Occurrence of Polycyclic Aromatic Hydrocarbons and Polychlorinated Biphenyls in Fogwater at Urban, Suburban, and Rural Sites in Northeast France between 2015 and 2021

Abstract

Polycyclic aromatic hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs) exist in the atmosphere in the vapor and particulate phases, as well as in solubilized form in fog/rain/cloud waters. In the current paper, fogwater samples are collected during 42 events between 2015 and 2021 at four different sites (Strasbourg, Geispolsheim, Erstein, Cronenbourg) in the Alsace region. Organics are extracted using liquid–liquid extraction (LLE) supported on a solid cartridge (XTR Chromabond), and then analyzed by gas chromatography–tandem mass spectrometry (GC-MS/MS). The total PAHs and PCBs concentrations in fog samples vary between 0.58 and 6.7 µg L⁻¹ (average of 2.70 µg L⁻¹), and 0.14 and 15.5 µg L⁻¹ (average of 2.75 µg L⁻¹). Low-molecular-weight (LMW) PAHs are predominant and highly detectable compared to high-molecular-weight (HMW) PAHs, while pentachloro-biphenyls are the dominant PCB congener. The PAHs and PCBs concentrations have increased over the sampling years at all sites, except for a slight decrease in PCBs level at Geispolsheim. A diagnostic ratio (DR) and principal component analysis (PCA) are applied to suggest potential contamination sources at Strasbourg metropolitan. Their results reveal that PAHs derive from a mixture of pyrogenic activities, while PCBs mainly come from industrial activities. The results also reveal, in some cases, inter-site variability for simultaneous and successive fog events.

1. Introduction

Fogwater is a meteorological phenomenon that is composed of water droplets condensed in the air [1,2]. Fog has the ability to scavenge inorganic compounds as well as organic compounds through various mechanisms, including nucleation scavenging, coagulation, Brownian diffusion, and uptake of precursor gases with subsequent transformation [3,4,5,6]. The main environmental benefit of fogwater is its strong ability to decrease the ambient pollutant concentration through continuous nucleation scavenging followed by wet deposition [7]. However, it may promote the formation of new particles via aqueous-phase reactions [8,9,10]. Overall, fog acts like a reservoir that accumulates the scavenged pollutants present in the air, water-soluble gases, and their chemical reaction species, and therefore, fog is a rich environment for studying air pollutants rather than ambient atmosphere [11]. Despite the complexity of fog analysis and sampling, it is still a unique and interesting topic to be investigated by researchers, which has rapidly grown over time.

Persistent organic pollutants (POPs), such as PAHs and PCBs, are amongst the organic species that can be scavenged by fogwater [12,13]. POPs are highly stable and long-lasting and may persist in the atmosphere for extended periods of time compared to other toxins [14]. They are characterized by their high potential for degradation and their ability to be transported through different environmental (air, water, and soil) and biological matrices. They can enter the environment through unintentional and intentional routes resulting from diesel, fossil fuel combustion, agricultural activities, etc. [15]. POPs are linked with adverse impacts on humans and the environment. Long-term exposure to some POPs has been associated with developmental abnormalities, endocrine disruption, and increased cancer risks [14,16]. Numerous studies have been performed on PAHs and PCBs occurrence and distribution in the aquatic system (river, lakes, etc.) [17,18,19], soil [20,21,22], and air [23,24,25,26], of which very few studies were performed in fogwater [2,27,28,29]. Although there are many PAHs, most analyses and regulations focus on a limited number of PAHs, which are recognized by the Environmental Protection Agency (EPA) as harmful components (16 PAHs) [30]. Once released into the atmosphere, they can be found in the gas or particulate phase. PAHs with two to three rings (LMW) tend to be more concentrated in the gas phase, whereas those with four to six rings (HMW) are found in the particulate phase [31]. PCBs have been widely used for a long time in a wide variety of industrial applications, such as electrical transformers and capacitors, hydraulic fluids, lubricants, paints, heat-transfer fluids, etc. PCBs are classified based on the degree of chlorination (low and high chlorinated PCBs). Their degree of toxicity increases with their number of chlorine atoms [32]. Therefore, it is crucial to investigate the occurrence of both PAHs and PCBs in fog events.

The Alsace region is known for its radiation fog that is formed during stable weather conditions (calm wind) at nights. Strasbourg is known for its high population density, which contributes to harmful contamination in the city itself and in its surrounding regions due to the release of various pollutants. To better understand the environmental and health aspects of PAHs and PCBs, 38 fog events are sampled at four different sites, which are Strasbourg (urban), Geispolsheim (suburban), Erstein (rural), and Cronenbourg (suburban) during 2015, 2016, and 2018. Additionally, only one fog event is sampled during 2017 at Geispolsheim, while three are sampled during 2021 at Cronenbourg due to the absence of fog events at other sites. The inorganic analysis of fogwater at Strasbourg started in the early 1990s and lasted until the end of 1999 [33,34]. Since then, no fog studies have been conducted in Alsace for either organics or inorganics. That is why this paper is novel, because it is the first to investigate organic contaminations in fog in Alsace, and even in France. To this end, organic matters are extracted using LLE on a solid support (XTR Chromabond) coupled with GC-MS/MS to assess their concentration levels in fog samples. The global aim of this study is to create a primary database regarding fogwater contamination levels and sources in Alsace. The specific objectives of this research include (1) monitoring the occurrences of PAHs and PCBs in fogwater at Strasbourg metropolitan (Alsace region) (2) checking the evolution of fogwater over the different sampling years (3) identifying the possible sources of PCBs and PAHs in Alsace. Besides, this research investigates the simultaneous and successive occurrences of fog events at different sites. To my knowledge, there are no previous field fog studies that have compared the occurrence of such events. However, the current paper deals only with some specific organic pollutants (PAHs and PCBs); thus, it is highly recommended to perform a global statistical analysis based on the whole dataset, including inorganics and organics, along with fog microphysics to study the factors that might be responsible for spatiotemporal variation (inter/intra date and site).

