect of contact time between adsorbent and adsorbate was evaluated in the adsorption studies. Approximately 0.2 g of the tire sample was weighed and placed in a 300 mL conical flask, followed by the addition of 100 mL of a mixed standard solution of the target micropollutants (initial concentration – 0.08 µg/mL, 6.2 pH in methanol, stored in – 20 °C in stable condition). The suspension was agitated on a mechanical shaker to ensure homogeneity.

Aliquots of 100 µL were collected into chromatographic vial inserts at the following contact times: 15, 30, and 60 min, as well as 4, 8, 12, and 24 h. All collected samples were subsequently analyzed by Liquid Chromatography-Mass Spectrometry (LC-MS/MS) in duplicate to ensure analytical reproducibility. Intrument perfmormance was verified using procedural blanks and calibration standards.

The removal efficiency (R,%) was calculated based on the following Eq. (1)¹⁴:

1

where C₀—the initial concentration of the micropollutants (µg mL ^–1), C_e—the equilibrium concentration of the micropollutants (µg mL ^–1).

The sorption capacity (q_e, µg g ^–1), which is the amount of micropollutants adsorbed on tire at equilibrium, was calculated based on the Eq. (2)¹⁷:

2

V is the volume of the solution (mL), m is the mass of the adsorbent (g).

Kinetic modeling

To evaluate adsorption kinetics, experimental data were fitted to common kinetic models to determine which best described the adsorption process. For most of the investigated compounds, the adsorption behavior followed the pseudo-second-order model, expressed by Eq. (3)¹⁵:

3

where t - the contact time (min), q_t - adsorption capacity at time t (µg g^–1), q_e - equlibrium adsorption capacity (µg g^–1), k₂ - pseudo-second-order adsorption rate constant (g µg^–1 min^–1).

The fitting was performed in Excel using the linear form of the model. Each kinetic point was measured in duplicate.

LC-MS/MS AAs and E3 analytical procedure

LC-MS/MS method was conducted using UHPLC Thermo Fisher Scientific Chromatograph coupled to a TSQ Altis Plus Mass Spectrometer equipped with electrospray ionization source and a C18 column (50 × 2.1 mm, 1.7 μm). Gradient elution mode consisted of two mobile phases: H2O: CH3OH 95:5% with 0.1% formic acid (phase A) and 0.1% formic acid in methanol (phase B). The flow rate of the mobile phase was 0.2 mL·min^-1 and the injection volume was 0.01mL. The total run time for the analysis was 19 min. The gradient of mobile phase is presented in Table 1. The temperature of the column and autosampler was set to 25 and 5 °C, respectively. Separation and measurements were carried out using Multiple Reaction Monitoring (MRM) method and positive/negative ionization mode. Ion transfer tube temperature and vaporizer temperature were set to 325 and 350 °C, respectively. Nebulizer pressure was 20 psi, sheath gas flow, auxiliary gas flow and sweep gas flow was 50, 10 and 1 Arb. Capillary voltage was 3500 V for positive ionization and 2500 V for negative ionization. Determination and calculation of AAs concentration was performed using Xcalibur software. The measurements allowed to determine not only the concentration of tested substances but also: the limit of detection, the limit of quantification, the linearity of calibration curves and the R2 parameters (Table 2). Chromatographic and spectrometric characteristics in the analysis of AAs and E3 are presented in Table 3. Underivatized estriol was analyzed in positive electrospray ionization mode. Although negative ESI is generally preferred for estriol due to higher sensitivity, positive ESI provided sufficient signal intensity for reliable quantification at the concentration level applied in the adsorption experiments (80 ng mL⁻¹). The obtained limits of detection and quantification were adequate for the study objectives, which focused on adsorption kinetics and equilibrium behavior rather than trace-level environmental monitoring. Erythromycin was analyzed in negative electrospray ionization mode. While positive ESI typically offers higher sensitivity for macrolide antibiotics, method optimization under multi-analyte conditions showed that negative ESI provided more stable signal intensity and improved linearity within the applied concentration range. The selected ionization mode ensured reproducible quantification at the experimental concentration level and consistent analytical performance across all target compounds.

