Investigating the concentration levels, distribution patterns, source identification and health risk assessment of PAHs, n-alkanes, hopanes, and steranes in deposited dust of Mashhad, Iran


 Deposited dust (DD) in urban environments contains carcinogenic organic compounds. The Indoor air quality is greatly affected by heating, ventilation, and air conditioning systems (HVAC), and in the Middle East most of the buildings are equipped by HVAC on top of them. It is possible that the DD on the roof near this equipment would be transferred to an indoor area. For these reasons, 40 samples of the over the roof DD were prepared, and organic compounds (16PAH compounds, 20n-alkane homologs, 8hopanes, and 6steranes) of DD were extracted using Soxhlet and analyzed by GC-MS. Source identification of organic compounds conducted by ring classification, diagnostic ratios, and factor analysis (FA). The results showed that the average (±SD) of total PAHs, n-alkanes, hopanes and steranes in DD were 1356.00 (±291.45) ng kg−1dw, 3211.65 (±969.18), 146.37 (±79.45) and 469.76 (±188.25) µg.g_1dw, respectively. The highest concentration of organic compounds was in the city center, where traffic congestion is common. Diagnostic ratios of n-alkanes results showed the dominant source is vehicular emission. FA results indicated vehicular emission and biogenic sources. In agreement, the results of sterane and hopane profiles confirm these results. On the other hand, the PAHs diagnostic ratios results indicated petroleum combustion sources. In this regard, FA findings showed combustion from vehicular emission and natural gas and wood combustion were the main factors. Furthermore, the incremental lifetime cancer risk was calculated as 8.45× 10−12 for children and 9.80 × 10−7 for adults, and the imposed risk was negligible.


Introduction
Since people in urban areas experience many indoor hours, it is crucial to maintain good indoor air quality ). Due to many prior studies, the applied system for heating, ventilation, and air conditioning (HVAC) in a building can considerably alter the quality of indoor air. The HVAC system can in uence supplied air quality, and a polluted supply air system can deteriorate indoor air quality (Holopainen et al. 2002). According to one report, HVAC dust could provide an integrated indicator of indoor airborne pollution levels (Tringe et al. 2008). Several researchers have also looked into the concentrations of heavy metals and PAHs in household dust, as well as their relationships with possible sources (indoor and outdoor) and distribution of particles' size (Al-Rajhi et al. 1996, Al Ali et al. 2017, Azimi et al. 2020, Chattopadhyay et al. 2003, Khedidji et al. 2013, Škrbić et al. 2019. PAHs concentrations in the air were assessed individually at indoor and outdoor sites, and higher concentrations of PAHs were detected in indoor samples than the outdoor ones (Pandit et al. 2001). Delgado-Saborit et al. investigated the PAHs concentrations in indoor and outdoor environments and their total carcinogenic potential. The results have shown that indoor PAHs concentrations are higher than the outdoor ones at buildings close by tra c (Delgado-Saborit et al. 2011).
Wintertime PAH concentrations in indoor and outdoor air samples at a large dormitory for students in Algiers revealed the superiority of indoor PAH levels in compared to outdoor levels, posing a signi cant health risk (Khedidji et al. 2013).
Particulate matter (PM) is a signi cant air pollutant that degrades air quality in arid and semi-arid areas (Soleimani et al. 2016). Residential exposure to indoor contaminants has been measured using deposited dust (DD).
Environmental and seasonal in uences, ventilation and air ltration, residents' habits, and indoor and outdoor pollutants sources all affect the composition of indoor dust (Maertens et al. 2004). Heating, ventilation, and HVAC systems substantially affect indoor air quality, and most buildings in the Middle East have HVAC on the roof. By providing an empirically based, post-hoc exposure metric at the household scale, DD on the roof can provide a valuable means of re ning exposure estimates in such contexts. Furthermore, the particle size of DD near the ground surface is smaller than that near the top of the roof (Kramer et al. 2020). It comprises of clumps of material that come primarily from outside sources, such as aerosol and soil deposition, rather than from domestic operations (Sajn 2005).