2. Materials and Methods

2.1. Study Sites

The fogwater samples assessed in the current study were sampled from northeastern France between October and December between 2015 and 2018 and in 2021. The sites were chosen based on their topological variability. Strasbourg (48.58461°, 7.75071°), which is located in the middle of the study area, is a typical urban site surrounded by huge residential areas (284.677 inhabitants) and traffic. It is in close proximity (3 Km away) with the biggest industrial zone in the region, known as the “Port du Rhin”. We also studied two suburban sites, Cronenbourg (48.59449°, 7.71621°) and Geispolsheim (48.51469°, 7.64373°), which are located, respectively in northwest and west Strasbourg. Geispolsheim is enclosed by small houses (7000 inhabitants) and agricultural fields, and is located around 1 Km from the A35 highway and 9 Km from Entzeim airport. Cronenbourg is surrounded by bigger residential areas (21,000 inhabitants) and limited industries. We also studied Erstein (48.42345°, 7.66326°), a rural site (11,000 inhabitants) situated in southwest Strasbourg characterized by its wider agricultural fields. The sampling sites are shown in Figure 1.

2.2. Sampling Campaign

The sampling campaign was mainly performed between 2015 and 2018 during October, November, and December (foggy periods in Alsace) to obtain a total of 39 samples. Additionally, three samples were gathered from Cronenbourg alone during 2021 due to the absence of foggy days. At each sampling site, the fog samples were collected using a Caltech Active Strand Cloudwater collector (CASCC2) described in detail elsewhere [35]. Field blanks were taken continuously after cleaning the collector to ensure the absence of any contamination during the sampling period. More details about the collection procedure are found in the work of Khoury et al. [12,36]. Table S1 in the Supplementary Materials shows the sampling information of the different samples.

2.3. Analytical Procedure of Fogwater Samples

Samples are analyzed for their potential contamination in 22 PCBs and 16 PAHs analyzed by GC-MS/MS. The analytical procedure used for the analysis is based on the work of Khoury et al. for the residual analysis of organic compounds in fogwater [12]. The validation parameters for PAHs and PCBs along with the internal standard mixture are shown in Tables S2–S6 in the Supplementary Materials.

2.3.1. Samples Treatment

The fog samples were filtered on-site (glass microfiber, 0.47 µm porosity), transported in refrigerated conditions to the ICPEES laboratory, and separated into many aliquots. The first aliquot was used to measure pH (pH meter), conductivity (conductivity meter), and dissolved organic carbon (total organic carbon analyzer); another aliquot was used to quantify cations and anions (ion chromatography); a third aliquot was used to measure trace metals (inductively coupled plasma mass spectrometry); and a last aliquot was used to analyze different organic families. The fog samples were maintained in clean glass bottles and frozen at −18 °C, then stored until further analysis. In the current paper, we present only the results of the determination of PAHs and PCBs for the different fog events in the studied areas. The full handling protocol is found in the work of Khoury et al. [12,36].

2.3.2. Extraction Procedure

The developed extraction protocol for the quantification of the organic fraction in fogwater, along with its optimization and validation parameters, are fully described by Khoury et al. [12]. LLE on a solid cartridge (XLB Chromabond) was used for extraction of PAHs and PCBs in fog samples. A volume of 50 mL of fog sample was loaded into the column and the elution was performed successively with 40 mL of dichloromethane (DCM) and 40 mL of ethyl acetate (EtAc). The extract (around 80 mL) was gently evaporated under a fume hood to 1 mL. If the collected fog volume was large enough (>50 mL), two extractions were performed separately to extract as much organic material as possible. The chromatographic analysis was performed by GC-MS/MS following the addition of the appropriate mixture of the internal standard (IS). Laboratory blanks were also performed continuously between samples injection to remove any trace in the analytical instrument (column, etc.) and to ensure the absence of contamination.

2.3.3. Chromatographic Analysis

The analytical parameters for the analysis of PAHs and PCBs are summarized in

Table 1. Analytical conditions for the analysis of PAHs and PCBs with GC-MS/MS.

Chromatographic Conditions
Device	GC-MS/MS (TraceTM, ITQTM 700)
Separation column	XLB (50% phenyl/50% methylsiloxane) (30 m length, 0.25 mm diameter, 0.25 μm film thickness)
Injection parameters
DCM Rinsing	2 Rinsing with 1 µL (pre-run and post-run)
Injection volume	1 µL
Injection type	Splitless mode
Injector temperature	250 °C
Purge	50 mL.min−1 after t = 2 min
Gas saver	15 mL.min−1 after t = 5 min
Chromatographic parameters
Carrier gas	Helium (purity > 99.99%)
Carrier gas flow	Constant at 1 mL.min−1
Pressure	≈10.253 psi (à t = 0 et T = 90 °C)
Oven temperature programming
Mass spectrometer parameters
Transfer line temperature	300 °C
Electron energy	70 eV
Source temperature	210 °C
Acquisition mode	MRM

. The identification of the different species is based on the multiple reaction monitoring (MRM) mode. The accurate detection of each compound detected is achieved by checking and comparing the observed retention times, parent ions, and fragmented (daughter) ions with those obtained for standard solutions. The results are treated using linear calibration curves which are performed on Xcalibur.