Table 1.

The gradient of flow rate for analysis of AAs and E3 by LC-MS/MS.

Time	% B
0.0	0
1.0	0
9.5	40
13.0	100
16.0	100
16.5	0
19.0	0

Table 2.

Validation parameters for LC-MS/MS analysis of AAs and E3.

Compound Parameter	5-hPZA	TMP	ERY	PmPen	E3
LOD, ng mL − 1	4.0	0.7	2.7	0.7	0.7
LOQ, ng mL − 1	12.0	2.0	8.0	2.0	2.0
Linearity, ng mL − 1	12.0–78.0	2.0–78.0	8.0–78.0	2.0–74.0	2.0–78.0
R2	0.9951	0.9960	0.9907	0.9985	0.9942

Table 3.

Chromatographic and spectrometric characteristics in the analysis of AAs and E3.

Compound	Ionization	Precursor ion mass (m/z)	Product ion mass (m/z)	Collision energy (V)
5-hPZA	+	140.9	95.0	14.4
		140.9	80.9	16.7
		140.9	55.0	21.0
TMP	+	291.1	261.1	24.1
		291.1	230.1	26.0
		291.1	123.0	26.0
ERY	–	732.6	700.6	22.6
		732.6	618.5	52.6
		732.6	564.6	55.8
PmPen	+	350.9	332.9	5.4
		350.9	300.8	13.4
		238.7	160.1	37.8
E3	+	289.1	190.0	12.4
		289.1	173.0	15.0
		289.1	171.9	22.9

LC-MS/MS OH-PAHs and COT analytical procedure

Following sample preparation, analyses were performed using an ACCUCOR BIPHENYL column (100 × 2.1 mm, 2.6 μm) with two mobile phases: water (phase A) and methanol (phase B). The total run time was 40 min, with an injection volume of 10 µL and a mobile phase flow rate maintained at 0.3 mL·min^-1, ensuring stable and accurate separation of sample components. The mobile phase gradient is presented in Table 4. The column and autosampler temperatures were set at 25 °C and 5 °C, respectively, while the ion transfer tube temperature was maintained at 325 °C. Argon was used as the collision gas at a pressure of 50 Arb (arbitrary units, as indicated by the instrument). A capillary voltage of 3.5 kV was applied to ensure efficient ionization of molecules for mass spectrometric analysis. Detailed parameters of the mass spectrometer are provided in Table 5. The analysis enabled the determination of both the limits of detection (LOD) and quantification (LOQ), as well as the linearity of calibration curves (R²), ensuring high analytical precision. Owing to these features, the LC-MS/MS method constitutes a key tool for evaluating differences in micropollutant adsorption on microplastics before and after the aging process, which is essential for assessing changes in material properties and adsorption efficiency under various environmental conditions. Validation results for individual compounds are presented in Table 6.

Table 4.

The gradient of flow rate for analysis of OH-PAHs and COT by LC-MS/MS.

Time	% B
0.0	0.0
2.0	5.0
7.00	30.0
34.5	85.0
37.0	85.0
40.0	0.0
0.0	0.0

Table 5.

Validation parameters for LC-MS/MS analysis of OH-PAHs and COT.

Parameter compound	LOD, ng mL− 1	LOQ, ng mL− 1	Linearity, ng mL− 1	R 2
COT	0.667	2	2–80	0.9920
1-OH-NAP	0.667	2	2–80	0.9982
2-OH-NAP	0.667	2	2–80	0.9968
2-OH-FLU	0.667	2	2–80	0.9973
3-OH-FLU	0.667	2	2–80	0.9973
9-OH-FLU	0.667	2	2–80	0.9973
1-OH-PHEN	0.667	2	2–80	0.9961
2-OH-PHEN	0.667	2	2–80	0.9974
3-OH-PHEN	0.667	2	2–80	0.9974
4-OH-PHEN	0.667	2	2–80	0.9979
9-OH-PHEN	0.667	2	2–80	0.9953
1-OH-PYR	0.667	2	2–80	0.9909
3-OH-BaP	2.67	8	2–80	0.9986

Table 6.