Several studies worldwide have looked into persistent organic pollutants (POPs) like PAHs that belong to PM and DD (Al Ali et al. 2017, Dong &Lee 2009, Mostafa et al. 2009, Omar et al. 2007, Škrbić et al. 2019, with high levels of PM and DD identi ed in some of Iran's major cities (Azimi et al. 2018, Ghanavati et al. 2019, Keshavarzi et al. 2020, Saeedi et al. 2012, Soltani et al. 2015. PAHs, n-alkanes, n-alkanoic acid, hopanes, steranes, and other compounds make up the organic fraction of DD. Some of these compounds, such as PAHs, hopanes, and steranes, are persistent and potentially harmful, posing a signi cant health risk as they can cause mutagenic and carcinogenic effects . These various organic compounds, on the other hand, can be used as a source recognition and health risk assessment marker (Javed et al. 2019).
Molecular marker analysis may also be used to classify organic matter origins, allowing for comparing identi ed sources and observed atmospheric organic compound mixtures (Mostafa et al. 2009, Omar et al. 2007, Zhang et al. 2017). PAHs and other organic pollutants may be adsorbed onto PM (Ghanavati et al. 2019, Javed et al. 2019, Najmeddin et al. 2018. PAH compounds are biodegradable and may not be stable in the environment after they have been released. Other markers, in addition to PAHs, could aid in the identi cation of particle sources. As a result of their low reactivity and volatility, n-alkanes are solid indicators for air transport and particle sources (Omar et al. 2007, Zhang et al. 2017. Photochemical and microbial degradation can hardly in uence other petroleum indicators, such as hopanes and steranes, because they are complex cyclic molecules (Alves et al. 2018). As a result, these markers will produce precise source information that is critical to environmental forensics investigations to research hydrocarbon fate and activity in the environment (Mostafa et al. 2009, Omar et al. 2007).
Comparing the diagnostic ratios (DRs) of commonly occurring organic organisms' pairs can lead to the characterization of organic pollutant sources' variation (Ravindra et al. 2008). Hence, there was a broad application of molecular DRs to qualitatively de ne the sources of air pollutants (Azimi et al. 2018, Bahry et al. 2009, Omar et al. 2007). However, these ratios should be used with caution due to the variations in pollution sources and the decomposition of most of the organic-based contaminants within the atmosphere due to their reactions with the oxidants present in the atmosphere (Moeinaddini et al. 2014a, Zakaria et al. 2000.
The DD could be transported indoors in the Middle East by HVAC from the roof's top. Because of the importance of DD on the roof, as noted previously, a source study and risk assessment of DD were needed. The project aimed to measure the concentration and spatial distribution of PAHs, n-alkanes, hopanes, and steranes present in DD, as well as pollution levels, source identi cation, and an assessment of the potential cancer risk related to PAH exposure from DD through dermal contact, ingestion, and inhalation.
It is located at an elevation of 985 meters above sea level and has a population of over three million (Fig. 1). The city has experienced rapid growth in recent decades, owing to its economic, social, and religious attractions (Azari &Arintono 2012). It also hosts 20 tourism events each year. Mashhad has public transportation, including a subway and a bus rapid transit system (BRT) (Azari &Arintono 2012).

Sampling and Extraction method
The 40 DD samples were taken from building roofs in different urban land uses (approximately 3-4 m above ground level) (Fig. 1). The sampling was carried out in the summer, when not only is the temperature the highest and the use of HVAC is the highest, but also the frequency of dust storms is higher than in other seasons (Azimi et al. 2018, Najmeddin &Keshavarzi 2019. The samples were obtained by sweeping 1 m 2 area of the building roof and preserving them in aluminum foil in a zip bag for transport to the laboratory (Jordanova et al. 2012, Wang et al. 2011).
