Chemical investigation of polycyclic aromatic hydrocarbon sources in an urban area with complex air quality challenges

This section presents the findings from our investigation of polycyclic aromatic hydrocarbons (PAHs) in particulate matter during winter in Kraków, Poland. The results provide an insight into the concentration levels, composition, and potential sources of PAHs in the ambient air, with a focus on variations between different particulate matter fractions (PM10 and PM2.5). This comprehensive analysis allows for a deeper understanding of the factors affecting PAH levels in the urban atmosphere, contributing to the broader field of aerosol chemistry and air quality management.

PM10 and PM2.5 concentrations during the sampling campaign

In exploring the chemistry of aerosols, the study notably focused on the particulate matter concentrations within Kraków’s atmosphere. To further understand the dynamics of airborne particulates, an examination of PM10 and PM2.5 levels was conducted throughout the sampling campaign. During the sampling period, PM10 concentrations exhibited a significant variability, ranging from 32.4 to 134.7 µg m⁻³, with an average concentration of 77.8 µg m⁻³. Concurrently, PM2.5 concentrations ranged from 18.2 to 101.7 µg m⁻³, with an average concentration of 62.0 µg m⁻³. Figure 3 presents the variations in PM10 and PM2.5 concentrations measured during the sampling campaign. To validate the reliability of the sampling procedure, the results obtained in this study are compared with data from two air quality monitoring stations in Kraków, representing both urban background and urban environments. The monitoring station data were averaged for time periods corresponding to the study’s sampling episodes, ensuring comparability.

The data demonstrate that PM concentrations measured in this study generally followed the temporal variation trends observed at both monitoring stations, confirming the representativeness of the collected samples. However, some deviations were noted, particularly for PM10, where values recorded in this study were occasionally higher or lower than those reported by the monitoring stations. The most noticeable discrepancy occurred on February 2/3, 2015, when PM10 concentrations significantly differed from both urban and background monitoring station values. A similar pattern was observed for PM2.5, where our measurements were generally consistent with the monitoring station trends, except for February 3/4, 2015, when the recorded PM2.5 concentration was substantially lower than the station measurements.

These variations can be attributed to several factors, including differences in measurement locations and local emission sources during the sampling period. The monitoring stations are positioned at fixed locations that may be more influenced by traffic emissions, whereas the study’s sampling sites may have been subject to localized variations in emissions and atmospheric dispersion processes. Additionally, variations in wind patterns, boundary layer height, and pollutant dispersion mechanisms can contribute to discrepancies between point measurements and station averages.

The highest concentration of PM10 was recorded on December 15/16, 2014. This peak was likely associated with the prevailing meteorological conditions that favored the accumulation of particulate matter. As shown in Table 1, during this period, wind velocity was at its lowest level, which, when combined with low temperatures and low relative humidity, created conditions conducive to PM accumulation in the atmosphere. The lack of strong wind flow likely limited the dispersion of pollutants, leading to an increase in particulate concentrations. For PM2.5, the highest concentration was observed in early February 2015, coinciding with a period of decreasing relative humidity. This suggests that atmospheric dry conditions may have influenced PM2.5 levels, potentially enhancing the suspension of fine particles in the air. Additionally, this period aligns with increased domestic heating activities, a dominant source of PM2.5 emissions during winter in Kraków. The lower humidity may have also reduced particle coagulation and wet deposition processes, leading to sustained high concentrations in the air²⁸.

Atmospheric concentrations of polycyclic aromatic hydrocarbons in PM10 and PM2.5

The research resulted in the evaluation of 18 different PAHs substances for 11 sampling short episodes from November 14th, 2014 to March 07th, 2015. Figure 4 summarizes the results of the concentration of polycyclic aromatic hydrocarbons associated with ambient PM10 (Fig. 4A and C) and PM2.5 (Fig. 4B and D). The concentrations of benzo(b)fluoranthene and benzo(k)fluoranthene were presented jointly due to their inseparability during the chromatographic measurement.