Chromatographic and spectrometric characteristics in the analysis of OH-PAHs and COT.

Compound	Ionization	Precursor ion mass (m/z)	Product ion mass (m/z)	Collision energy (V)
COT	+	177.0	80.0	34
1-OH-NAP	–	143.1	115.1	40
2-OH-NAP	–	143.0	115.0	40
2-,3-,9-OH-FLU	–	181.1	180.1	30
1-OH-PHEN	–	193.1	165.1	40
2-OH-PHEN, 3-OH-PHEN	–	193.1	165.1	40
4-OH-PHEN	–	193.1	165.1	40
9-OH-PHEN	–	193.1	165.1	40
1-OH-PYR	–	217.1	189.1	40
3-OH-BaP	–	267.1	239.1	48

Results

Confocal Raman spectroscopy

The obtained Raman spectrum of the tire material is presented in Fig. 1.

Fig. 1.

Confocal Raman microscopy images (100x magnification) of tire particles after grinding process.

The Raman spectrum from tire particles (T) shown peaks at 1160, 1336, 1595, and 2911 cm⁻¹, which are characteristic of carbon-rich materials and rubber compounds¹⁸. The most prominent signal at 1595 cm⁻¹ corresponds to the G-band, indicative of graphitic C=C vibrations commonly found in carbon black used as a filler in rubber^1,19. The peak at 1336 cm⁻¹ (D-band) reflects the presence of structural disorder or defects within the carbon network¹. A weaker signal at 1160 cm⁻¹ is likely associated with C–C or C–H vibrations, typical of polymer backbones such as those found in synthetic rubber. Additionally, the 2911 cm⁻¹ band arises from aliphatic C–H stretching, confirming the presence of hydrocarbon chains in the material^1,19.

ATR-FTIR results

The ATR-FTIR spectrum of the tire material is presented in Fig. 2.

Fig. 2.

IR spectrum of ground tire obtained using the ATR-FTIR technique.

The ATR-FTIR spectrum obtained for the tire sample shows the presence of characteristic bands corresponding to aliphatic C–H bonds in the 2800–3000 cm⁻¹ region, confirming the presence of long hydrocarbon chains in rubber polymers¹⁹. A weak peak around 3072 cm⁻¹ may indicate the presence of aromatic groups, while bands in the 1600–1640 cm⁻¹ region correspond to C=C stretching vibrations, typical of unsaturated bonds and aromatic rings in synthetic rubbers such as SBR or polybutadiene. Distinct bands in the 1300–1500 cm⁻¹ range are consistent with the presence of –CH₂– and –CH₃ groups, and a strong peak at 1067 cm⁻¹ suggests the presence of silica filler, as indicated by Si–O vibrations¹⁹. Additionally, signals in the 900–700 cm⁻¹ region may arise from aromatic compounds and specific polybutadiene vibrations. The very intense peak at 451 cm⁻¹ indicates the presence of inorganic fillers such as silica or zinc oxide¹⁴. The FTIR results complement the Raman analysis, which identified bands characteristic of carbon black (D and G bands) and confirmed the presence of hydrocarbon polymer backbones^1,19.

Sorption experiments

The adsorption capacities of the tire material in adsorption process of tested AAs, E3, OH-PAHs and COT were calculated based on the Eq. 2. The removal efficiencies for all of determined micropollutants are based on the Eq. 1 and presented in Figure S1 and S2 in Supplementary Materials. The results of kinetic modelling are presented in Table 1S, 2S and 3S including the time-resolved adsorption capacities (q_e), removal efficiencies (R), and the corresponding standard deviations (SD) for all analysed compounds. These full datasets are included in the Supplementary Materials to ensure transparency and allow detailed evaluation of adsorption behaviour over the entire experimental period.