2.4. PAHs Diagnostic Ratios

In this investigation, PAHs are divided into LMW and HMW, according to the number of benzene rings. LMW PAHs are composed of less than four aromatic benzene rings including Naphtalene (NaP), Acenaphtene (Ace), Fluorene (Flu), phenanthrene (Phe), and anthracene (Ant). HMW PAHs are composed of more than four aromatic benzene rings including Fluoranthene (Flo), Pyrene (Pyr), Chrysene (Chry), Benzo(a)anthracene (BaA), Benzo(b)fluoranthene (BbF), Benzo(k)fluoranthene (BkF), Benzo(e)pyrene (BeP), Benzo(a)pyrene (BaP), Dibenzo(a,h)anthracene (DBhA), Indenol(1,2,3)pyrene (IndP), and Benzo(g,h,i)perylene (BghiP).

PAHs are considered to be single hoppers, suggesting that all sites might be affected by the emissions released from neighboring industries and power plants. In the current study, qualitative source identification of PAHs was conducted by using the diagnostic ratio (DR) method, which has been widely employed [37,38]. It is used to differentiate between gasoline and diesel combustion emissions and between biomass burning processes and different crude oil processing products [39]. Subsequently, the DR method is performed through establishing an index based on the ratio of the PAHs having the same molecular weight (MW) [37]. A LMW/HMW ratio of less than 1 implies combustion sources, whereas a ratio higher than 1 indicates petroleum sources [40]. PAHs with molecular weights of 178 and 202 are commonly used to differentiate between combustion and petroleum source. For a PAH with a molecular weight of 178, a ratio of Ant/(Ant + Phe) lower than 0.1 indicates petroleum origins, while a ratio of higher than 0.1 indicates combustion emissions [41]. For an individual PAH compound with a molecular weight of 202, the Flo/(Flo + Pyr) ratio if less than 0.4 indicates petroleum source or oil spills. A ratio between 0.4 and 0.5 indicates fossil fuel combustion (vehicle and crude oil), whereas the source is suggestive of grass, wood, or coal combustion when the ratio is beyond 0.5. Besides, when the ratio of Flu/(Flu + Pyr) is lower than 0.5, it indicates that PAHs originate from petroleum emissions, whereas a ratio higher than 0.5 indicates diesel emissions [37,42]. Other DRs such as IndP/(IndP + B(a)P) and B(a)A/(B(a)A + Chry), are commonly used in previous studies [37,40,43,44,45], but cannot be applied in our case because their DFs are very low in our samples.

2.5. PCBs Risk Assessment

Risk assessment has a particular focus on dioxin-like (DL) PCB congeners due to their high toxicity. The quantification of their potential risk is performed by calculating the total toxic equivalent quantity (TEQ). TEQ is defined as the sum of the product of each PCB multiplied by its toxic equivalency factor (TEF) (Equation (1)) [46]. It is an estimate of 2,3,7,8-tetrachlorodibenzo-p-dioxin-like activity (2378-TCDD) and is frequently used for legislation and risk assessment and management.

$$\mathrm{T}\mathrm{E}\mathrm{Q}=\sum \mathrm{i} \mathrm{C}\mathrm{i} \mathrm{\ast } \mathrm{T}\mathrm{E}\mathrm{F}$$

(1)

where $\mathrm{C}\mathrm{i}$ is the individual PCB-DL concentration (ng L⁻¹) and $\mathrm{T}\mathrm{E}\mathrm{F}$ is the toxic equivalency factors for the six PCBs which are, respectively, 0.0001, 0.0003, 0.00003 for PCB 77, PCB 81, and PCBs 105, 114, 118, and 167, according to standards set by the World Health Organization (WHO) in 2005 [47].

2.6. Principal Component Analysis

In this study, principal component analysis (PCA) is executed using R (version 4.3.2) correlated with RStudio (version 2023.12.0 + 369). PCA can semi-quantitively describe the main PAHs and PCBs source contribution [48,49,50,51]. The concentrations of PAHs and PCBs are standardized before executing PCA to avoid the dominance of very high PAH or PCB levels compared to very low ones. Principal components (PCs) with eigen values higher than 1 are retained, and a correlation is considered significant when the loading of the variable is greater than 0.5.

3. Results and Discussion

3.1. PAHs Analysis in Fogwater

The common PAHs detected in most analyzed fog samples mostly belong to the LMW PAHs (2–3 rings) like NaP, Flu, Phe, and Ant. HMW PAHs (4–6 rings) are rarely found in fog water. LMW PAHs tend to be more concentrated in the gas phase and could be scavenged by fog water due to their higher solubility, while those with a higher molecular weight are often associated with particulates due to their higher hydrophobicity. For instance, BaA, Chry, BbF, BkF, BeP, BaP, and DBhA are monitored once or twice per site. BaP is regarded as a marker of total carcinogenic PAH compound; fortunately, it was not detected much in this investigation. The detection frequency (DF) for LMW PAHs is higher than 96%, whereas that of HMW PAHs varies from 0 to 96%. The DF of Flo, Pyr, IndP, and BghiP is, respectively, 96%, 91%, 77%, and 56%, whereas the rest is detected with less than 30%. Over the sampling years, the concentrations of LMW PAHs are several times higher than those of HMW PAHs at all sampling sites.