The sampling period revealed PAHs patterns with different ring structures. Specifically, 4-ringed PAHs (Fla, Pyr, BaA, and Chr), 5-ringed PAHs (BeP, BbF + BkF, BaP, and Per), and 6-ringed PAHs (BghiP and IcdP) were the primary constituents throughout the study. This trend was consistent for both PM10 and PM2.5 fractions. In fact, these three groups of PAHs collectively accounted for a substantial portion, constituting 85.5% of the total PAHs in PM10 and 83.6% in PM2.5. The authors are aware of the challenges associated with the quantitative analysis of 2- and 3-ringed PAHs. These compounds are characterized by higher volatility and susceptibility to degradation during sampling and analytical procedures, potentially leading to underestimation of their concentrations. For this reason, the concentrations of 2- and 3-ringed PAHs may be slightly higher than it was reported in the paper. However, we may assume that this underestimation do not influence the conclusions of the research presented here.

During the initial days of sampling in November, concentrations of these 4-, 5-, and 6-ringed PAHs remained relatively low, ranging from 0.7 ng m⁻³ (BeP) to 8.0 ng m⁻³ (BbF + BkF) in association with PM10 and from 0.1 ng m⁻³ (BeP) to 6.4 ng m⁻³ (BbF + BkF) in association with PM2.5 (Table S2). However, as the concentrations of both PM10 and PM2.5 increased on November 24/25, 2014, the total PAH concentrations followed suit.

Notable episodes of significantly elevated concentrations of PM10-bound PAHs occurred on December 08/09, 2014, and December 15/16, 2014, with the sum of PAHs reaching 222.2 ng m⁻³ and 207.8 ng m⁻³, respectively. A similar pattern emerged in February 2015, with elevated concentrations of PAHs, particularly on February 12/13, 2015. During this period, concentrations ranged from 2.4 ng m⁻³ (IcdP) to 74.8 ng m⁻³ (BaA), resulting in a summarized PAHs value of 275.6 ng m⁻³ (Table S2). For PM2.5-bound PAHs, episodes of elevated concentrations were observed on December 09/10, 2014, February 03/04, 2015, and February 20/21, 2015, with average values of 145.7 ng m⁻³, 187.1 ng m⁻³, and 132.3 ng m⁻³, respectively (Table S3). The results obtained within this study are comparable with other works. For instance, Rogula-Kozłowska et al.²⁹ examined the concentrations of PAHs in three different locations in Katowice city (Silesian Voivodeship, Poland). The sum of PAHs related with PM2.5 was equal to 107.26 and 196.80 ng m⁻³ at urban background and urban site of Katowice city.

Among the various PAHs examined, benzo(a)pyrene (BaP) stood out as one of the most concentrated PAHs. The consistent ratio of BaP to total PAHs, both for PM10 (0.11 ± 0.03) and PM2.5 (0.11 ± 0.04), underscores BaP’s effectiveness as a reliable reference substance for assessing overall PAH concentrations³⁰. It exhibited an average concentration of 18.2 ng m⁻³ when associated with PM10 and 12.6 ng m⁻³ when associated with PM2.5 (Tables S2 and S3). This values are similar or slightly lower to these ones obtained by Kaleta and Kozielska³¹. They measured PM-bound BaP concentration in several cities in Silesia Voivodeship in Poland. In the 2018, 2019, 2020 and 2021 heating seasons in Rybnik city, the concentrations were 23.5, 27.7, 15.9 and 15.7 ng m⁻³.

The prevalence of 4-, 5-, and 6-ringed PAHs can be a suggestion of the influence of specific sources and contributing factors during the sampling period. During periods with the lowest PM/PAHs concentrations (specifically on Nov 14/15, 2014, Nov 19/20, 2014, and Feb 02/03), a noticeable shift in the PAHs profile was observed. This shift was observed in both fractions PM10 and PM2.5 (Fig. 4C and D). During these days, the proportion of lighter PAHs (2- and 3-ringed PAHs) increased to approximately 20–30% of total mass concentration of PAHs. Moreover, the share of PM2.5-bound PAHs in PM10-bound PAHs differed even in episodes of elevated total concentrations. On Nov 24/25, 2014 PAHs associated with smaller PM fraction contributed in 42.5% in PM10-bound PAHs mass, whereas on Feb 12/13, 2015—in 83.9%. This observation can be a hint for the origin of PAHs in the atmosphere. This will be discussed in 3.3.