DD samples were dried at room temperature for 24 h, sieved with a 1-mm stainless steel sieve to remove large materials, and placed in a refrigerator at −4°C for analysis (Dong &Lee 2009, Škrbić et al. 2019. After that, the samples were freeze-dried for 72 hours, and 5 g samples was prepared to extract organic compounds. For quality control (QC), the PAH surrogate internal standard mixture (200 µg kg −1 each of naphthalene-d 8 , anthracene-d 10 , perylene-d 12 and chrysene-d 12 ) was added to samples. Before analysis, about 5 g of each freeze-dried sample was spiked with 100 µL of n-alkanes surrogate internal standard (5 µg g −1 of n-dodecane-d 26 ) for aliphatic compounds (n-alkanes, hopanes, and steranes). The extraction and fractionation procedure was based on Zakaria et al. (2000) ( Zakaria et al. 2000) 's Soxhlet process, which used 100 mL of dichloromethane for 12h. In summary, this process involves two steps of silica gel chromatography followed by gas chromatography-mass spectrometry analysis of the analytes (GC-MS). Glassware was rinsed with methanol, acetone, and hexane to prevent any contamination, and then heated under a 60°C temperature for two hours. The external standard solutions of targeted petroleum hydrocarbons were procured from the Sigma Chemical Company. 16 PAHs were analyzed containing naphthalene (Nap), acenaphthylene (Acy), acenaphthene (Ace), uorine (Flo), phenanthrene (Phe), anthracene (Ant), uoranthene Gas chromatograph-mass spectrometer (GC-MS) analyses were performed using GC-7890A with quadrupole MS-5975C, Agilent Technologies (PaloAlto, CA, USA). Helium was used as the carrier gas. Sample peak results were detected by comparing the sample results with the retention time of the authentic standard run on the same day con rmed by GC-MS. In the alkane fraction of samples, biogenic hopanes and diagenetically modi ed tri-and tetracyclic terpanes, hopanes, and steranes were quanti ed. The single ion monitoring (SIM) mode was used to identify petroleum biomarkers: m/z 191 for tri-and tetracyclic terpane and hopane, m/z 217 for ααα-steranes and m/z 218 for αββ-steranes (Rogge et al. 1993b). The retention times of n-alkanes were compared to known standards of n-alkanes ranging from n-C 14 to n-C 32 to identify them.

Quality assurance and control (QA/QC)
A eld blank, a spiked blank, a procedural blank, and a matrix spiked sample were used for each sample (Bakhtiari et al. 2009, Zakaria et al. 2000. Deuterated n-alkanes spiked in each sample were used to estimate the losses of all compounds during sample processing. The average recovery for PAHs ranged between 87% and 104%, n-alkanes ranged between 81% and 94%, while for hopanes and steranes it varied from 86-101%. The relative recoveries of PAHs, n-alkanes, and biomarkers were used to correct their concentrations.

Organic Compounds Source identi cation
In this research, widely used DRs for PAHs, hopanes, steranes, and n-alkanes indices were measured and compared to the literature to get an initial implication of the organic aerosol emission sources (Javed et al. 2019, Mostafa et al. 2009, Omar et al. 2007. DRs have been effectively used as indicators of PAH sources by various studies (Al Ali et al. 2017, Dong &Lee 2009, Mon et al. 2020, Ravindra et al. 2008, Wang et al. 2011) by considering the relative thermodynamic stability of different parent PAHs and the characteristics of different sources (Bian et al. 2016). The methods used to determine these ratios and indices can be found elsewhere (Bakhtiari et al. 2011, Yadav et al. 2013). To nd potential PAH sources, DRs for Flt/(Flt + Pyr), BaA/(BaA + Chr) and InP/(InP + BghiP) were used. An Flt/Flt+Pyr ratio of less than 0.4 indicates mostly petrogenic sources, 0.4-0.5 indicates fossil fuel combustion, and greater than 0.5 indicates coal and biomass combustion. If the BaA/BaA+Chr ratio is less than 0.2, it means petroleum evaporation, between 0.2 and 0.35, coal combustion, and greater than 0.35, vehicular emission or combustion (Liu et al. 2009, Yunker et al. 2002. A ratio of InP/InP+BghiP of less than 0.2 implies gasoline, 0.2-0.5 petroleum combustion, and > 0.5 combustion of coal or biomass (Yunker et al. 2002).
The carbon numbers of n-alkanes, as measured by a carbon preference index (CPI), the percentage of wax (WNA%), and the ratio of unresolved components to resolved components in gas chromatograms (U:R) have all been used in the literature to identify between petrogenic (evaporation or combustion) and biogenic sources of n-alkanes (Moeinaddini et al. 2014b). The CPI is de ned as the ratio of the total concentration of odd carbon number homologues in the sample to the total concentration of even carbon number homologues (Andreou et al. 2008, Duan et al. 2010, Zhang et al. 2017. CPI values are used to compare the contribution of vehicular and human activities to biological sources (Moeinaddini et al. 2014a, Ravindra et al. 2008, with values close to unity indicating vehicle emissions and other human activities, and values >3 indicating a predominant origin from biological materials. The WNA% value (WNA = C n -[(C n+1 + C n−1 )/2] where n is the odd carbon congener, WNA% = (∑WNA/∑n-alkanes)*100) is calculated for odd carbon number homologues, and it is used to indicate biogenic as opposed to petrogenic sources (Javed et al. 2019, Omar et al. 2007, Zhang et al. 2017. (Andreou et al. 2008, Rogge et al. 1993a).