To evaluate the adsorption mechanisms of all investigated compounds, the experimental kinetic profiles were fitted to the pseudo-second-order model (Table 7). The model showed an excellent fit for the majority of analytes, with R² values typically exceeding 0.98, confirming that the adsorption process was predominantly governed by chemisorption-driven interactions.

Table 7.

Equilibrium adsorption capacities q_e and adsorption parameters for pseudo-second-order kinetic models in adsorption of AAs, E3, OH-PAHs and COT (k₂ - adsorption rate constants, R² - correlation coefficients).

Adsorbated compound	Pseudo-second-order kinetic model		qe (µg g⁻¹)
	k2 (g µg− 1 min− 1)	R 2
hPZA	0.0483	1.000	38.18
TMP	0.0402	0.9915	5.26
ERY	0.0048	0.9931	18.79
pmPEN	0.0321	0.9980	9.68
COT	0.0007	0,9934	90,91
1-OH-NAP	0.1434	0.9949	2.52
2-OH-NAP	0.0096	0.9575	10.96
2-,3-,9-OH-FLU	0.0929	0.9916	4.77
1-OH-PHEN	0.0006	0.8862	7.36
2-, 3-OH-PHEN	0.0031	0.9601	4.77
4-OH-PHEN	0.0005	0.975	24.63
9-OH-PHEN	0.0168	0.9992	22.03
1-OH-PYR	0.0004	0.9931	13.55
3-OH-BaP	0.3917	0.9843	0.12

Among the antibiotics, the model performed exceptionally well for 5-hPZA, TMP, ERY and PmPen, with correlation coefficients ranging from 0.9915 to 1.000. The highest rate constants were obtained for 5-hPZA (k₂ = 0.0483 g µg⁻¹ min⁻¹) and TMP (k₂ = 0.0402 g µg⁻¹ min⁻¹), indicating their rapid adsorption on tire-derived particles. PmPen also showed relatively fast kinetics (k₂ = 0.0321 g µg⁻¹ min⁻¹), whereas ERY adsorbed more slowly (k₂ = 0.0048 g µg⁻¹ min⁻¹).

The equilibrium adsorption capacities (qₑ) reflected a similar pattern, with the highest sorption observed for 5-hPZA (38.18 µg g⁻¹), followed by ERY (18.79 µg g⁻¹), PmPen (9.68 µg g⁻¹) and TMP (5.26 µg g⁻¹).

For COT, the adsorption rate constant was low (k₂ = 0.0007 g µg⁻¹ min⁻¹), yet the high correlation (R² = 0.9934) confirmed that the pseudo-second-order model accurately described its adsorption behavior. Importantly, COT exhibited a high adsorption capacity (qₑ = 90.91 µg g⁻¹), indicating strong interactions with the tire matrix despite slow kinetics.

The OH-PAHs demonstrated diverse kinetic characteristics and sorption potentials. Rapid adsorption was observed for 1-OH-NAP (k₂ = 0.1434 g µg⁻¹ min⁻¹) and fluorenes (0.0929 g µg⁻¹ min⁻¹), both showing strong model fits (R² ≈ 0.99). Lower rate constants were found for 2-OH-NAP (0.0096) and 9-OH-PHEN (0.0168), whereas very slow kinetics characterised 1-OH-PHEN (0.0006), 4-OH-PHEN (0.0005) and 1-OH-PYR (0.0004).