The mean concentrations of LMW and HMW PAHs at the four sites during the sampling years are summarized in Figure 2 (see Table S7 in the Supplementary Materials). LMW PAHs account for between 67 and 85% (average of 78%), 61 and 89% (average of 77%), 82 and 95% (average of 88%), and 81 and 88% (average of 83%) of the total PAH concentrations, respectively at Geispolsheim, Erstein, Strasbourg, and Cronenbourg. Their mean concentrations, respectively vary from 0.17 to 1.15 µg L⁻¹, from 0.05 to 0.51 µg L⁻¹, from 0.31 to 0.87 µg L⁻¹, and from 0.03 to 1.58 µg L⁻¹. The total PAH concentrations vary according to the sampling site and year. The highest total mean concentration was obtained at Erstein (2.99 ± 0.25 µg L⁻¹) followed by Geispolsheim (2.91 ± 0.34 µg L⁻¹), Cronenbourg (2.77 ± 1.70 µg L⁻¹) and Strasbourg (2.40 ± 0.93 µg L⁻¹). A slight increase in the total mean concentration was observed yearly at all sampling sites. For instance, the total mean concentrations increased at Geispolsheim from 2.53 to 3.34 µg L⁻¹ (+33%) between 2015 and 2018, at Erstein from 2.71 to 3.21 µg L⁻¹ (+18%) between 2015 and 2018, at Strasbourg from 1.75 to 3.06 µg L⁻¹ (+74%) between 2016 and 2018, and at Cronenbourg from 1.57 to 3.97 µg L⁻¹ (+153%) between 2018 and 2021. The high increase at Cronenbourg might be due to the post-COVID pandemic period. France experienced a total confinement period during COVID-19 in which people stayed at home, suggesting that PAH concentrations are closely associated with additional anthropogenic activities such as more wood and coal burning for domestic heating.

The contributions of PAHs, shown in Figure 3, correspond to the sum of the average concentration for each PAH compound at each site to the total PAH concentrations at that site during all years. Nap, Phe, and Ant have the highest contributions in all analyzed fog water samples at all locations. Ant has the highest contribution which varies between 25.3 and 38.9% (average of 35.2%), followed by Phen, which varies between 19.1 and 30.1% (average of 24.5%), and Nap varies between 10.5 and 20.2% (average of 14.9%). Their average concentrations are, respectively, 2.68, 1.79, and 1.21 µg L⁻¹. The % contributions of other PAHs are less than 10%. HMW PAHs have low contributions in fog water samples (as low as 1%), except for Pyr and Flo, whose average % contributions are, respectively, 6.9 and 4.9%, with average concentrations, respectively, of 0.54 and 0.39 µg L⁻¹.

3.2. PCBs Analysis in Fogwater

Figure 4 illustrates the total PCB concentrations at the investigated sites during all years (see Table S8 in the Supplementary Materials). The highest total PCB concentration was obtained at the urban site Strasbourg, followed by Geispolsheim, Cronenbourg, and Erstein. Their average concentrations were, respectively, 8.99 ± 2.69 µg L⁻¹, 4.24 ± 2.81 µg L⁻¹, 3.59 ± 1.08 µg L⁻¹, and 2.38 ± 1.90 µg L⁻¹. The total PCB concentrations increased at Strasbourg, Cronenbourg, and Erstein, respectively by 53% (between 2016 and 2018), 54% (between 2018 and 2021), and 188% (between 2015 and 2018). However, a slight (13%) decrease occurred at Geispolsheim (between 2015 and 2018), despite the substantial decrease between 2015 and 2017 (−85%). It seems that secondary emissions are, in the long term, more important than primary emissions.

The average concentrations of the different PCB congeners at all sites are illustrated in Figure 5. In the current analysis, PCBs are classified into different PCB congeners: trichlorobiphenyls, including PCBs 18, 28, 31; tetrachlorobiphenyls, including PCBs 52, 70, 81; pentachlorobiphenyls, including PCBs 101, 105, 114, 118, 123, 126; hexachlorobiphenyls, including PCBs 138, 149, 153, 156, 157, 169; and heptachlorobiphenyls, including PCBs 189. The repartition of the different PCB congeners is similar at all sampling locations. The figure also shows that pentachlorobiphenyls are the dominant congener at all sites, accounting for 69, 60, 64, and 51%, respectively at Geispolsheim, Erstein, Strasbourg, and Cronenbourg. Their mean concentrations are, respectively, 2.94, 1.37, 5.75, and 1.83 µg L⁻¹. Hexachlorobiphenyls are the second dominant congeners, accounting for 17, 25, 17, and 31% of the total PCB fraction, respectively, at Geispolsheim, Erstein, Strasbourg, and Cronenbourg. Their average concentrations are, respectively, 0.70, 0.57, 1.55, and 1.08 µg L⁻¹. Tetrachlorobiphenyl congeners come third, and they contribute between 7 and 13% of the total PCB fraction. After that come heptachlorobiphenyls, whose contributions comprise less than 10%, followed by trichlorobiphenyls, which have the fewest contributions (between 1 and 4%) and lowest concentrations.