The U:R was applied to determine the pollution level caused by the residues of petroleum, and it can also be used to differentiate between the sources; i.e. biomass and fossil fuel combustion (Moeinaddini et al. 2014a, Mostafa et al. 2009, Rogge et al. 1993a, Shirneshan et al. 2017. Hopanes, as shown in Table S1, are a series of 17α (H), 21β (H) compounds used in this analysis (C 27 to C 34 ) ( Table   S1 in supplementary section). To characterize petroleum inputs, several geochemical ratios derived from hopane biomarkers were used (Shirneshan et al. 2017). The following ratios were used to analyze hopane distribution patterns further: T s /T s +T m (ratio of 17α-22,29,30-trisnorhopane relative to 17α-22,29,30-trisnorhopane+18α-22,29,30 trisnorhopane), C 29 /C 30 (ratio of 17α,21β(H)-30norhopane to 17α,21β(H)-hopanes), C 31 HS/C 31 H(S+R) (homehopane index) C 32 HS/C 32 H(S +R) (bishomohopane index) SC 31 _C 35 /C 30 (ratio of sum 17α,21β(H)-C 31 homohopane to 17α,21β(H)-C 35 homohopane relative to 17α,21β(H)hopane), C 28 αββ/(C 27 αβα + C 29 αββ) and C 29 αββ/(C 27 αββ + C 28 αββ) Besides that, factor analysis (FA) was applied to nd the PAHs' and n-alkanes' sources (SPSS version 20.0 for Windows, SPSS Inc.). The Kaiser-Meyer-Olkin (KMO) and Bartlett's tests were also used to assess the suitability of the dataset concerning the FA; for KMO values of more than 0.6, the data was considered suitable (Moeinaddini et al. 2014a, Ravindra et al. 2008. The FA of PAHs and n-alkanes were conducted separately based on the KMO test results. The varimax rotation method was used. The principal components were chosen where eigenvalues had more than 1. The representative species of the factor were PAHs and n-alkanes with a factor loading of >0.5 (Ravindra et al. 2008, Wang et al. 2011).

Cancer risk assessment
The toxic equivalency factor (TEF) of the PAHs was assessed to calculate their carcinogenic potential. The TEF value of Benzo[a]pyrene (BaP) was set to 1 as the reference chemical, and the TEF values of other PAHs were calculated against BaP (Nisbet &Lagoy 1992, Škrbić et al. 2019. Total Equivalency Factors (TEFs) should be compared to those of BaP, the most toxic member of the PAH family, to measure the carcinogenicity of individual PAHs (BaP eq ). (Nisbet &Lagoy 1992). TEFs were added to each PAH compound, and the toxic equivalent concentration (TEQ) was determined by adding each PAH concentration and its TEF using equation 1: Where C i is the PAH i concentration and TEF i is the PAH i toxic equivalency factor. Table 1 presents the TEF values for all PAHs based on Malcolm and Dobson (1994) and Nisbet and LaGoy (1992) (Malcolm &Dobson 1994, Nisbet &Lagoy 1992. The incremental lifetime cancer risk (ILCR) was determined using standard USEPA models to calculate the exposure risk of PAHs (Ma et al. 2017, Martuzevicius et al. 2011. Equations 2-4 were used to measure the ILCRs of ingestion, dermal touch, and inhalation:

Results And Discussion
3.1. Spatial distribution and Source identi cation of n-alkanes Figure 2 illustrates the spatial distribution of n-alkanes (n-C 14 -n-C 33 ) concentration. The highest concentrations ranged from 2131.3 to 5964.6 µg.g _1 dw, with an overall mean of all tests equal to 3159.1 µg.g _1 dw. S1 (6562.0 µg.g _1 dw), S2 (5964.6 µg.g _1 dw), S3 (5485.7 µg.g _1 dw) and S4 (5286.3 µg.g _1 dw) had the highest concentrations ( Figure 2). These sites were in the city's core, near the holy shrine, and were highly tra cked. S9 has a concentration of 4004.56 µg.g _1 dw and is located near some petrol and car service stations, making it relatively polluted due to the heavy tra c load. At the other hand, the residential area (S26) had the lowest n-alkanes concentration (2131.28 µg.g _1 dw). The DD from Mashhad has lower n-alkanes concentrations than street dusts from other cities, such as Portugal (197-9982 µg.g _1 dw (Alves et al. 2018)), Malaysia (7360 µg.g _1 dw (Omar et al. 2007)) and Singapore (3760 µg.g _1 dw (Zhang et al . 2017)). Table S3 shows the DRs results, including the CPI, WNA%, U:R ratio, and Pr:Ph. The CPI values indicated that petroleum residues derived from vehicular emissions are the chief source of n-alkanes in most sampling sites, whereas higher plant waxes are found in a few sites. The lower CPI values (0.84-1.52) backed up these ndings (Table S3) (Bakhtiari et al. 2011, Omar et al. 2007). The majority of sites were close to unity, but S8, S14, and S28 were all >1 at 1.52, 1.41, and 1.36, respectively. S8 is near Alandasht (10 ha), S14 is near Malek-Abad orchard (>50 ha), and S28 is located on the Shandiz-Torqaba road, where many orchards are located.
