Organoclay containing filtration systems as nature based solutions to treat PAHs in road runoff

Combined impact of increasing impermeable surfaces in urban areas and elevated occurrence of high intensity rainfalls due to climate change results in higher urban runoff. This urban runoff transports Polycyclic Aromatic Hydrocarbons (PAHs), frequently emitted from vehicles and accumulated on impermeable urban surfaces such as roads or parking lots and causes spread of these contaminants among many others. PAHs as persistent and bioaccumulative priority organic pollutants end up in surface water, groundwater, or soil environments degrading their quality. Therefore, controlling pollutants at the source is considered the best long-term solution. On-site filtration systems can be deployed as Nature Based Solutions (NBS) to remove pollutants from urban runoff and protect receiving water bodies in addition to their other benefits like climate change mitigation or adaptation. Clay minerals in filter media might serve to remove ionic and polar pollutants but they are ineffective in treating nonionic PAHs. Indeed, their percentage in a filter is limited due to their low hydraulic conductivity. On the other hand, organoclays which are synthesized by integrating surfactants into clay structure are found successful to reduce PAH concentrations without interfering with the hydraulic conductivity of the filter. So, in this study filter columns consisting of sand and organoclay (HDTMA-bentonite) (4:1) are operated under dynamic conditions for a month. Synthetic road runoff is prepared with 5 of the most commonly observed PAHs in dissolved phase of road runoff (Naphthalene (NAP), Acenaphthene (ACE), Phenanthrene (PHE), Fluoranthene (FLA) and Pyrene (PYR)), heavy metals, and ammonium. This synthetic road runoff is pumped through the filter continuously and effluent concentrations of PAHs are determined by gas chromatography-mass spectrometry (GC-MS). The results showed that removal performance is higher than 94%, 88% and 76% at initial concentrations of 100, 50 and 25 μg/L (of each PAH compound), respectively. This study demonstrates a filter system containing 20% organoclay has high potential as an NBS to be used on site to reduce PAH load in urban road runoff and protect water bodies.

Globally, drought and water scarcity are regarded as primary challenges necessitated by climate change. However, for many regions, increasing extreme rainfall and storm events accompanying climate change are emerging as other problems (Abbasnia and Toros, 2018; Konapala et al., 2020; Bhatti et al., 2021). Indeed, precipitation patterns in New Cairo, Egypt, have been altered by climate change, with the frequency of extreme precipitation events increasing by 17% between 2000 and 2020 compared to the pre-1999 period (Hassan et al., 2022). Another study examining precipitation changes in the US between 1991 and 2015 observed that heavy precipitation events have increased in both intensity and frequency across most of the United States since 1901. Furthermore, although rainfall regimes vary regionally and seasonally, a 4% increase in the national annual average was recorded (Easterling et al., 2017). Also, the study projects that the frequency and intensity of heavy precipitation events will continue to rise throughout the 21st century. Similarly, an increase of approximately 290–530 mm in annual rainfall is expected across a significant portion of India toward the end of this century under various models and scenarios (Sarkar et al., 2023).

Similar scenarios and projections regarding climate-induced shifting precipitation patterns also exist for Türkiye (Aziz and Yucel, 2021; Yilmaz, 2015; Abbasnia and Toros, 2018; Nigussie and Altunkaynak, 2019; Kömüşcü and Aksoy, 2023; Nacar, 2023). A study conducted for the Marmara Region, covering data from 1961 to 2016, expects the general trend of precipitation events in the coming decades to shift toward shorter but more intense rainfalls. Furthermore, it highlights the increased risk of floods and landslides (Abbasnia and Toros, 2018). Similarly, an increasing precipitation trend of 7 mm/yr along the Black Sea coasts is also noteworthy (Kömüşcü and Aksoy, 2023). Another example can be drawn from the Mediterranean Region, specifically for the province of Antalya. In a study examining the impact of climate change on extreme precipitation events in Antalya, rainfall intensities were calculated for both the current (1970–1989) and future (2080–2099) periods, and it was projected that future rainfall intensities will be 23% higher than those of the current period (Yilmaz, 2015). Increasing trend in high intensity precipitations are expected affect surface runoff especially in cities. Larger portion of impervious surfaces such as concrete buildings, roofs, asphalt roads, and pavements in urban areas produce runoff faster and in greater quantities compared to permeable land cover with vegetation or soil (Hassan et al., 2022; Gray et al., 2023)and this will worsen due to more intense rainfall events exacerbated by climate change (Kiprotich et al., 2021; Yonaba et al., 2023).

In New Cairo, the proportion of impervious surfaces has risen to 91% alongside urbanization over the last 30 years, resulting in a 135% increase in surface runoff (Hassan et al., 2022). In another study covering the years 2005–2020 in Bangladesh, it is emphasized that impervious urbanized areas and vegetation land cover are the primary determinants of changes in surface runoff. The results associated changes in runoff depth positively with urban expansion and negatively with changes in vegetation. While impervious urban areas increased by 5.44%, the average runoff also showed an increase of 17.00 mm (Das and Esraz-Ul-Zannat, 2022). In Xiamen City, China, between 1980 and 2015, it was observed that the total volume of surface runoff increased significantly from 38.2% to 48.4% (Shrestha et al., 2021). These results are consistent with a monitoring study conducted in two similar drainage areas located 1 km apart in North Carolina (one in a residential area with 53% impervious cover, the other consisting of forest and agricultural land). Surface runoff in both areas was monitored simultaneously for 5.6 years, revealing that runoff volume from the residential development was 68% higher. Furthermore, the average peak discharge rates were 420% greater than those of the undeveloped area (Line and White, 2007).