The identification of PCB sources in the environment is more complex than the identification of PAHs, since there are no DRs specific for PCB congeners found in the literature. Even though their use and production are banned worldwide, they continue to be found at important concentrations in the atmosphere resulting from the deposition released from diesel generators, electric equipment, and vehicle exhaust. It should be common knowledge that some equipment and materials containing PCBs, such as capacitors, vessel paints, double framed glazing windows, etc., are still in use or in stock in some regions [52]. Moreover, in high-density population areas, the high levels of PCBs are not only related to the volatilization process from soil and equipment, but also to the thermal processes that mainly occur during industrial processes, waste incineration, vehicle exhaust, and combustion of organic matter. Since the distribution of PCB homologs is approximately the same at all sites, their emission sources could be assumed to be almost the same at the four sampling locations. PCBs 118 and 138 are detected at important concentrations at all sites, proving the influence of vehicles and electric equipment as primary PCB sources. Their total mean concentrations are the highest at the urban site (Strasbourg), followed by the two suburban sites (Geispolsheim and Cronenbourg), and the rural site (Erstein). In 2018, PCBs 118 and 138 together have the highest total concentrations (0.98 ± 0.10 µg L⁻¹) at Strasbourg, followed by Geispolsheim (0.22 ± 0.20 µg L⁻¹), Cronenbourg (0.15 ± 0.05 µg L⁻¹), and Erstein (0.06 ± 0.03 µg L⁻¹). The same decreasing order is also observed for 2016, with Strasbourg having the highest total concentrations of PCBs 118 and 138 (0.67 µg L⁻¹), followed by Geispolsheim (0.48 ± 0.61 µg L⁻¹), and Erstein (0.15 ± 0.09 µg L⁻¹). This can be associated with the fact that Strasbourg is the nearest site among others to the industrial port in Kehl “Port du Rhin”. Besides, a high load of traffic and vehicles (50,000 vehicles/day) crosses the highway, which is only 2 Km away from the sampling point. These two reasons demonstrate the high fraction of PCBs at Strasbourg. The high concentrations of PCBs at Geispolsheim could be also due to its proximity to the highway (A35), along which a huge number of vehicles and trucks passes daily. Other possible sources might be due to the presence of small aluminum and steel factories, as well as some energy industries near this region. Moreover, the presence of Entzeim airport may release PCBs due to the incomplete combustion of PCB impurities in fuel, airplane engines, and electric generators. The PCB contamination at Cronenbourg is primarily the result of some traces of PCBs in the soil resulting from one of the biggest breweries “Kronenbourg” in France and Grandest, even though it closed in 2000. This proves the secondary emission sources of PCBs in the air (soil accumulation). At Erstein, PCBs could result from the long-range transport (LRT) of PCBs, in particular, the low chlorine levels congeners (<5 atoms) that are easily degraded in the atmosphere. Another source could also be attributed to the dependency of rural people on coal and wood for residential heating.

The DF of the analyzed PCBs varies between 27 and 100%. PCBs 105 and 108 are detected in all samples (100%), followed by PCB 189 (98%), PCB 157 (97%), PCB 126 (84%), PCB 70 (72%), and PCB 52 (67%). The least-detected PCBs are PCBs 18 and 157, whose DF is 27%. Among the detected PCBs, seven of them are known as the PCBs indicator (∑₇PCBs) due to their high abundance in the atmosphere, whereas twelve of them are known as the PCBs dioxin-like (DL) (∑₁₂PCBs) due to their high toxicity and persistency in the atmosphere and their detrimental health effects. The seven PCBs indicator are 28, 52, 101, 118, 138, 153, and 180, of which six are detected almost in all samples. The 12 PCBs DL are 77, 81, 105, 114, 118, 123, 126, 156, 157, 167, 169, and 189, of which nine are highly detectable in this study. The ∑₇PCBs at Strasbourg, Geispolsheim, Cronenbourg, and Erstein are found, respectively in the range of 1.25–2.38 µg L⁻¹ (average of 1.81 µg L⁻¹), 0.26–2.09 µg L⁻¹ (average of 0.89 µg L⁻¹), 0.24–0.46 µg L⁻¹ (average of 0.35 µg L⁻¹), and 0.15–0.54 µg L⁻¹ (average of 0.33 µg L⁻¹), whereas the ∑₁₂PCBs vary, respectively in the range of 5.14–7.28 µg L⁻¹ (average of 6.21 µg L⁻¹), 0.82–2.29 µg L⁻¹ (average of 1.69 µg L⁻¹), 1.34–2.39 µg L⁻¹ (average of 1.86 µg L⁻¹), and 0.53–1.40 µg L⁻¹ (average of 0.99 µg L⁻¹). The highest average of both groups is obtained at Strasbourg, followed by Geispolsheim, Cronenbourg, and Erstein. ∑₁₂PCBs dominates over ∑₇PCBs at all sampling sites.

DL-PCBs are classified as non-ortho PCBs with high likelihood of producing DL-effects, mono-ortho substituted PCBs with a weak ability to produce DL-effects, and multiple-ortho substituted PCBs with no DL-effects [53]. In the following investigation, six PCBs, PCB 77, PCB 81, PCB 105, PCB 114, PCB 118, and PCB 167, are of major concern due to their similar toxic effect to 2378-TCDD. The decreasing order for ∑₆PCBs is as follows: Strasbourg > Geispolsheim > Cronenbourg > Erstein. The total average concentration at Strasbourg is 2.67 µg L⁻¹, Geispolsheim is 0.94 µg L⁻¹, Cronenbourg 0.46 µg L⁻¹, and Erstein is 0.32 µg L⁻¹. The highest ∑₆PCBs is obtained at Strasbourg, accounting for 30% of the total average PCBs and 43% of ∑₁₂PCBs. The fewest ∑₆PCBs are obtained at Erstein, accounting for 13% of the total PCBs and 25% of the ∑₁₂PCBs. Thus, these pollutants should be considered a strong cause for concern during pollution management. The average TEQ is the highest at Strasbourg (0.119 ng WHO-TEQ. L⁻¹) due to the high-density population and its location near to the industrial zone, followed by Geispolsheim (0.039 ng WHO-TEQ. L⁻¹), and Cronenbourg and Erstein (0.022 ng WHO-TEQ. L⁻¹). Thereby, more attention and concern should be drawn to the urban site at Strasbourg, since people there are more exposed to the negative risks of PCBs.