GC-MS traces of hydrocarbons or total extracts detected in urban DD usually consist of large unresolved complex mixtures (UCM) composed of branched and cyclic (petroleum) compounds (Fig. S1, supplementary section). The majority of these hydrocarbons come from the use of fossil fuels, which comprise the main components of both diesel and auto engine exhaust (Rogge et al. 1993b, Simoneit 1984, 1985. Biogenic hydrocarbons derived from higher plants exhibit no UCM (Simoneit &Mazurek 1982). A UCM in the range of n-C 20 -n-C 33 alkanes was observed in all samples, indicating samples were polluted by biodegraded petroleum residues (Andreou et al. 2008, Mostafa et al. 2009).
The U:R ratio (i.e., the UCM ratio to plant wax n-alkanes) can be used to determine pollution from petroleum product combustion of vehicular sources since the UCM is primarily formed by fossil fuel applications (Jafarabadi et al. 2018, Jafarabadi et al. 2019, Omar et al. 2007. Besides that, the U:R ratio was used as a contaminant input diagnostic criterion, where values >2 show signi cant pollution by petroleum products (Jafarabadi et al. 2019). The U:R ratios for all samples were greater than 2, indicating petrogenic sources. This was in line with the UCM pro le and low CPI values (Table S3). The U:R ratios were applied to express the extent of vehicular exhaust contribution to street dust hydrocarbons (Simoneit 1999) which increases as vehicular exhaust contribution increases. The U:R ratio is much less than that estimated in road dust in Athens, Greece (Andreou et al. 2008) but much higher for street dust in Egypt (Mostafa et al. 2009). A comparison of n-alkanes DRs in DD with PM and street dusts of other cities worldwide are given in Table S4.
The most abundant petroleum isopranoids are pristane (2,6,10,14-tetramethylpentadecane, C 19 isoprenoid) and phytane (2,6,10,14-tetramethylhexadecane, C 20 isoprenoid) (Shirneshan et al. 2017). They can be found in diesel fuel, lubricating oil, from gasoline and diesel engine exhaust (Simoneit 1984(Simoneit , 1985. The presence of pristane and phytane in the UCM, and also the Pr:Ph ratio of 0.18 to 0.76 implied petroleum residue pollution (Omar et al. 2007). Figure S2 (supplementary section) and Table 2  23.57% of the total variance, is associated with longer chain n-alkanes (>nC 24 ). As a result, it was concluded that the signi cant sources of n-alkanes in the DD are most likely vehicle exhaust and fuel combustion.  Fig. S3.
The mean concentration of hopanes was 2209.8 (±SD) µg.g _1 dw. The highest hopanes concentrations were observed at S2, S1, S3 and, S4, equal to 4860.1, 4241.0, 4193.9, and 3940.6 µg.g _1 dw, respectively. These stations are in the city center, close to the holy shrine, and have heavy tra c. S38 (3120.50 µg.g _1 dw) is located at the passenger bus terminal, where heavy bus tra c leads to high hopane concentrations. On the other hand, the lowest concentration was found in the residential area at S19 (979.34 µg.g _1 dw). Hopanes levels in this study were relatively lower than the ones measured in street dust from Portugal (2706 µg g −1 (Alves et al. 2018)) and Anzali, Iran (5225.58 µg g −1 (Azimi et al. 2018)), but higher than those in Singapore (153.47 µg g −1 (Zhang et al. 2017)) and Malaysia (389 µg g −1 (Omar et al. 2007)).