Increased frequency of intense precipitation due to global climate change, combined with the expansion of impervious surfaces through urbanization, had caused urban surface runoff to become very critical. Surface runoff management should be considered as a critical water management issue since urban runoff can contain various inorganic and organic pollutants which were accumulated on the impervious areas over which runoff flows. These contaminants are then transported to receiving water bodies such as seas, rivers and lakes, as well as to soil and other environmental matrices, potentially posing risks to living organisms and adversely affecting water quality (Ekin Kali, 2025; Valtanen et al., 2014; Pamuru et al., 2022). It is observed that urban runoff contains a variety of contaminants such as heavy metals, microplastics, phthalates, polycyclic aromatic hydrocarbons (PAHs), and pathogens which vary depending on the region and the degree of urbanization. Given their widespread presence in urban environments and their potential harm to human health and environmental media, PAHs are among the most concerning pollutants in urban runoff. They are classified as persistent and bioaccumulative organic pollutants. Indeed, 16 PAHs have been included in the list of priority pollutants defined by the United States Environmental Protection Agency (USEPA). These specific PAHs were selected due to their higher toxicity, increased probability of exposure, and widespread environmental occurrence (Zhang et al., 2023; 2024; de Souza et al., 2024).

PAHs primarily originate from incomplete combustion of hydrocarbons. PAHs released into the atmosphere through exhaust emissions, are adsorbed onto particulate matter and accumulate on various surfaces (roads, streets, asphalt, etc.) through dry or wet precipitation. During rainfall events, these compounds are incorporated into surface runoff and transported into aquatic environments (Prabhukumar and Pagilla, 2010; Abdel-Shafy and Mansour, 2016), thereby increasing PAH concentrations in the runoff and leading to significant contamination (Zhang et al., 2024).

A review of the literature indicates that urban surface runoff is a notable and concerning source for the release of PAHs into various water bodies or environmental media (Hwang and Foster, 2006; Zhang et al., 2024). Urban areas with intense traffic activity represent a continuous source of PAH contamination in surface runoff; accordingly, numerous studies have correlated high traffic density with high PAH concentrations in urban runoff (Mitsova et al., 2011; Xiong et al., 2022; Vijayanand et al., 2023). Results revealed that suspended sediments (SS) from a 100% urban catchment contained Σ16PAH levels ten times higher than those in street dust. In another study, the effluent from a constructed wetland designed to treat road runoff was analyzed. Among the EPA’s 16 priority PAHs, ten compounds—NAP, ACYL, ACEN, F, PHEN, A, FLA, PYR, CHR, and BaP—were consistently detected in the dissolved phase (Neary and Boving, 2011).

The concentrations of PAHs, which are among the most harmful chemicals transported via surface runoff to water resources, vary significantly depending on runoff type, geographic region, and specific PAH compound. For instance, a study examining 50 storm events across 21 regions in the USA, Canada, New Zealand, and Australia reported individual PAH concentrations reaching up to 40,400 μg/L (Pamuru et al., 2022). In contrast, mean total PAH levels were recorded at 161.9 ng/L in South Korea (Nguyen et al., 2021), 1,341–11,495 ng/L in Paris, France (Flanagan et al., 2018), and 146.58–3,636.59 μg/L in Tianjin, China (Shi et al., 2017). When examining the concentrations of different PAHs in surface runoff, variations are observed. In the study by S.T. Pamuru et al. (2022), PAH concentrations are found to span a wide range; PHEN, PYR, FLA, and NAP concentrations reach 15,800 μg/L, 40,400 μg/L, 22,800 μg/L, and 14,700 μg/L, respectively, while the maximum value for BaA is 7 μg/L. Similarly, BaP concentration reaches up to 12 μg/L. Although these maximum values are significantly higher than typical literature ranges, the median concentrations reported in the same study (10 ng/L for NAP (n = 188), 60 ng/L for PHE (n = 184), 170 ng/L for PYR (n = 181), and 110 ng/L for FLA (n = 183)) are more similar to global averages.

Furthermore, both total and individual PAH concentrations are heavily influenced by the land use of the runoff source. Literature indicates that total PAH levels in highway runoff range between 1,341 and 11,495 ng/L, whereas runoff from partially urbanized areas shows much lower levels (161.9 ng/L). Interestingly, runoff from artificial athletic surfaces (e.g., turf, badminton, and football fields) can contain total PAH concentrations as high as 72,230 ng/L. The composition of PAH species also shifts with the surface type; while NAP levels in highway runoff range from 21 to 485 ng/L, PYR concentrations are significantly higher (160–2,298 ng/L). In comparison, PYR levels in runoff from artificial turf reached 1,670 ng/L, whereas roof runoff showed a negligible concentration of only 1 ng/L (Ngabe et al., 2000; Flanagan et al., 2018; Fardel et al., 2020; Nguyen et al., 2021; Gasperi et al., 2022; Zhang et al., 2023).