Despite their slow kinetics, the adsorption capacities were substantial for several OH-PAHs. High qₑ values were observed for 4-OH-PHEN (24.63 µg g⁻¹), 9-OH-PHEN (22.03 µg g⁻¹) and 1-OH-PYR (13.55 µg g⁻¹). In contrast, 3-OH-BaP, despite its exceptionally high adsorption rate (k₂ = 0.3917 g µg⁻¹ min⁻¹), showed only minimal capacity (qₑ = 0.12 µg g⁻¹) at the tested concentration.

Lower R² values for 1-OH-PHEN (0.8862), 2-OH-NAP (0.9575) and 2-,3-OH-PHEN (0.9601) suggest that additional mechanisms beyond chemisorption may influence their adsorption behavior.

For the tire-derived microplastic material, the adsorption experiments further confirmed its considerable sorption potential toward antibiotics, COT, OH-PAHs, and the steroid hormone E3. The equilibrium adsorption capacities obtained after 24 h aligned with the kinetic trends described above, showing the highest retention for 5-hPZA, followed by ERY, PmPen, TMP, and E3. Their corresponding removal efficiencies ranged from 6.59% for TMP to 100% for 5-hPZA, demonstrating that TWP surfaces can effectively retain selected antibiotics and endocrine-active compounds. COT exhibited both a high sorption capacity (qₑ = 90.91 µg g⁻¹) and a substantial removal efficiency (57.37%), confirming strong TWP-analyte interactions despite its relatively slow kinetics.

The adsorption of OH-PAHs was highly differentiated, reflecting structural differences within this group. Among the naphthalene derivatives, 1-OH-NAP exhibited low adsorption capacity, whereas 2-OH-NAP showed markedly higher retention. Fluorene demonstrated moderate adsorption, while phenanthrene derivatives covered a broad range of qₑ values, with the highest capacities observed for 4-OH-PHEN and 9-OH-PHEN. Increased adsorption was also noted for the pyrene metabolite 1-OH-PYR. In contrast, 3-OH-BaP displayed only minimal adsorption at equilibrium despite its high kinetic constant, indicating that its equilibrium behaviour differed substantially from the other OH-PAHs investigated.

Discussion

Our results confirm that TWPs exhibit considerable sorption capacity toward diverse organic micropollutants, with the highest adsorption observed for COT (90.91 µg g⁻¹), followed by 5-hPZA (38.18 µg g⁻¹) and several OH-PAHs such as 4-OH-PHEN (24.63 µg g⁻¹) and 9-OH-PHEN (22.03 µg g⁻¹), while compounds like TMP (5.26 µg g⁻¹) and 1-OH-NAP (2.52 µg g⁻¹) showed only limited removal. This compound-specific variability reflects the influence of molecular structure, polarity, and substitution pattern on sorption behaviour, in agreement with trends reported for PAHs and other aromatic pollutants. It should be noted that the analytical sensitivity achieved in this study was optimized for comparative adsorption assessment rather than trace-level environmental detection, and the selected ionization modes were therefore considered adequate for the applied experimental design.

Particularly concerning are the high sorption efficiencies observed for 5-hPZA, ERY, COT, and OH-PAHs such as 9-OH-PHEN and 1-OH-NAP. These compounds are known for their ecotoxicological and human health relevance, ranging from promoting antibiotic resistance (ERY) and endocrine disruption (E3) to mutagenic and carcinogenic effects (3-OH-BaP). The strong retention of such bioactive molecules on TWPs underscores their role not only as environmental vectors but also as potential facilitators of chronic human exposure through inhalation or ingestion pathways^20–25. While the strong retention of AAs (e.g., ERY, PmPen, TMP) and endocrine-active molecules (e.g., E3) on TWPs highlights their potential role in wastewater environments²⁶, where such compounds are frequently detected in effluents and sewage sludge their significance for inhalation exposure is likely limited. In contrast, OH-PAHs deserve particular attention in the context of atmospheric exposure, as their occurrence in airborne particulate matter and urban aerosols has been repeatedly confirmed^27–29. The sorption of OH-PAHs such as 9-OH-PHEN and 3-OH-BaP onto TWPs therefore directly links our laboratory findings with environmental pathways of inhalation exposure. On the other hand, in wastewater and sludge matrices, TWPs, representing a specific category of MPs, may act as long-term reservoirs of bioactive pollutants including both AAs and OH-PAHs, potentially prolonging their environmental persistence and ecotoxicological effects^26,30.