Comparing these results with our data, we observe similar patterns, where PAHs with a higher number of rings dominate in both particulate matter fractions, indicating a significant influence of fossil fuel combustion and traffic emissions. In a study conducted in Nanjing³², PAHs associated with PM10 and PM2.5 were analyzed in different functional areas of the city. It was found that PAHs with a higher number of rings (5- and 6-ring compounds) dominated in both particulate matter fractions, suggesting a significant contribution from vehicle emissions and coal combustion. In a comparative study conducted in various cities in China³³, PAHs associated with PM2.5 were analyzed to identify major emission sources and assess health risks. It was found that the dominant sources were coal combustion, industrial emissions, and traffic, with higher PAH concentrations observed during winter, which was attributed to the increased use of fossil fuels for heating.

The study of potential origin of polycyclic aromatic hydrocarbons

Examination of correlations to other PM components

The examination of statistically significant correlations (p 2 values equal to or greater than 0.6) for PM10 and 5 for PM2.5 (Table 3). The most pronounced correlation was observed between organic carbon and total PAHs (0.89 for PM10 and 0.95 for PM2.5). This indicates a direct link between variations in PAH concentrations during the winter and factors responsible for organic carbon production, such as the combustion of fossil fuels. Interestingly, a similar relationship was not identified for elemental carbon, mitigating the potential influence of transportation³⁴.

Furthermore, a substantial correlation (0.87 for PM10 and 0.84 for PM2.5) was identified concerning the excess of chlorine ions (ExCl⁻), calculated as the discrepancy between equivalent concentrations of total chlorides and chlorides bound with sodium cations to form NaCl. As reported by Szramowiat-Sala et al.⁴ combustion processes involving coal and coal-related fuels generate elevated chloride concentrations. Therefore, based on the significant correlation of PAHs with ExCl⁻, we can infer that the combustion of fossil fuels serves as a primary source of PAHs. Additionally, the correlation between PAHs and levoglucosan (0.88 for PM10 and 0.94 for PM2.5) suggests the combustion of biomass as a further potential origin for PAHs. This finding is further substantiated by the significant correlation between PAHs and K⁺ (0.65 for PM10).

However, in addition to the strong correlation between PAHs and ExCl⁻ it is important to consider the potential influence of biomass burning as well. Biomass combustion, particularly from wood burning in residential heating, is a significant source of chloride emissions in atmospheric aerosols³⁵. Studies have shown that biomass burning releases substantial amounts of chlorinated organic compounds and inorganic chlorides, including KCl, which can undergo chemical reactions in the atmosphere, leading to the formation of secondary chlorine-containing species²⁸. The observed correlation between PAHs and ExCl⁻ could, therefore, be partially linked to biomass combustion processes, which co-emit both species.

The simultaneous presence of levoglucosan, a widely recognized tracer of biomass burning³⁶, in the analyzed samples further supports this hypothesis. The correlation between levoglucosan and PAHs (0.88 for PM10 and 0.94 for PM2.5) suggests that part of the PAH load may derive from biomass rather than exclusively from fossil fuel combustion. Furthermore, potassium (K⁺), another key marker of biomass combustion, also exhibits a moderate correlation with PAHs (0.65 for PM10), reinforcing the idea that biomass burning plays a non-negligible role in PAH emissions in Kraków.

It is essential to consider that the chemistry of biomass burning emissions differs from that of coal combustion. While coal-related emissions are often enriched in sulfur compounds (such as SO₂), biomass combustion produces a distinctive mixture of organic compounds, including oxygenated PAHs and chlorinated derivatives³⁵. These differences could influence the atmospheric processing and deposition of PAHs, potentially altering their reactivity, transport dynamics, and toxicological impacts. Given the high correlation of PAHs with ExCl⁻, further chemical characterization of chloride-containing species in PM could help delineate the relative contributions of coal versus biomass burning, improving source apportionment efforts and air quality mitigation strategies.