Road runoff contaminated with PAHs transport these micropollutants into water bodies causing water pollution. PAHs in surface water resources are attributed to PAH presence in urban runoff in many studies (Manoli and Samara, 1999; Göbel et al., 2007; Oros et al., 2007; Nguyen et al., 2021) emphasized that surface runoff is a significant and widespread source of PAHs in the surface waters of Ulsan, South Korea. Similarly, research conducted in San Francisco Bay reported that 51% of the total PAH load in the bay originated from surface runoff (Oros et al., 2007). Another investigation in the Le Havre Basin reported that surface runoff accounted for 35% of the total PAH loading into the basin (Motelay-Massei et al., 2006). Furthermore, in a study in Oslo, Norway, two sedimentation ponds receiving road runoff were examined. The concentrations of 16 EPA PAHs in the pond sediments were found to be high (1,900–4,200 ng/g), and it was observed that the health of fish within these ponds was adversely affected by the road-derived pollutant (Grung et al., 2016).

Moreover, PAHs are also found in WWTP discharge and sludges because of conveyance by stormwater infrastructure even though domestic use is not a significant source for them. PAH presence is more likely especially if sewer and stormwater collection system is combined. A study in San Francisco Bay reported that 28% of the total PAH load in the bay originated from WWTP effluents (Oros et al., 2007). In another investigation at the Montreal Urban Community (Canada) WWTP, the average PAH concentrations in raw and treated wastewater were measured at 1.5 and 0.4 μg/L, respectively, leading to an estimated annual discharge of approximately 0.26 tons of PAHs into the St. Lawrence River (Pham, and Proulx, 1997). Similarly (Shi et al., 2023), examined the distribution of PAHs in the sewage collection and treatment systems of four cities along the Yangtze River. The average 16 ΣPAH concentrations were found to be 1,489.45 ng/g in sewer sediments and 781.78 ng/g in sewage sludge. Through diagnostic ratio methods and Positive Matrix Factorization (PMF), the study identified that PAHs in sewage sludge primarily originated from petroleum sources, coal, and coking activities. In contrast, PAHs in sewer sediments were mainly attributed to wood combustion, vehicular emissions, and diesel engine exhaust.

As climate change leads to high intensity rainfall events and urbanization induce increase in impervious surfaces in urban areas, the surface runoff has become more in quantity and poses a growing environmental threat. Given its role as a major non-point source of contaminants, managing these pollutants at their point of origin is evaluated as the most sustainable strategy for long-term environmental protection (Fuerhacker et al., 2011; Pereira et al., 2020). Various practices are recommended to mitigate the risks of flooding and contamination caused by surface runoff, as well as to ensure the source control and removal of these pollutants. Recommended and implemented solutions include natural and constructed wetlands, bioswales (rain trenches), permeable pavements, infiltration trenches, filtration systems, rain gardens, and green roofs (Boller et al., 2007; Scholz and Kazemi Yazdi, 2009; Prabhukumar and Pagilla, 2010; Fuerhacker et al., 2011). While some of these runoff management methods focus primarily on controlling the runoff volume, other specifically designed approaches, such as advanced filtration systems, are implemented to reduce the concentrations of contaminants within the runoff water (Fuerhacker et al., 2011; Kang et al., 2019; Nguyen et al., 2021). Filtration systems are specialized retention structures designed to control runoff pollutants and reduce overall pollutant mass loadings (Kang et al., 2019; Pereira et al., 2020). These systems function by filtering surface runoff to mitigate contamination while simultaneously facilitating groundwater infiltration, which reduces the risk of urban flooding. Such filtration systems incorporate various adsorbent media, including zeolite, peat, activated carbon, compost, bentonite, silt, limestone, clay, or sand, to sequester contaminants as the runoff passes through the filter bed (Fuerhacker et al., 2011; Pritchard et al., 2023). Furthermore, a significant advantage of these systems is their ability to be easily integrated into and accessed within existing urban infrastructure (Kang et al., 2019).