Comparable trends have been reported by other researchers, who showed that TWPs adsorb various AAs, such as CTC and AMX, more strongly than conventional microplastics². The stronger retention observed in our study for AAs like ERY and PmPen is consistent with these findings, whereas the weaker sorption of TMP highlights that less hydrophobic and structurally simpler molecules may interact less effectively with TWP surfaces². Interestingly, while previous works have mainly focused on AAs, our findings extend this evidence to structurally diverse compounds such as COT and OH-PAHs. The efficient binding of COT (qₑ = 80.78 µg g⁻¹) indicates that TWPs can also retain small polar metabolites, a compound group rarely considered in the context of microplastic sorption.

In the case of OH-PAHs, the observed compound-specific variability mirrors trends reported for parent PAHs and other hydrophobic organic contaminants¹⁶. Hydrophobicity is often considered a factor influencing sorption affinity; however, the OH-PAHs examined here showed a wide range of qₑ values. For instance, 4-OH-PHEN and 9-OH-PHEN exhibited higher adsorption, whereas 1-OH-NAP and 1-OH-PHEN were only weakly retained, despite having comparable hydrophobicity. These observations indicate that sorption within this group of compounds can vary substantially, although the available data do not allow for determining the underlying causes of these differences.

Kinetic analysis showed that sorption of several compounds (including 5-hPZA, ERY, and PmPen) followed the pseudo-second-order model, in agreement with previous studies². In contrast, some OH-PAHs, such as 1-OH-PHEN, showed lower R² values, suggesting that additional or competing processes may influence their adsorption behaviour. For 3-OH-BaP, the combination of a very high rate constant (k₂ = 0.3917 g µg⁻¹ min⁻¹) and extremely low qₑ (0.12 µg g⁻¹) indicates rapid initial interaction without substantial accumulation at equilibrium, underscoring the diversity of sorption profiles within this class. Our kinetic findings are also broadly consistent with studies using powdered waste tire rubber as a biosorbent, where pseudo-second-order kinetics and high sorption capacities have also been reported^12,31. These parallels support the interpretation that TWPs can act as effective sorbents across multiple compound classes, although the precise mechanisms remain unresolved.

Taken together, these comparisons indicate that TWPs can act as versatile sorbents for a broad spectrum of organic contaminants, consistent with previous findings for tire-derived materials². The combination of high surface area, heterogeneous surface chemistry, and their persistent presence in the environment suggests that TWPs may interact with a variety of organic pollutants. Recent field studies confirm that TWPs are significant contributors to airborne PM₁₀ in high-traffic areas, sometimes representing the dominant fraction of road traffic particulates^32–34. In addition, high environmental abundances have been reported, with average concentrations of 372 ± 50 fragments kg⁻¹ in soils and 515 ± 20 fragments kg⁻¹ in road dust, highlighting that TWPs are widespread substrates capable of interacting with micropollutants³⁵.

Raman analysis confirmed the presence of carbonaceous phases through the D and G bands, while ATR-FTIR identified inorganic fillers such as silica and zinc oxide^19,35. This multicomponent structure is consistent with previous reports describing TWPs as mixtures of rubber polymers, carbon black, fillers, and various additives, and is further supported by our spectroscopic characterisation. Such heterogeneous composition, combined with the irregular morphology and small particle size of TWPs, provides a wide range of surface domains capable of interacting with organic contaminants. In addition, the frequent co-occurrence of metals (e.g., Zn, Fe, Al) on TWP surfaces reported in environmental studies³² further illustrates the compositional complexity of these particles. These characteristics help explain the broad variability in adsorption capacities observed across the different classes of micropollutants in this study, even though the available data do not allow identification of the specific mechanisms responsible for these differences. Overall, these findings emphasize that TWPs function as effective sorbents under laboratory conditions and are also widespread and persistent carriers in real environments^13,34.