In this study, the ααα-and αββ-steranes ranging from C 27 to C 29 were analysed to be 20S and 20R isomers in the DD (Omar et al. 2007). The mean concentration of steranes was 469.79 (±SD) µg.g _1 dw. The highest concentrations were found in the same locations as for the hopanes (S2, S3, S1, and S4). On the other hand, the lowest concentration of steranes was in residential areas at S19 (185.24 µg.g _1 dw) and S22 (193.25 µg.g _1 dw). Steranes levels of street dust samples in this study were relatively higher than the ones measured in Singapore (30.18 µg g −1 (Zhang et al. 2017)), Malaysia (50 µg g −1 (Omar et al. 2007)) and Anzali, Iran (250.98 µg g −1 (Azimi et al. 2018)). The comparison of hopanes and steranes DRs from the present study with other cities worldwide is shown in Table S5.
Petroleum biomarkers such as diasteranes, steranes, and hopanes are molecular biomarkers found in crude oils used to monitor petrogenic-related inputs into the environment, including street dust and airborne PM (El Haddad et al. 2009, Rogge et al. 1993b. Hopanes and steranes are released in the most signi cant amounts by gasoline and diesel-fueled vehicles (Alves et al. 2017, Mostafa et al. 2009).
The T s /(T s + T m ) ratio was calculated in this study using Han et al. (2015) 's ndings for vehicle exhaust particles (0.53-0.61) (Han et al. 2015). The homopane index, C 31 αβS/ (C 31 αβS+C 31 αβR), is one of the most common indices (Alves et al. 2017, Omar et al. 2007), those for this study were between 0.4 to 0.77 (Table S3 in supplementary  section), and for the bishomohopane index C 32 [S/(S+R)] ratios were 0.34 to 0.72 (Table S3) (Han et al. 2015) found that the C 32 [S/(S + R)] ratios or each form of coal smoke were lower (<0.44) than those for vehicle exhausts (0.52-0.64). Rogge et al. (1993) (Rogge et al. 1993b) recorded similar ratios for gasoline (0.57) and diesel exhaust (0.59), which are consistent with the ndings of this study. Furthermore, the C 31 -C 35 /C 30 ratio provides additional information for determining the source of airborne particles (Alves et al. 2017, Bahry et al. 2009, Zakaria et al. 2002, Zhang et al. 2017. Except for high-rank (anthracite) coal, C 29 αβ/C 30 αβ ratios >1 were recorded for both residential and industrial coal burning, while vehicle exhausts had values of 0.42-0.59 (Han et al. 2015).
These results, therefore, support that tra c-related emissions are the primary sources of hopanes and steranes. The mixture of lubricating oil from automobiles and street dust and soot particles, and their subsequent re-suspension, according to Bahry et al. (2009), is likely to be a signi cant source of atmospheric hopanes.

Spatial distribution and Source identi cation of PAHs
16 prior PAH compounds introduced by USEPA were present and detected in all samples; statistical analysis of the results is presented in Fig. 2. The mean ΣPAHs content was 945.80 (±SD) µg.kg _1 dw. The highest concentrations were observed in S2, S1, S3, and S4, that the measured concentrations were 1580. 79, 1457.25, 1409.25, and 1389.51 µg.kg _1 dw, respectively. These sampling sites are located in the city's center near the holy shrine, and are surrounded by congested roads (Fig. 1).
One of the key explanations for the high concentrations of PAHs observed at some of the sampling stations is likely due to the limited dispersion of pollutants from their primary source (Wang et al. 2011). S38, for example, is present in a bus terminal and has a relatively high concentration of 1008.24 µg.kg −1 dw, which may be due to engine oil and diesel leakage. On the other hand, S19 had the lowest concentration (464.26 µg.kg _1 dw), even though it is a lightly tra cked residential area with no industrial activities. Maliszewska-Kordybach (1996) classi ed the levels of contamination into four classes according to the PAHs concentrations: (1) not contaminated (<200 µg·kg −1 ); (2) slightly contaminated (200-600 µg·kg −1 ); (3) contaminated (600-1000 µg·kg −1 ), and (4) heavily contaminated (>1000 µg·kg −1 ).