High PAH removal efficiencies are observed in both pilot-scale and field studies summarized in the literature. In pilot-scale studies, column systems acting as a filter were created, and PAH removals were examined. In a study, a column containing biochar completely removed PHE while removing 76% of NAP (Reddy et al., 2014). In another study, highway runoff containing 24 PAH types was treated in a bioretention column containing biochar and fungi, with an average PAH removal of 98% (Mitchell et al., 2023). Breuss (2019) showed that a bioretention column containing biochar, compost, and sand removed 16 PAHs 99%. Similarly (Yuan et al., 2025), achieved 98.2%–99.8% removal of 5 PAHs in a column containing biochar and activated carbon. Additionally (Ekanayake et al., 2021), investigated removal of 3 PAHs using column experiments. While soil-based filter with 10% zeolite significantly removed 30%–50% of PAHs, soil-based filter with 10% zeolite and 0.3% granular activated charcoal removed 65%–99% of PAHs in 24 h of operation. Field studies also demonstrate high efficiencies in PAH removal. For instance, research conducted in France (Flanagan et al., 2018) showed that pollutants predominantly associated with particulate matter, including Pb, Zn, and PAHs, were removed with high efficiency (90%) through a filter strip and a swale treating road runoff. Similarly, in Tennessee, USA, a peat-based filtration system installed within a sinkhole to treat highway runoff reduced PAH concentrations by up to 95% (Zhou et al., 2003) Furthermore, at the University of Maryland campus (USA), the performance of a bioretention cell—an infiltration/filtration practice featuring a media mix of soil, sand, and organic matter planted with appropriate vegetation—was evaluated for the removal of 16 PAHs from urban runoff. The study reported that the mass loading of PAHs decreased by an average of 87%, dropping from 0.0180 kg/ha/year to 0.0025 kg/ha/year (Diblasi et al., 2009).

Among filter materials deployed for different NBS, clay minerals can be applied due to their adsorptive capacities. Owing to their high surface area, hydrophilic surfaces and high cation-exchange capacities (CEC), they exhibit high adsorption capacity, especially for ionic and polar pollutants. Furthermore, clays are widely utilized as adsorbents in pollutant removal processes because they are ubiquitous, cheap, and non-toxic (Lee et al., 2012; Funes et al., 2023). However, due to their hydrophilic surface properties, natural clays are relatively ineffective in removing non-polar organic contaminants, such as PAHs (Shang and Sun, 2019). To address this limitation, the use of organoclays—modified with surfactants to convert clay surfaces to hydrophobic—is suggested for enhanced treatment performance (Younker and Walsh, 2015; Shang and Sun, 2019; Shang et al., 2020). Organoclays, synthesized by modifying various clay minerals with different cationic surfactants, serve as highly effective adsorbents for PAHs, volatile organic compounds (VOCs), pentachlorophenol (PCP), and toxic heavy metals in aqueous solution (Ake et al., 2003; Shang, 2021). The high removal of PAHs by the organoclay can be linked to adsorption, hydrophobic partitioning, and specific cation–π and π–π interactions (Javadi et al., 2017; Hedayati et al., 2021; Satouh et al., 2021; Shang et al., 2020). Through surfactant modification, cationic surfactants replace the native inorganic exchangeable cations within the clay structure, expanding the interlayer basal spacing and establishing an organophilic microenvironment that enhances the retention of hydrophobic PAHs. Interactions between PAHs and surfactants also play a critical role in this retention (Ekanayake et al., 2021; Javadi et al., 2017). While PAHs supply delocalized π-electrons via their aromatic benzene rings, the surfactant contains positively charged, electron-accepting cationic ammonium groups. Thereby It enable π–π and cation–π interactions between the PAHs and the adsorbent matrix (Hedayati et al., 2021).

For instance, Javadi et al. (2017) observed that HDTMA intercalation of clay and use of 5% organoclay increased the linear partitioning coefficient (Kd) for naphthalene by approximately 300-fold (from 4.53 mL/g in raw clay to 1355.5 mL/g in organoclay), contributing to an elevated column retardation factor (Rd = 6.98) and reduced contaminant mass flux from 0.113 mg/cm 2 × day to an average of 0.004 mg/cm 2 × day. Similarly, Shang et al. (2020) demonstrated that incorporating 10% organobentonite into cementitious media increased the partitioning coefficient (Kd​) of naphthalene from 0.0026 L/g to 0.1615 L/g, while a favorable Freundlich sorption intensity parameter (n = 1.675, corresponding to 1/n = 0.597) confirmed strong solid-phase affinity. Another study developing green pervious concrete for PAH removal observed that the inclusion of small amounts of organoclay significantly enhanced the adsorption capacity of the cement paste (Shang et al., 2020). Another research investigated the adsorption capacities of natural, synthetic, and modified clays for the removal of 16 priority PAHs. The adsorption results demonstrated that modified clays achieved removal efficiencies approaching 100% (Satouh et al., 2021). (Hedayati et al., 2021) investigated the potential adsorption capacities of clinoptilolite (MC) modified with three different surfactants—cetylpyidinium chloride (CPC-MC), didodecyldimethylammonium bromide (DDAB-MC), and hexadecyltrimethylammonium bromide (HDTMA-MC)—for the removal of PAHs. In the treatment of landfill leachate, the removal efficiencies for A, FLA, F, PHE, and PYR using CPC-MC were reported as 97.8%, 98.6%, 95.7%, 97.0%, and 98.5%, respectively. Furthermore, removal rates reached up to 99.6% for DDAB-MC. Ake et al. (2003) utilized a column composed of an organoclay and sand mixture, positioned downstream of an oil-water separator, to treat contaminated groundwater containing various PAH compounds. They reported that this column system consistently removed PAHs with efficiencies exceeding 90%. Furthermore (Lemić et al., 2007), conducted experiments using an organo-zeolite column with a synthetic mixture of five PAHs to obtain breakthrough curves. Their findings demonstrated that PAHs with higher hydrophobicity were retained in the column at rates of 70%–100% and were not desorbed.