An intriguing aspect worth noting is the correlation between PAHs and another component of organic carbon, namely HULIS (0.69 for PM10). This correlation becomes even more pronounced when considering heavier PAHs, with 5 and 6-ringed PAHs associated with PM10 showing correlations with HULIS ranging from 0.66 to 0.84, and 6-ringed PAHs associated with PM2.5 exhibiting correlations with HULIS in the range of 0.63 to 0.82. The production of humic-like substances is linked to both combustion processes and the formation of secondary organic aerosols in heterogeneous atmospheric reactions^37,38. This suggests that PAHs may undergo secondary formation processes, similar to humic-like substances, especially on days when PAH concentrations are relatively low (e.g., Feb 02/03, 2015, and Nov 14–20, 2014). During periods of elevated PAH concentrations, which coincide with increased PM10 and PM2.5 levels, the origin of PAHs is likely more complex. In addition to emissions related to combustion and secondary reactions within Kraków, consideration must be given to transboundary long-range transport as a factor contributing to pollutant origins. To illustrate this, Fig. 5A–C displays backward trajectories (calculated over 72 h for three different altitudes: 300, 800, and 1200 m above sea level) for three sampling periods characterized by varying PAH profiles: December 15/16, 2014, Feb 12/13, 2015, and Mar 05/06.

During these periods, air masses arrived from both the West (Fig. 5A and C) and the South (Fig. 5B). In Fig. 5A and C, the trajectory of the highest air mass appears quite typical. However, in Fig. 4B, the lowest air masses initially arrived from higher altitudes and then descended even lower than 300 m above sea level. Conversely, the highest air masses originated from lower altitudes. This observation suggests the occurrence of temperature inversion phenomena and the substantial accumulation of pollutants above the city, resulting in elevated concentrations of PM and PAHs on Feb 12/13, 2015. During this period, the origin of PAHs likely had a predominantly local character.

A similar situation occurred on Mar 05/06, 2015, when the proportion of heavier PAHs was highest (as depicted in Fig. 4C and D). However, the extended green line in Fig. 5C serves as evidence of long-range transport, which, in addition to local emission sources, evidently contributed to PAH formation.

Investigation of PAHs origin through diagnostic ratios analysis

To ascertain the source of PAHs, researchers commonly employ diagnostic ratios (DRs) derived from the profiles of polycyclic aromatic hydrocarbons. This practice is rooted in the understanding that varying combustion conditions result in distinct PAH profiles³ and different compositions of particulate matter⁴. Binary diagnostic ratios of PAHs, constructed from pairs of isomeric individual PAHs, serve as effective tools for pinpointing potential sources³⁹. The indicators used and their potential associations with specific sources are detailed in Table 4.

The diagnostic ratio Anth/(Anth + Phe) less than 0.1 signifies petroleum emissions, whereas values greater than 0.1 indicate biomass combustion. Similarly, a Fla/(Fla + Pyr) ratio below 0.4 is associated with petroleum emissions, values between 0.4 and 0.5 indicate natural gas combustion, and values exceeding 0.5 suggest biomass or coal combustion. For IcdP/(IcdP + BghiP) and BaA/(BaA + Chr), ratios below 0.2 signify petrogenic sources. Ratios between 0.2 and 0.35 for BaA/(BaA + Chr) and between 0.2 and 0.5 for IcdP/(IcdP + BghiP) suggest mixed sources, such as fossil fuel combustion, crude oil, or vehicle emissions. Ratios exceeding 0.5 for IcdP/(IcdP + BghiP) and BaA/(BaA + Chr) are indicative of biomass and coal combustion.

Additionally, the ratio BaP/(BaP + BeP) provides insights into the aging of PAHs. Values greater than 0.5 indicate aged emissions, typically associated with biomass and coal combustion. In contrast, values less than or equal to 0.5 represent fresh emissions, commonly linked to traffic and vehicular sources. BaP/(BaP + BeP) is sensitive to photochemical degradation, as BaP degrades faster than BeP in the atmosphere due to its higher reactivity⁶.

Table 5 presents the mean values and ranges of specific DRs calculated for PM10 and PM2.5. The results consistently indicate that biomass and coal combustion were the dominant PAH sources, with some influence from vehicular emissions. The Flu/(Flu + Pyr) and Fla/(Fla + Pyr) ratios, which help differentiate between fossil fuel combustion and biomass/coal burning, show values exceeding 0.5 for both PM10 and PM2.5, confirming that combustion processes were the primary contributors to PAH pollution. The IcdP/(IcdP + BghiP) ratio, which distinguishes between fossil fuel combustion and biomass/coal combustion, shows low values (0.09 for PM10 and 0.10 for PM2.5), suggesting a strong influence of fossil fuel combustion, likely from coal and traffic emissions. Similarly, the BaP/(BaP + Chr) and BaA/(BaA + Chr) ratios exceed 0.35, further reinforcing the dominance of pyrogenic sources, particularly biomass and coal burning.