3.3. Source Analysis

3.3.1. Diagnostic Ratio

The ratios of HMW/LMW PAHs are all lower than 1 (varying from 0.1 to 0.5), suggesting that pyrogenic activities (petroleum sources) predominate over petrogenic activities. These results are in accordance with previous fog studies [13,27]. Pyrogenic sources might include the incomplete combustion of organic matters (fossil fuel, coal, wood, and petroleum), forest fires, by-products of industrial processing and vehicle engines powered by gasoline or diesel fuel [54]. The average ratios of Ant/(Ant + Phe), Flu/(Flu + Pyr), and Flo/(Flo + Pyr) at the four sites during the sampling years are shown in

Table 2. Average diagnostic ratios of PAHs at all sites and years.

Year	Ant/(Ant + Phe)
	Geispolsheim	Erstein	Strasbourg	Cronenbourg
2015	0.50 ± 0.2	0.40
2016	0.56 ± 0.18	0.46 ± 0.15	0.46
2017	0.72
2018	0.65 ± 0.1	0.65 ± 0.13	0.60 ± 0.12	0.56 ± 0.28
2021				0.52 ± 0.12
Average	0.6 ± 0.09	0.51 ± 0.13	0.53 ± 0.09	0.54 ± 0.02
	Flu/(Flu + Pyr)
	Geispolsheim	Erstein	Strasbourg	Cronenbourg
2015	0.41 ± 0.18	0.63
2016	0.73 ± 0.22	0.47 ± 0.28	0.51
2017	0.36
2018	0.74 ± 0.10	0.72 ± 0.13	0.86 ± 0.06	0.60 ± 0.02
2021				0.42 ± 0.15
Average	0.56 ± 0.2	0.60 ± 0.12	0.68 ± 0.24	0.52 ± 0.12
	Flo/(Flo + Pyr)
	Geispolsheim	Erstein	Strasbourg	Cronenbourg
2015	0.58 ± 0.1	0.38
2016	0.52 ± 0.13	0.49 ± 0.21	0.39
2017	0.42
2018	0.55 ± 0.12	0.36 ± 0.11	0.5 ± 0.11	0.36 ± 0.18
2021				0.48 ± 0.12
Average	0.52 ± 0.07	0.41 ± 0.07	0.45 ± 0.07	0.42 ± 0.08

. The results reveal that the ratios of Ant/(Ant + Phe) are all higher than 0.1, proving the presence of pyrogenic emissions. For instance, the average ratios at Geispolsheim, Erstein, Strasbourg, and Cronenbourg are, respectively, 0.61 ± 0.09, 0.51 ± 0.13, 0.54 ± 0.09, and 0.54 ± 0.02. In addition, most of the Flu/(Flu + Pyr) ratios are higher than 0.5, suggesting that diesel emissions are one of the main sources in the region. Diesel emissions are mainly released from the transportation sector (automobile traffic with diesel vehicles and heavy-duty trucks) [55]. For instance, the average ratios at Geispolsheim, Erstein, Strasbourg, and Cronenbourg are, respectively, 0.56 ± 0.2, 0.6 ± 0.12, 0.68 ± 0.24, and 0.52 ± 0.12. Furthermore, the results show that the average ratios of Flo/(Flo + Pyr) at Erstein, Strasbourg, and Cronenbourg are, respectively, 0.41 ± 0.07, 0.45 ± 0.07, and 0.42 ± 0.08. These are all between 0.4 and 0.5, suggesting that PAHs also originate from fossil fuel combustion. Geispolsheim is the only site at which most of the average values of Flo/(Flo + Pyr) are higher than 0.5 (0.52 ± 0.07), indicating that PAHs are released from the combustion of grass, wood, and coal. In the Alsace region, more than 87% of PAHs originate either from domestic heating (based on wood or coal) during the winter time (fog period) or from the transportation sector (diesel and gasoline combustion). These data are in accordance with the results obtained in this work [56].

3.3.2. Principal Analysis

The determination of emission sources for PAHs in fogwater is further assessed based on the PCA. The significant loading factors (>0.5) obtained via the varimax rotation for PAHs are summarized in

Table 3. Factor loadings of PAHs after PCA varimax rotation in the whole sampling campaign.

Variables	Factor 1	Factor 2	Factor 3	Factor 4
Flu			0.51
Phe		0.61
Flo		0.81
Pyr		0.74
BaA	0.69
BbF				0.62
BkF	0.51			0.57
BeP	0.61		0.64
BaP	0.57		0.59
Eigen values	3.14	2.00	1.89	1.55
Variance (%)	22.50	14.35	13.47	11.08
Cumulative (%)	22.50	36.85	50.31	61.39

. The PCA analysis arranged the PAHs dataset into four PCs which control 61.39% of the total data. The table shows that Factor 1 is highly loaded by HMW PAHs such as BaA, BeP, BaP, BkF, and Chry, suggesting the dominance of diesel and gas engine emissions. Chry is a marker of coal combustion, coke production, and wood combustion; BaP is a tracer of vehicle, gasoline, and coal emissions; and BkF is a tracer of diesel-powered vehicle emissions [57,58,59,60,61,62,63,64,65]. Factor 2 is highly loaded with Flo, Pyr, and Phe. The presence of these compounds suggests that coal combustion is their main source [28,62]. Flo and Pyr are considered to be markers of natural gas and biomass emissions (coal combustion, wood burning, etc.), and Phe can be also a tracer of fossil fuel combustion and traffic emission [59,61,66]. Factor 3 is highly loaded BeP, BaP, and Flu. The first two are tracers of natural gas and vehicle emissions, while the third one can be further released from fossil fuel combustion. Factor 4 is highly loaded with BkF and BbF. The latter is emitted from diesel and heavy oil combustion, while the former has been previously reported as a tracer of diesel vehicles [62,65,67]. Based on factor analysis, PAH are primarily released into the Alsace atmosphere either from wood/coal combustion or vehicle emissions (diesel/gasoline combustion).