As can be understood, climate change and urbanization have a synergistic impact on surface runoff in cities. Especially, road runoff is identified as a diffuse source of PAHs which in turn cause environmental problems. Nature based solutions such as filter boxes can be helpful to capture and treat this type of diffuse pollution at the source. Organoclays are proposed as an effective adsorbent component in such filter systems to remove PAHs that are frequently observed in urban road runoff. Adsorption of PAHs on many materials had been studied in literature through isotherm studies which were conducted at batch conditions. However, adsorption coefficients obtained by isotherm studies do not reflect removal capacity of PAHs in filter columns. There is a need to integrate adsorbents into filter columns and operate them under dynamic conditions to evaluate their removal efficiency (Kasaraneni et al., 2014). Moreover, removal of contaminants in filter columns under different loads was not studied extensively although impact of flowrate on removal efficiency was reported (Reddy vd., 2014). Thus, the novelty of this study is based on the column experiments reflecting dynamic conditions similar to road runoff at different initial concentrations to consider impact of different loads on removal efficiency. Therefore, the aim of the study is to evaluate the PAH removal performance of filter systems containing organoclay as the reactive component and sand to treat PAHs from synthetic road runoff in continuously operated columns. The results of this study aim to provide output to assess the potential effectiveness of such filter systems if they will be applied in full-scale roadside applications for treating urban road runoff.

Calcium (Ca) bentonite and quartz sand used in this study were provided by ESAN Eczacıbaşı. The specific surface area (SSA) (Santamarina et al., 2002), and cation exchange capacity (CEC) (Amonette, 1994) of raw Ca-bentonite were determined. X-ray diffraction (XRD) to determine basal spacing and X-ray fluorescence (XRF) to determine elemental compositon of the raw and synthesized clays were used. Scanning electron microscopy (SEM) images were also taken to further characterize organoclay.

For organoclay synthesis, Ca-bentonite was pulverized and passed through a 0.075 mm sieve. Then, 100 g of Ca-bentonite was dispersed in 1 L of distilled water and stirred for 24 h using a magnetic stirrer to ensure complete hydration. After that, Hexadecyltrimethylammonium bromide (HDTMA-Br, ISOLAB, >97% purity, solubility: 50,000 mg/L) was utilized as the cationic surfactant for surface modification and added at a dosage equivalent to 100% of the clay’s CEC. The mixture was stirred for an additional 24 h. The resulting organoclay was separated by filtration, then it was washed with 500 mL of distilled water to remove excess ions and dried in an oven at 105° C for 24 h. Before using in column experiments, organoclays are crushed in mortar and sieved down to 0.25 mm.

A first flush sample was taken from the side of the highway in Gebze, Kocaeli, and characterized to serve as a base to prepare synthetic road runoff. Based on this preliminary sample, synthetic road runoff was prepared containing heavy metals at concentrations of 0.6 mg/L Cu, 0.18 mg/L Fe, 0.007 mg/L Zn, 0.03 mg/L Cr, 0.02 mg/L Pb, 0.004 mg/L Ni, and 0.004 mg/L Cd, and NH 4 + at concentration of 15 mg/L. Heavy metals were used as Cu(NO 3) 2, Fe(NO 3) 3, Zn(NO 3) 2, Cr(NO 3) 3, Pb(NO 3) 2, Ni(NO 3) 2, Cd(NO 3) 2 standard solutions and NH 4 + was added in the form of NH 4 Cl. Heavy metals and NH 4 Cl were dissolved in distilled water. Before adding PAHs, pH of the prepared synthetic runoff was adjusted to 6.5–6.6 using 0,1 NaOH solution. Among PAHs, NAP, ACE, PHE, PYR, and FLA were selected for this study because those were frequently present in dissolved phase of road runoff as also reported in the literature. PAHs (AccuStandard Inc., USA) were obtained in solid form and due to their low solubility, PAH stocks were prepared in acetone. This stock solution was used to prepare synthetic road runoff. Concentration of the selected 5 PAHs were 25, 50 and 100 μg/L of each compound.

Synthetic road runoff treatment experiments were conducted using glass columns purchased from Borucam with a total height of 10 cm and an internal diameter of 5.4 cm. The filter media consisted of 45 g mixture (20% organoclay and 80% natural sand). The sand-organoclay mixture was compacted in the column in 4 layers, with each layer being pressed down 20 times. Resulting bed had a height of 1.3 cm and a porosity of 0.30. To prevent the loss of filter media, a stainless-steel mesh was secured at the bottom of the column. The columns were saturated with distilled water by being left to stand for 2 days. A layer of glass beads was placed on top of the media to ensure uniform flow distribution across the column surface. Synthetic road runoff was delivered continuously to the column at a constant flow rate of 20 mL/h using a peristaltic pump (Kamuer:M1-STP). 2 replicate columns were operated simultaneously for each initial concentration. Moreover, two replicate samples were extracted for each column resulting in an overall number of 4 samples (n = 4) to calculate the removal efficiency. Teflon pipes were used in the transfer of the synthetic sample to the columns. The column effluent was collected in a beaker located at the base of the setup. The schematic representation of the experimental column assembly is illustrated in Figure 1.