Differences between the two particulate fractions indicate that PM2.5 was more affected by atmospheric aging and long-range transport, while PM10 contained relatively fresher emissions from local combustion sources. The BaP/(BaP + BeP) ratio, which assesses photochemical aging, is higher in PM2.5 (0.63) than in PM10 (0.56), suggesting that PAHs bound to fine particles underwent more atmospheric degradation. Additionally, the BaP/BghiP ratio, which differentiates between traffic-related and non-traffic sources, remains close to 0.7 in both fractions, pointing to a mixed influence of vehicle emissions and other combustion sources.

In order to enhance the PAHs source apportionment the diagnostic ratio cross plots have been created (Fig. 6). The Flu/(Flu + Pyr) vs. IcdP/(IcdP + BghiP) (Fig. 6A) plot allows differentiation between petrogenic and pyrogenic sources. In the studied samples, values of Flu/(Flu + Pyr) exceeding 0.5 suggest that PAHs originated predominantly from biomass and coal combustion, while values below 0.5 indicate contributions from fossil fuel combustion. Similarly, the IcdP/(IcdP + BghiP) ratio helps distinguish between fossil fuel combustion (values below 0.5) and biomass/coal combustion (values above 0.5). The clustering of data points in the pyrogenic region further confirms that incomplete combustion processes, particularly residential heating with coal and biomass, were dominant PAH sources in the study area. Another important plot, BaP/(BaP + Chr) vs. IcdP/(IcdP + BghiP) (Fig. 6B), provides insights into the emission sources by distinguishing between petrogenic emissions (BaP/(BaP + Chr)  0.35). The results indicate that most data points fall into the biomass and coal combustion category, with only a few samples showing indications of mixed-source pollution. This suggests that domestic heating emissions significantly contributed to PAH levels during the sampling period, which aligns with the peak wintertime pollution episodes recorded in Kraków. The Anth/(Anth + Phe) vs. Flu/(Flu + Pyr) plot (Fig. 6C) supports these findings by further distinguishing between biomass burning and fossil fuel combustion. The Anth/(Anth + Phe) ratio exceeding 0.1 in most samples confirms the dominance of combustion-related PAHs, while Flu/(Flu + Pyr) values between 0.4 and 0.5 indicate a mix of fossil fuel and biomass burning. These observations align with the region’s known pollution sources, where both coal-fired heating and wood burning contribute significantly to ambient air quality deterioration. Lastly, the BaP/BghiP vs. IcdP/(IcdP + BghiP) plot (Fig. 6D) provides additional evidence of source contributions, specifically differentiating between traffic-related and non-traffic sources. The results show that BaP/BghiP values are generally above 0.8, suggesting a dominant non-traffic source, such as residential heating and industrial emissions. The IcdP/(IcdP + BghiP) values confirm this trend, with most values exceeding 0.5, further reinforcing the strong influence of biomass and coal combustion.

The diagnostic ratio analysis in this study aligns well with previous research on PAH source apportionment, reinforcing the dominance of biomass and coal combustion as primary contributors to air pollution. Consistent with Sofowote et al.⁴², the Ant/(Ant + Phe) and BaA/(BaA + Chr) ratios effectively differentiate combustion sources from petrogenic emissions, though our results indicate a stronger influence of residential heating rather than industrial emissions. Similar to Tobiszewski and Namieśnik⁷, our study highlights the importance of combining multiple diagnostic ratios, allowing for a clear distinction between vehicular emissions and biomass/coal burning. The Flu/(Flu + Pyr) and IcdP/(IcdP + BghiP) ratios confirm the predominance of pyrogenic PAHs, with a notable contribution from coal combustion. Moreover, in agreement with Onaiwu and Ifijen⁴³, our findings reveal that major contributors to PAHs included gasoline combustion, diesel combustion, traffic emissions, and emissions from vehicle panel welding. However, they also concluded that application of diagnostic ratios has some limitations. Mainly due to the fact that PAHs can originate from various sources, and ratios may not always definitively indicate the primary source.

PAHs toxicity and their relative carcinogenicity

Polycyclic aromatic hydrocarbons belong to the group of most hazardous chemical substances that can be found in the environment due to their carcinogenic and mutagenic properties. Their influence is especially dangerous when their concentrations, either in the atmosphere, water or other media are elevated continuously⁴⁴.