Future research should prioritise quantitative determination of TWP concentrations across different environmental matrices, particularly in airborne particulate matter, where their abundance remains insufficiently characterised. Such data are essential for improving exposure assessments and for evaluating the potential relevance of inhalation as a pathway of concern. In addition, extending sorption studies to a broader set of micropollutants - including pharmaceuticals, pesticides, transformation products and emerging contaminants - would allow better identification of compound classes that are most likely to associate with TWPs. Investigations into competitive and simultaneous sorption, as well as correlation analyses between co-occurring pollutants, could further clarify how mixtures of contaminants interact with TWP surfaces. These approaches would collectively support more accurate environmental risk assessments and contribute to a more comprehensive understanding of the role of TWPs in multipollutant systems.

Conclusions

This study demonstrates that tire-wear particles (TWPs) exhibit substantial sorption capacity toward a wide range of organic micropollutants. The highest adsorption was observed for COT (qₑ = 90.91 µg g⁻¹), followed by 5-hPZA (38.18 µg g⁻¹) and several OH-PAHs, including 4-OH-PHEN (24.63 µg g⁻¹), 9-OH-PHEN (22.03 µg g⁻¹), and 1-OH-PYR (13.55 µg g⁻¹). In contrast, compounds such as TMP (5.26 µg g⁻¹), fluorenes (4.77 µg g⁻¹), and 1-OH-NAP (2.52 µg g⁻¹) showed markedly lower retention. These findings confirm a strong compound-specific variability in sorption behaviour across structurally diverse pollutant classes.

Kinetic modelling showed that most analytes followed the pseudo-second-order model, with high correlation coefficients (typically R² > 0.98), indicating chemisorption-driven behaviour for compounds such as 5-hPZA, TMP, ERY and PmPen. Several OH-PAHs deviated from this pattern; for example, 1-OH-PHEN showed a lower R² (0.8862), whereas 3-OH-BaP displayed an exceptionally high rate constant (k₂ = 0.3917 g µg⁻¹ min⁻¹) but very low equilibrium adsorption (0.12 µg g⁻¹), illustrating the diversity of sorption profiles within this group.

Overall, TWPs were shown to act as effective sorbents for multiple classes of micropollutants, and their ability to retain both highly sorbing compounds (e.g., COT, 5-hPZA, 4-OH-PHEN) and low-sorbing analytes highlights the heterogeneity of interactions occurring at their surfaces. Considering their documented abundance in soils, road dust and airborne particulate matter, TWPs represent environmentally relevant substrates capable of influencing the persistence and mobility of organic contaminants. Continued research aimed at quantifying TWP concentrations in different environmental matrices and examining competitive sorption among co-occurring pollutants will support a more comprehensive understanding of their role in multipollutant systems.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

D.U. was responsible for results of sorbtion measurement and kinetic modelinig, X.B and J.P. were responsible for results of Raman/ATR-FTIR , L.W and K.S were responsible for results of LC-MS/MS. All authors reviewed the manuscript.

Funding

This research did receive funding. Dominika Uchmanowicz received funding from AGH University of Krakow; Grant ID grant number 16.16.210.476. Xymena Badura received funding from Republic of Poland Ministry of Education and Science; Grant ID Project INiG-PIB no. 0064/TA/24. Katarzyna Styszko received funding from AGH University of Krakow; Grant ID grant number 16.16.210.476. Laura Węgrzyn received funding from AGH University of Krakow; Grant ID grant number 16.16.210.476. Justyna Pyssa received funding from AGH University of Krakow; Grant ID grant number 16.16.210.476.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding authors on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.