The column effluent samples were liquid-liquid extracted by hexane. 50 mL sample and 25 mL hexane were added into the separating funnel, shaken for 5 min, and waited for 10 min before water and hexane are separated. This process is repeated two more times with 25 mL hexane each time. Afterwards, hexane phases were collected in an Erlenmeyer flask, and 5 g of Na 2 SO 4 were added. A vacuum-controlled rotary evaporator (Buchi, V100) was used to evaporate the hexane containing PAH at a water temperature of 35 °C, in addition to N 2 gas flow to reduce the total extract volume to approximately 1–2 mL.

Isotopically labeled analytical standards were employed as surrogate and internal standards to account for potential mass losses during the multi-step sample pre-treatment process. Specifically, a certified internal standard mixture Z-014J (AccuStandard, Inc., USA), containing Naphthalene-d8, Acenaphthene-d10, Phenanthrene-d10, Chrysene-d12, Perylene-d12, and 1,4-Dichlorobenzene-d4 at a 0.4% concentration in dichloromethane (DCM), was spiked into all samples prior to the extraction process.

For PAH measurements, an Agilent GC/MS (5975C) instrument located in the GTU Environmental Engineering Laboratories was used. The instrument was operated in Electron Impact (EI) mode, and during the analysis, the injector, ion source, and quadrupole temperatures were set to 295, 300, and 180° C, respectively. After the extracted sample was injected using pulse spitless, a temperature program was implemented for the chromatographic separation of PAHs in the sample, which was transported to the injection block and capillary column-detector systems via helium (99.99% purity) used as the carrier gas in the system. An HP5-MS (30 m, 0.25 mm, 0.25 μm) column system was used for column separation.

The temperature program (holding at 50° C for one minute, increasing to 200° C at a rate of 25° C/min, increasing to 300° C at a rate of 8° C/min, and holding at this temperature for 5.5 min) was selected as the most suitable method in terms of peak shape and integration, using the injection volume (1 µL) and gas flow rate (1.5 mL/min). The calibration of the GC/MS was performed with standards prepared at six different concentrations (50–500 μg/L) before the analysis of the samples. The standard solution at the middle range of calibration curve was injected in every 10 samples to check the calibration. The device was recalibrated for deviations greater than 10%.

SSA of raw bentonite was measured as 539 m 2 /g whereas it increased to 844 m 2 /g for organoclay. CEC of raw bentonite was 96 meq/100 g, and after surfactant intercalation it became 151 meq/100 g. Additionally, XRD pattern of the raw material Ca-bentonite and synthesized material HDTMA-bentonite (Figure 2) revealed that basal spacing (d (001)) of pure clay increased from 9.8 Å to 12.5 Å which supported integration of the surfactant into interlayer space of the clay. Also, SEM image of the HDTMA-bentonite (Figure 3) further indicated the conversion of the layered structure of the clay into a block-like aggregated structure. Surface of the layers were covered by surfactant so the edges were not as sharp as in clay sheets. When elemental analysis of the Ca-bentonite and organoclay were compared (Table 1), Bromide was detected in organoclay whereas it did not exist in Ca-bentonite which is the main proof of Bromide containin surfactant integration into structure. Also, percentage of cations such as Ca and Mg decreased in organoclay because they were replaced with HDTMA. Stable SiO 2 ve Al 2 O 3 ratios demonstrated maintenance of the crystal stucture of clay while organoclay synthesis.

Effluent samples from columns operated continuously with synthetic road runoff were taken at different times and concentration of 5 PAHs (NAP, PHE, ACE, FLA and PYR) were determined. All of the columns were operated for 1 month although sampling times differ due to practical sample processing schedules and problems in sample processing. Removal efficiency of the columns at different influent concentrations of PAHs were summarized in Figures 4–6.

In the column operated with an inlet concentration of 100 μg/L, the average removal efficiencies were calculated as 98.78% ± 1.05% for NAP, 99.77% ± 0.77% for PHE, 99.66% ± 0.53% for ACE, 99.75% ± 0.06% for FLA, and 99.79% ± 0.04% for PYR. For NAP, which has a lower molecular weight, higher solubility and higher volatility, a slight fluctuating trend was observed, starting at 96% in the first few days and reaching levels of 97% and 98% in the following period. In contrast, no significant time-dependent changes were observed in other PAH components with a more hydrophobic character compared to NAP, such as PHE, ACE, FLA, and PYR; the removal efficiencies remained at around 99% throughout the 1-month operation period.

When the inlet concentration was reduced to 50 μg/L and the mass pollutant load was halved, the average removal rates were 98.27% ± 0.98% for NAP, 99.35% ± 0.35% for PHE, 98.97% ± 0.77% for ACE, 99.47% ± 0.11% for FLA, and 99.59% ± 0.09% for PYR. For all compounds except NAP, the time-dependent removal rates consistently remained around 99%. This less stable removal efficiency of NAP might be the result of more volatile nature of the compound or lower recovery rates of NAP during sample preparation stage. Moreover, when the influent concentration was 25 μg/L and, the average removal rates were 92.56% ± 0.13% for NAP, 98.60% ± 0.18% for PHE, 97.68% ± 0.01% for ACE, 99.31% ± 0.16% for FLA, and 99.51% ± 0.06% for PYR. For all compounds except NAP, the time-dependent removal rates consistently remained around 98%–99%. The high and similar average removal rates in both pollutant loads indicate that the adsorption capacity of the organoclay-sand containing columns has not yet reached saturation.