Then, PCA is performed on the entire dataset in order to evaluate any potential relationships among the different pollutants and to suggest potential PCBs emission sources. The significant loading factors (>0.5) obtained via the varimax rotation for PAHs, together with PCBs, are summarized in

Table 4. Factor loadings of PCBs and PAHs after PCA varimax rotation in the whole sampling campaign.

Variables	Factor 1	Factor 2	Factor 3	Factor 4
PCB18	0.68
PCB28			0.71
PCB31	0.82
PCB52	0.94
PCB70	0.93
PCB81	0.78
PCB105	0.66
PCB114	0.83
PCB118				−0.61
PCB123	0.84
PCB138	0.61
PCB149	0.86
PCB153	0.81
PCB157	0.60
PCB189	0.75
Flu				0.58
Phe				0.59
BaA		0.70
BeP		0.53
Eigen values	9.22	3.39	2.43	2.33
Variance (%)	28.84	10.60	7.61	7.30
Cumulative (%)	28.84	39.44	47.05	54.36

. Factor 1 explaining 28.84% of the total variance has high loading of PCBs (DL and indicators), but no correlations are found with any of the PAHs. This reveals that PCBs and PAHs sources are not the same, and the main PCBs source is different than vehicle exhaust and heating (coal, wood, etc.). Tri-PCBs (18 and 31), tetra-PCBs (52, 70, and 81), penta-PCBs (105, 114, and 123), hexa-PCBs (138, 149, 153, and 157), and hepta-PCBs (189) are well correlated together. This reveals that the Alsace region is affected by different sources of PCBs, in particular those emitted from industries (steel, iron, etc.) located in the industrial zone in Kehl bordered with Strasbourg, with some possible emissions released from Entzeim airport (fuels, generators, engines, etc.). Additionally, Factor 4 shows that PCB118 (vehicle exhaust indicator) is negatively correlated with Flu and Phe, demonstrating that its main emissions originate from electrical equipment rather than vehicle exhaust.

3.4. Inter-Site Variability for Simultaneous Events

Figure 6 illustrates, respectively, the total PAHs and PCBs concentration (µg L⁻¹) for some fog samples that have occurred simultaneously (at the same time) at different sites, and successively at the same site.

Total PAHs levels reveal an inter-site variability (among the different sites) for fog events that occurred concurrently. As mentioned before, Erstein usually relies more on coal and wood for heating during October, November, and December (the fog sampling period), which is clear for most of the samples. Fog samples that occurred on 27 October 2016, 30 October 2016, 15 December 2016, 14 November 2018, and 23 November 2018 were more concentrated at Erstein (rural site) rather than Geispolsheim (suburban site). Fog samples that occurred on 28 October 2016 and 15 December 2016 had almost similar PAH levels at both sites, while on 5 November 2018, the total PAHs level at Geispolsheim was higher than that observed at Erstein. Concerning PCBs, fog events that occurred on 30 October 2016, 15 December 2016, and 5 November 2018 had almost the same PCBs level. Fog events that occurred on 27 October 2016, 28 October 2016, and 14 November 2018 were more enriched in PCBs at Erstein than Geispolsheim, but the opposite was observed for the event that occurred on 23 November 2018. This might be explained by some air masses that hit the sampling site. Nevertheless, 48 h backward trajectories showed that both sampling sites are influenced by the same air currents (west). In such cases, physical characteristics such liquid water content (LWC), temperature, volume, etc., become decisive in controlling fog chemistry. Additionally, on 13 November 2018, Strasbourg, the urban site, had the lowest total PAHs level among all proving the dominant influence of fossil fuel combustion at other sites compared to other sources. In contrast with PAHs, Strasbourg is the most concentrated site in PCBs, followed by Geispolsheim and Erstein. As suggested by PCA, PCBs are mainly released from industries, generators, etc., which are more important at the first two sites due to their proximity to the industrial zone in Kehl and Strasbourg airport, respectively. Besides, two successive fog events occurred on 18 December 2018 at Geispolsheim. The first one appeared at 9:00 and disappeared at 13:00 (GMT+2), whereas the second one appeared at 17:00 and disappeared at 21:00 (GMT+2). The total PAHs concentration level decreased from 4.4 to 2.4 µg L⁻¹ (almost by 55%). This could be explained by the wet deposition phenomenon, which proves the ability of subsequent fogwater to decrease the ambient pollutant levels and wash the atmosphere. However, the total PCBs concentration level greatly increased after the first event, which might be due to the re-volatilization of some PCBs, which allows them to become more active, acting as future condensation nuclei for the re-occurrence of a new fog event.