These categories imply that most samples were contaminated (55%) or slightly contaminated (32.5%). As shown in Table S4, results also show that the Σ16PAH concentrations were lower than those found in street dust from Mashhad (2183.5 µg.kg _1 dw, (Najmeddin et al. 2018)), and other Iranian cities, including Bushehr (1116.2 µg.kg _1 dw, (Keshavarzi et al. 2020)), Isfahan (1074.6 µg.kg _1 dw, (Soltani et al. 2015)) and Ahvaz ( (Table S6). It can be concluded that due to the lack of impact of tra c, i.e. the effect of tyre particles, eroded asphalt, lubricating oil leakage and fossil fuels, the concentration of compounds in DD is lower than that of street dust in these cities. However, according to Table S6, the mean concentration of Σ16PAHs was slightly higher or in the same range as some of those obtained from street dusts in tropical and dry areas (Abbasnejad et al. 2019, Hassanien &Abdel-Latif 2008, Rastegari Mehr et al. 2016. As shown in Fig. 3a, PAHs was classi ed into three classes depending on the number of aromatic rings: 2-3, 4, and 5-6, with the percentage of rings being: 4 ring (33.6-44.9%, mean 38.8%), 5-6 rings (27.3-39.1%, mean 34.4%) and 2-3 rings (22.3-31.6%, mean 26.89%) (Fig. 3a). The relatively minor proportion of lower molecular weight PAHs (2-3 rings) such as naphthalene, acenaphthylene, acenaphthene and uorine could have its roots in high sensitivity to oxidation through weathering and high solubility in aqueous environments (Al Ali et al. 2017, Marynowski et al. 2011, Omar et al. 2002. The larger proportion of PAHs with a high molecular weight may be due to their common source in vehicle emissions (Hassanien &Abdel-Latif 2008, Lorenzi et al. 2011, Wang et al. 2011. In DD, combustion PAHs, the sum of Flt, Pyr, BaA, Chr, B(b+k)F, and BaP , Rogge et al. 1993a) constitute a notable contribution of the total, in the range of 986.60-1069.42 µg.kg _1 dw with a mean value of 740.04 µg.kg _1 dw. DRs for PAHs are shown in Fig. 3b that it demonstrated the source for the most of the samples were petroleum combustion section.
According to reports issued in Mashhad's transportation statistics in 2018, the number of passengers who traveled to Mashhad in 2018 was more than 25 million, 32% of which traveled to this city in the summer. As a consequence, summer brings a rise in the number of passengers on buses, trains, and ights. According to the meteorological statistics of Mashhad in 2018, the months of June (25.5 •C), July (30.17 •C) and August (28.7 •C) had the highest temperature and the minimum rainfall was in the three months of summer. In these seasons of the year, a lack of rainfall and high temperatures cause a decrease in soil moisture, resulting in the suspension of soil particles due to wind and a rise in dust and suspended particles (Abbasi &Keshavarzi 2019). In addition, according to Mashhad's annual report on environmental pollutants for 2018, the rst half of the year saw higher levels of PM 10 . Higher concentrations of suspended particles were found in the studied stations during the spring and summer, which is consistent with the ndings of the current study's surface dust.
Two PCs were de ned as re ecting different source categories or physicochemical behavior in the results of PCA analysis, as shown in Fig. S2 and Table 3. 4-5 ring compounds, including Ant, Flt, Pyr, BaA, BbF, BkF, InP, DahA and BghiP. BghiP dominated the rst PC (with 37.20% of the total variance) (Mostafa et al. 2009), (Boonyatumanond et al. 2007, Rogge et al. 1993a, and Flt and Pyr are associated with diesel emissions (Mostafa et al. 2009). DahA which represents fossil fuel combustion (Boonyatumanond et al. 2007, Rogge et al. 1993a with InP speci cally associated with gasoline emissions (Mostafa et al. 2009). In summary, Factor 1 was assigned to vehicular tra c contamination.  (Lin et al. 2010, Park et al. 2002, napthalene emissions with petroleum application due to its direct relevance with incomplete combustion sources (Jiang et al. 2014, Moeinaddini et al. 2014a) and Chr emissions with natural gas combustion (Harrison et al. 1996).
Factor 2 represents wood, biomass, and incomplete combustion-related sources, as well as natural gas combustion, based on the high loading values of these PAHs (Fang et al. 2004).