The carcinogenicity of PAHs usually increases with the number of organic rings in the structure. Thus, five-ringed aromatic hydrocarbons, such as benzo(a)pyrene, benzo(b)fluoranthene and dibenzo(a,h)anthracene have the strongest mutagenic and carcinogenic properties⁴⁵. Benzo(a)pyrene was classified as a PAHs presence indicator due to its common occurrence, high concentrations among them and relatively high carcinogenicity.

The relative carcinogenicity coefficient is a base for the determination of the PAHs exposure index which indicates the intensity of adverse affection for the mixture of given PAH compounds, giving information about the potential hazard of carcinogenic air pollutants. The Interdepartmental Commission on NDS and NDN Registration (ICNNR) in Poland introduced a specific index to describe the level of PAH exposition (A) which is defined as shown in Eq. (2):

where:

where A_i—reference index of carcinogenicity for given i-PAH. C_i—i-PAH average concentration during the study period; k_i—relative carcinogenicity coefficients for given i-PAH⁴⁶.

The reference index of carcinogenicity for Kraków during the study period was calculated using the average PAHs concentrations measured for PM2.5 fraction due to its probably higher hazardous impact on human health. The reference index of carcinogenicity reached the value of 0.001 mg m⁻³. Table 6 shows the results of exposition indexes computed for every sampling day. It can be noticed that both daily and summary emissions do not exceed the specified ICNNR limit of 0,002 mg m⁻³. As the computations proceeded for 11 samplings, the obtained results cannot be compared to the annual limit of PAHs exposition as mentioned before.

The mutagenic and toxic health hazard of the PAHs mixture were quantitatively expressed as the MEQ and TEQ, respectively, relating to the mutagenic potential of BaP⁴⁶ and toxicity of the 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)⁴⁷. The TEQ and MEQ were calculated as expressed by Eqs. (4) and (5) for each site, separately for PM2.5 and PM10, based on concentrations and toxicity/mutagenicity equivalence factors for given PM-bound PAHs. In addition, the contribution of carcinogenic PAHs to total measured PAHs content (ΣPAH_carc/ΣPAH) was also determined for PM10 and PM2.5 fractions (Table 7), with the approach described by Eq. (6).

As presented in Table 7, the carcinogenic PAHs contributed to the total PAHs in ca. 50% both in PM10 and PM2.5 in November, December and February. The mutagenic activity of PAHs was mostly visible in the PM2.5 fraction exhibiting the higher mean values of MEQ in the contrary to MEQ_PM10. A similar relation was reported in the toxic activity of PAHs: PM2.5-bound PAHs were more toxic during the conducted measuring campaigns than PM10-bounded PAHs. On the basis of comparison of MEQ and TEQ values, it may be stated that mutagenic activity was from ca. 20 to 30 times higher than toxic activity of these pollutants. Liu et al.³² calculated the toxic equivalency factors (TEFs) to assess the carcinogenic risk of PAH mixtures, expressing the toxicity in terms of benzo[a]pyrene equivalent concentrations (BaPTEQ). They found that the average BaPTEQ concentrations were 3.14 ± 1.27 ng/m³ for PM2.5 and 8.23 ± 1.55 ng/m³ for PM10, indicating significant health risks associated with PAH exposure in these particulate fractions. The toxicity and mutagenicity values reported by Kozłowska and Kozielska²⁹ were significantly lower than those observed in our study, with differences reaching several orders of magnitude. Despite this discrepancy in absolute values, both studies indicate a consistent trend—PAH mutagenicity was found to be stronger than its toxicity, suggesting that PAHs have a considerable potential to induce genetic mutations, even at lower toxicity levels. Furthermore, their findings revealed that PAHs bound to PM1 exhibited higher carcinogenic potential compared to those associated with PM2.5, a pattern that aligns with our results. This observation supports the hypothesis that finer particles (PM1 and ultrafine fractions) pose a greater risk to human health, as they not only contain a higher proportion of high-molecular-weight PAHs but also have an increased ability to penetrate deeper into the respiratory system. Our study reinforces this conclusion, as PAHs bound to smaller particulate fractions demonstrated a greater carcinogenic and mutagenic impact, highlighting the importance of monitoring fine and ultrafine PM-bound PAHs in air pollution assessments.