PCA is then applied for the whole dataset for those samples in order to identify probable sources and check inter-site variability. Figure 7 illustrates the biplot for variables (PCBs and PAHs) along the repartition of the different samples based on 50% of variability explained by the first two PCs: PC1 (35.5%) and PC2 (14.3%). PC1 is positively correlated with most PCBs, while PC2 is positively correlated with HMW PAHs and negatively correlated with LMW PAHs. The graph reveals that samples that occurred simultaneously at different sites in the Alsace were, in some cases, not statistically different. Examples include G2 and ER2 (28 October 2016), G3 and ER3 (30 October 2016), G4 and ER4 (15 December 2016), G5 and ER5 (5 November 2018), and G6 and ER6 (13 November 2018). The rest of the samples are statistically different from each other (G1 and ER1 (27 October 2016), G7 and ER7 (14 November 2018), G8 and ER8 (23 November 2018), and STG6 with G6 and ER6). Even the two successive events (1G9 and 2G9) that occurred on 18 December 2018 are statistically different from each other. Furthermore, the graph demonstrates the diversity of emission sources at all sampling sites in Alsace. This shows that a mixture of sources affects the contamination levels at Geispolsheim and Erstein such as (petroleum, diesel engines, electric equipment, etc.). Comparing the variability between the events that occurred on 13 November 2018, the result reveals that STG6 is highly correlated with PC1, proving the stronger effect of electric waste (e-waste) coming from the use of electrical equipment transported from nearby industrial areas, whereas E6 is more correlated with LMW PAHs, indicating the significant effects of petroleum sources, and G6 is in between, showing a mixed effect of both.

4. Comparison with Previous Studies

Table 5. Total PAHs concentration range (ng L⁻¹) in fog samples compared with other sites.

Site	Mount Taishan (China) [28]	Shanghai (China) [27]	Northwestern Mountains (Spain) [13]	Geispolsheim (France) This study	Erstein (France) This study	Strasbourg (France) This study	Cronenbourg (France) This study
Compounds
Nap	n.a	376 (2–1448)	n.a	534 (11–1367)	412 (255–1141)	333 (158–556)	314 (66–554)
Flu	17 (5–63)	66 (3–520)	18 (n.d–134)	112 (38–307)	173 (22–559)	58 (11–88)	180 (12–328)
Acy	24 (n.d–62)	13 (n.d–27)	n.a	n.a	n.a	n.a	n.a
Ace	28 (3–53)	30 (n.d–114)	n.a	n.a	n.a	n.a	n.a
Phe	80 (21–222)	138 (3–1043)	n.a	546 (145–1920)	750 (298–2432)	626 (252–1219)	1064 (206–1770)
Ant	13 (2–25)	172 (3–1281)	n.a	1076 (152–3181)	851 (137–3566)	996 (263–2600)	1175 (114–1315)
Flo	42 (19–95)	34 (n.d–178)	n.a	179 (23–356)	217 (57–496)	226 (114–460)	150 (103–178)
Pyr	12 (1–45)	34 (n.d–133)	24 (n.d–70)	115 (n.d–262)	298 (n.d–1259)	97 (24–197)	273 (63–553)
BaA	13 (4–51)	41 (n.d–189)	0.1 (n.d–1.2)	46 (n.d–364.1)	61 (n.d–91)	46 (n.d–57)	n.d
Chry	9 (3–35)	19 (n.d–86)	1 (n.d–15)	27 (n.d–72)	52 (n.d–73)	n.d	57 (n.d–67)
BeP	9 (n.d –47)	2 (n.d–9)	n.a	68 (n.d–79)	57 (n.d–80)	n.d	n.d
BbF	23 (1–102)	4 (n.d–22)	0.9 (n.d–10)	51 (n.d–69)	52 (n.d–86)	n.d	46 (n.d–56)
BkF	6 (n.d–38)	6 (n.d–17)	0.6 (n.d–2.1)	67 (n.d–79)	30 (n.d–45)	n.d	n.d
BaP	6 (n.d–27)	n.d	0.7 (n.d–1.7)	97 (n.d–170)	41 (n.d–56)	22 (n.d–43)	n.d
Total	273 (90–975)	982 (30–6670)	45 (8–216)	2959 (451–5866)	2994 (520–6725)	2404 (985–5132)	2765 (578–5097)

n.a: not analyzed; n.d: not detected; Ace: acenaphtene; Acy: acenaphylene.

and

Table 6. Total PCBs concentration range (ng L⁻¹) in fog samples compared with other sites.

Site	PCBs	Reference
Zürich (Switzerland)	(7000–22,000)	[68]
Northwestern Mountains (Spain)	(n.d–319)	[13]
Geispolsheim (France)	(137–12,058)	This study
Erstein (France)	(434–5787)	This study
Strasbourg (France)	(5383–15,515)	This study
Cronenbourg (France)	(934–4979)	This study

summarize some of the previous results showing the total PAHs and PCBs concentration range, respectively. Previous fog studies conducted in China [27,28] and Peninsula [13] prove that fog water is mostly enriched with LMW PAHs rather than HMW PAHs because of the high solubility of LMW PAHs in water. Same dominant compounds, Nap, Flu, Phe, Ant, and Flo, were found at all sites. The maximum total PAH concentrations observed in Alsace were lower than those at other sites, except Erstein, which has a higher maximal concentration than that obtained at Shanghai (China). However, the minimum total PAHs concentrations observed are all much higher than those obtained in China and Spain. The mean total PAH concentrations in Alsace are at least 2.5 times higher than those observed in China. The total concentration of PCBs at Strasbourg varies almost within the same range as that at Zurich (Switzerland), while those observed at other sites are much lower. The total concentrations at the four sites are much higher than those found at Northwestern (Spain).