Overall, within the studied concentration range (25–50–100 μg/L) of PAHs mostly found in dissolved phase of road runoff (NAP, PHE, ACE, FLA and PYR) high removal efficiencies ranging from 92% to over 99% was achieved. In order to compare results and evaluate the impact of initial concentration on removal efficiency, data was analyzed by one-way ANOVA. Time-dependent measurements were not analyzed statistically, instead average removal rates of independently operated columns were used. For concentrations below LOD, LOD/2 was used to calculate removal efficiencies. It appeared that PAH removal of columns operated at different initial concentrations (and loads) were not statistically different at significance level 0.05 (F: 3.13, p:0.06, F crit: 3.35). However, it must be noted that statistical significance (p:0.06) closer to the threshold value of 0.05 could indicate an impact of initial load on removal rates even if it is insignificant.

Method validation was performed to determine the analytical reliability and data quality of the experimental analyses conducted within the scope of the study. Mean recovery values ​​of 88.2% ± 2.4% for NAP-d8, 86.1% ± 2.1% for PHE-d10, and 77.1% ± 2.1% for ACE-d10 were obtained for the internal standards used in the analyses. These values ​​are within the range of 70%–117% reported in similar studies in the literature (Fu et al., 2012). In all measurements conducted between parallel columns (n = 4), the relative standard deviation (%RSD) 1%, and the standard deviation (SD) remained below 1%, except for a single value of the standard deviation that showed a negligible deviation of 2.32%. For all reported PAHs, the correlation coefficients (R 2) of calibration curves had values higher than 0.99. Level of detection (LOD) (analyte concentration in the blank plus three times of standard deviation) was determined 2.3, 2.8, 3.0, 6.2, and 4.3 μg/L for NAP, ACE, PHE, FLA and PYR, respectively. 9 μg/L was determined as the level of quantification (LOQ) where signal to noise ratio was above 10.

Although removal efficiencies calculated using the operational baseline reading of the equipment were above 92%–99%, effluent concentrations were below LOD in this case. So, reported removal rates could be considered as the potentially maximum removal percentage. However, the guaranteed minimum removal rate should be reported as higher than 94%–98%, 88%–96% and 76%–92% based on LOD values at initial concentrations of 100, 50 and 25 μg/L, respectively. Therefore, breakthrough curves in Figures 7–9 were drawn. Here, the relative threshold levels for breakthrough (C LOD /C o) were determined where effluent concentration is equal to LOD for each PAH compound. Columns remove PAH at the minimum removal rate until the breakthrough thresholds where the actual removal might be higher but could not be quantified analytically. As can be seen from Figure 7 (breakthrough curve at initial concentration of 100 μg/L), the only PAH compound exceeding its relative concentration threshold based on LOD (2.3 μg/L) was NAP. Breakthrough occurred nearly 20 days after operation. So, although maximum removal rate was found as 98.78% and the minimum removal rate could be 97.7%, there was a risk of saturation of the column for NAP. For the rest of the other PAH compounds, breakthrough curves had not reached their corresponding threshold limits based on their LOD, indicating service time of the column continued at a high removal rate. Minimum removal rates of the column could be reported as 97.2 for PHE, 97.0% for ACE, 93.8% for FLA and 95.7% for PYR. Breakthrough curves at initial concentration of 50 μg/L were demonstrated in Figure 8 and it appeared that relative concentration of all the studied PAH compounds were below thresholds calculated by LOD. So, it can be concluded that the service life of the column had not finished. Although using baseline readings of the instrument resulted in higher removal rates, removal rates based on LOD of the compounds were slightly lower as 95.4% for NAP, 94.4% for PHE, 94.0% for ACE, 87.6% for FLA, and 91.4% for PYR. Those removal rates could be considered as the minimum removal achieved by the column. At the lowest initial concentration of 25 μg/L from each PAH compound, breakthrough curves shown in Figure 9 were obtained. Evaluated together with LOD-based thresholds, non eof the PAH compounds reached breakthrough providing removal rates at minimum of 90.8% for NAP, 88.8% for PHE, 88.0% for ACE, 75.2% for FLA and 82.8% for PYR. Overall, NAP is the compound, which was removed at a lower efficiency and reached breakthrough at initial concentration of 100 μg/L after 20 days of operation. This can be explained by the more soluble and volatile nature of NAP. Due to these characteristics, it is the least favorable PAH compound to adsorb. Among the 5 PAHs which are observed in dissolved phase of road runoff, FLA and PYR had the lowest solubility and highest affinity for adsorption owing to the higher number of rings in their structure. As a result, they were adsorbed and retained in the column at the highest ratio.