Risk assessment
Using the DD, The compounds that contributed the most to carcinogenic potency, as shown in Table 4, were BaP and DahA, which accounted for 24.73% and 47.91% of the total BaP TEQ , respectively. In most monitoring systems, an ILCR of 10 −6 -10 −4 indicates a potential risk, as the ILCR < 10 −6 implies relative safety and the ILCR > 10 −4 illustrates high risk (Gope et al. 2018, Guo et al. 2003. The carcinogenic risk from ingestion and dermal touch was around 10 −7 for adults and infants, while the risk from inhalation was between 10 −11 to 10 −12 . In a comparison of exposure methods [ingestion and dermal contact (Wu et al. 2018)], dermal contact was identi ed as the main exposure pathway to cancer risk, the latter exposure risk was almost negligible.  (Maertens et al. 2008) estimated higher surplus lifetime cancer risks for preschool children, ranging from 10 −6 and 10 −4 . So, in the currrent study, the total carcinogenic risk for children (8.45 x 10 −12 ) indicated no risk, while for adults (9.80 x 10 −7 ) was the nearest to the initial admissible risk of 10 −6 . Despite the relatively low risks involved, if the in ux of visitors grows without any control over tra c movements, and with DRs suggesting sources of PAHs from vehicular tra c, these risks are likely to increase unless urgent action is taken.

Relationships between variables
The aim of cluster analysis was to discover the relationship between variables (stations, n-alkanes, PAHs, hopanes, steranes and some diagnostic ratios). Standardization was done prior to clustering, due to the distinctive scales of the variables. Fig. 4 shows the DHCA results.
These clusters are identi ed by inspecting the correlation matrix of the parameters, and the parameters generally in each cluster are strongly correlated. The colors describe the relationship between the parameters. The vertical dendogram illustrates how sampling locations are clustered based on different variables. The horizontal dendogram shows how the various variables are grouped together based on their similarities (Fig. 4). HCA separates the sampling sites into two clusters in the horizontal dendogram. The ndings suggest that the distribution of ∑16PAHs, ∑20n-alkanes, ∑8hopanes and ∑6steranes varies by sampling site in the city, with the colors implying the relationship between parameters and sampling stations. In terms of vertical dendrogram, DHCA groups sampling points into 2 clusters. Cluster 1 is made up of high-tra c sampling points and sampling points near highways. Cluster 2 is composed of sampling points in close proximity to residential areas.
According to the results of DHCA and the impact of urban tra c on the clustering of sampling points, it is fair to believe that the aromatic and aliphatic hydrocarbon emissions in this area could have an adverse effect on tourism, socioeconomic conditions, and air quality in this city.

Conclusions
In DD from Mashhad, Iran, high levels of aromatic and aliphatic hydrocarbons (PAHs, n-alkanes, hopane and steranes) were found near the heavily tra cked city center and bus terminal. The majority of n-alkanes in urban DD originate in vehicular emissions, with minor contribution from higher plant waxes. The low values of CPI (CPI ~ 1) and the high values of U:R support this nding. The presence of pristane, phytane, hopanes, and steranes in DD, even the UCM indicates that it has been polluted by petroleum residues originated in vehicular emissions. The molecular distribution of Hopanes and steranes is very similar to that of vehicular exhaust. The urban historical structure in Mashhad also affected the occurrence and distribution of PAHs in DD. In contrast to residential areas, the oldest urban areas (near the holy shrine) have more polluted dust, with around three times the concentrations of some PAHs. The results of DRs and PCA analysis show that the major compounds contained in the DD were high molecular weight PAHs, implying a combustion origin.
Based on the results of this study, diesel and gasoline-powered vehicles are responsible for a signi cant proportion of the hydrocarbon pollution in Mashhad; technological monitoring and improvements in the e ciency of diesel and gasoline vehicles will help to minimize this emissions. The ndings of this study are applicable to similar urban areas around the world, principally those with known heritage sites or tourist in uxes. It also operates as a scienti c foundation for relevant authorities and policy-makers to formulate and implement policy and air pollution mitigation measures to regulate and monitor tra c-related ne aerosol emissions, protecting the environment and public health. At last, it should be noted that total cancer risk for both children (8.45 ×10 −12 ) and adults (9.80×10 −7 ) in this research is in the range of virtual safety, making the risk of cancer for teens and adults who studied and worked in these buildings insigni cant.