In the literature, column studies using different adsorbents have generally reported high removal rates. While (Reddy et al., 2014) reported complete removal of PHE in a column containing biochar, similar higher PHE removal in this study remained at approximately 88%–97%. On the other hand, while Reddy et al. (2014) reported over 76.0% removal for Naphthalene (NAP), this study achieved an efficiency of over 92%, and this demonstrate that both the type of PAH and adsorbent plays an important role in the study of removal efficiency.

(Yuan et al., 2025) reported very high removal efficiencies at a 5.75 μg/L initial concentration ranging from 98.2% to 99.8% for compounds such as FLA, PYR, and Benz [a]anthracene in columns using biochar and activated carbon (AC) together. Application of higher inlet concentrations (100–50–25 μg/L) in this study resulted in decreasing removal efficiencies of PYR (over 84%) and FLA (over 76%) still indicating retention performance by organoclay (Fronczyk, 2017). operated three column systems using different adsorbents in addition to valent iron (ZVI) and activated carbon (AC) (silica spongolite (SS) in system 1, zeolite (Z) in system 2, and limestone (LS) in system 3, achieving 99.0%, 96.5%, and 99% removal rates for PY, 96% removal rate for FLA in all systems, and 93.0%, 86.0%, and 96.0% removal rates for PHE, respectively. The removal rates for PYR (84–96), PHE (88–97), and FLA (76–94) were close, especially at higher initial concentrations in this study showing removal rates similar to system using zeolite and indicating parallelism.

PAH removal efficiencies obtained in this study, when compared with field studies in the literature (Zhou et al., 2003; Diblasi et al., 2009; Flanagan et al., 2018), exhibit both alignment and divergence. Specifically, the PYR (over 84%) and FLA (over 76%) efficiencies obtained in this study are lower than the removal rates reported by Flanagan et al. (2018) for plant swale (BFS) and filter strips (VFS) (91%–96%). Similarly, Wanfang Zhou et al. (2003) reported a total PAH removal rate of 95% in a peat filtration system used in highway drainage. Furthermore, while our study showed NAP removal above 92%, Flanagan et al. (2018) reported low NAP removal in their field study, and even negative removal in some measurements.

The fact that the filter system exhibits a similar and stable removal graph at different pollutant loads indicates that the adsorbent bed was able to remove the studied PAHs under detection limits. Thus, the proposed filter material could be used more successfully for road runoff heavily polluted with PAHs. However, due to the higher mass transfer driving force, it is predicted that the column with a 100 μg/L initial concentration will be saturated earlier compared to the 50 μg/L and 25 μg/L columns when the operating time is further extended. It has been observed that the organoclay-sand matrix maintains its long-term stability under the current hydraulic and pollutant loads and can serve as an on-site filter system to reduce the PAH load of road runoff and protect water sources.

Urban road runoff from impermeable surfaces such as roads, paths, and parking lots frequently causes the transport of PAHs, critical micropollutants posing a threat to environmental health. Filtration systems on the sideways of roads or parking lots as NBS can be applicable in urban areas to manage this problem through removing contaminants and infiltrating runoff into subsurface.

This study is designed to evaluate PAH removal of filter materials containing organoclays as the adsorptive components of the filter. Synthetic road runoff containing NAP, PHEN, FLA, ACE and PYR as the representative PAHs at water phase of road runoff in addition to heavy metals (Cu, Ni, Cd, Zn, Pb, Fe) and NH 4 + pumped continuously for 1 month through columns consisted of %20 HDTMA-bentonite and %80 sand. Removal of selected PAHs at three different initial concentrations (25, 50 and 100 μg/L) is observed under dynamic conditions similar to field conditions to assess the response of filter material to different contaminant loads. The most striking finding from the experimental studies is that even 1.3 cm deep adsorbent bed can provide high removal rates for a month, even when operated at PAH concentrations much higher than those encountered in the field. Since real-world field conditions will involve much lower inlet PAH concentrations and deeper filter depths, it is predicted that these organoclay-based filtration systems will have a much longer service life in the field. Considering the data in the study, it is predicted that adding 20% ​​organoclay to the filter media could be an effective solution to reduce PAH load from urban road runoff on site reducing the risk of contaminant spread. This could also provide infiltration of treated runoff into subsurface restoring urban water cycle and reducing risk of floods, so serving as a sustainable NBS and contributing to circular city.

This study determined the removal efficiency of a filter with organoclay adsorbent at a relatively high mixture ratio (20%) on removal of dissolved PAHs in synthetic road runoff. Operation at dynamic conditions with syntetic road runoff similar to field are the novel components of this study. However, further studies are required to operate columns at a larger scale to understand filter system’s removal efficiency at higher time periods completely. Also, life cycle of such a system should be assessed to investigate environmental impacts and sustainability of the proposed system. Moreover, it must be considered that the service life of filter material is limited as removal mechanism is based on adsorption. Regeneration of spent adsorbents are frequently applied, but regeneration of organoclays for further adsorption includes thermal processes which is likely to degrade organoclay structure (Lee and Tiwari, 2012). So, regeneration of organoclay for reuse could not be considered as an option. Instead, reuse of column media as an aggregate for concrete production could be evaluated as a sustainable and circular method which might be a topic for further studies.