Characterization of PAHs Trapped in the Soot from the Combustion of Various Mediterranean Species

: Climate change causes more frequent and destructive wildﬁres even transforming them into megaﬁre. Moreover, all biomass ﬁres produce emissions of carbon compounds in the form of soot to the atmosphere with a signiﬁcant impact on the environment and human health. Indeed, the soot is causing the formation of PAHs from (a) the high temperature thermal alteration of natural product precursors in the source organic matter and (b) the recombination of molecular fragments in the smoke. However, these molecules are known to have carcinogenic effects on human health. It is therefore interesting to quantify the 16 PAHs concentration extracted from soot emitted in open diffusion ﬂame of biomass combustion. To achieve this objective, an analytical method developed for the study of kerosene combustion has been adapted for soot from biomass. This new method allowed to quantify the 16 PAHs deﬁned as priority pollutants by the US EPA for their carcinogenic mutagenic effect and on human health.


Introduction
Climate change is increasing the risk of wildfires in temperate biomes, with more extreme landscape drying combined with extreme weather events leading to increased occurrence of destructive fires [1]. In European countries bordering the Mediterranean Basin, over 40,000 fires per year were reported between 2010 and 2016 [2]. The impact of Mediterranean wildfires involves several sectors entailing environmental, human, and economic losses.
Wildfires emit carbon dioxide and other greenhouse gases that will contribute to warm the planet well into the future. They damage forests that would otherwise remove CO 2 from the air and they inject soot and other aerosols into the atmosphere, with complex effects on warming and cooling. On a global scale, wildfires emit approximately 34% of total atmospheric soot mass [3] and the carbon released from fires during combustion alters the global carbon balance [4]. For example, black carbon strongly absorbs solar radiation and atmospheric particulate matter (PM) also acts as cloud condensation nuclei (CCN), which are important for the radiation balance and the hydrological cycle [5].
With the impact on the environment, the other worrying aspect related to wildfires is the adverse effect on health caused by the exposure to biomass combustion emission [6,7]. It is well known that wildfires can produce substantial increases in the concentration of gaseous pollutants such as carbon monoxide (CO), nitrogen oxides (NOx), ozone (O 3 ), and volatile organic compounds (VOCs) [8][9][10] as well as particulate matter (PM) [11,12] and particular carbon nanoparticles, namely soot. Soot is an important component of very fine suspended dust, often referred to in the literature as elemental carbon (EC); alternative names are black carbon (BC) or black smoke. This waste of combustion process consists of compounds including amorphous carbon, graphite-like connections, fullerenes, ashes, and metal compounds. Epidemiological studies have shown that inhalation of soot particles can cause pulmonary disease, cardiovascular damage and mortality [12]. Despite the cause and effect relationship, the mechanism(s) by which this occurs is yet unclear. One hypothesis is that these ultra-fine particles exert toxicity by generating reactive oxygen species, OH*, in a reaction analogous to the Fenton reaction [13]. The reactivity of soot in radical formation processes could be highly dependent upon its nanostructure. Additionally, a possible explanation of the hazardous health effects of soot is their association with Polycyclic Aromatic Hydrocarbons (PAHs). Although detailed mechanisms are not well understood, PAHs formed in combustion processes are thought to be precursors of soot [14]. Among these PAHs, the International Agency for Research on Cancer (IARC) has determined the Benzo[a]pyrene as carcinogenic and Dibenzo[a,h]anthracene as probably carcinogenic to humans. The following PAHs: benz[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, and indeno [1,2,3-cd]pyrene are possibly carcinogenic to humans; while acenapthene, fluorene, phenanthrene, anthracene, fluorenthene, pyrene, and benzo[ghi]perylene are not classifiable as to their carcinogenicity to humans. This list is quite similar to the 16 priority PAHs listed by the Environmental Protection Agency of United States (US EPA). Acenaphtylene has not been classified for carcinogenic effects by the IARC or US EPA. Moreover, it has been observed that the total PAH content is higher for combustion particles from wood smoke as compared to vehicle exhaust [15].
Aware of the impact on the environment and the health of populations of the combustion of biomass, some authors have focused on the regulatory aspect in terms of health and safety at work. So, they are turned towards the domestic use of wood as a heating system [16] with experiments carried out in a combustion chamber under perfectly controlled conditions. Other authors have already identified and quantified PAHs emitted during forest fires. These previous papers dealing with the smoke emissions of prescribed fire plumes [17][18][19][20] are focused on particulate matters such as PM 2.5 or PM 10.
However, to our knowledge, no study has proposed to quantify the PAHs extracted from the soot obtained as close as possible to the flames. However, it seems important to us to get closer to the real working conditions of firefighters because they present a significant risk of inhalation of soot.
In this context, the objective of this study is to characterize the PAHs contained in the soot emitted in wildland fires conditions. To achieve this goal, laboratory experiments were carried out in order to be as close as possible to the source and to ensure a good collection of the soot. The samples studied were representative of the Mediterranean shrublands and it was characterized by the presence of the following dominant species: rockrose, rosemary, strawberry tree, lentisk, and pine. The 16 PAHs which are a priority for the US EPA based on concerns that they cause or might cause cancer in humans were quantified using a multi-step analytical method requiring extraction, purification. and quantification by HPLC of the target molecules.

Fuels
Plant material was collected from a natural Mediterranean ecosystem located at 450 m height above sea level and situated far away from urban areas (near the little town of Corte in Corsica) in order to prevent any pollution on the samples. The selected plants: rockrose (Cistus monspeliensis: CM), rosemary (Rosmarinus officinalis: RO), strawberry tree (Arbutus unedo: AU), lentisk (Pistacia lentiscus: PL), and pine (Pinus pinaster: PP) are representative species of the Mediterranean vegetation concerned with wildland fires. Aerial parts of each plant were collected at the beginning of April. For each species, a bulk sample from six individual plants was collected in order to minimize interspecies differences. Current year mature leaves were selected, excluding newly developed tissues at the top of the twigs. To characterize the fuels, elemental analysis was performed according to the standard EN ISO 16,948 and results are presented in Table 1.

Soot Production (Combustion Process)
Samples were previously dried in an oven at 60 • C for 24 h before experimental burning in order to get closer to summer conditions. Then, 20 g of plants were subjected to an external radiative heat flux and burned as open diffusion flame, under the hood, such as they are burned in the field. Using radiant panels, the fuels were submitted to a heat flux of 20 kW·m −2 (843 K), selected to be in the temperature range producing soot. The soot produced, including PAHs, was collected on a stainless steel grid of 0.5 mm diameter holes placed inside an aspiration hood. the size of the pores of the grid has been determined to collect a maximum of soot [21]. After sample collection, the soot was weighted using an analytical balance with a resolution of 0.01 mg, and placed in borosilicate glass test tubes and stored at 4 • C until analysis. Three events of combustion were carried out for each of the five biomass fuels in the same experimental conditions.

PAHs
PAHs are ubiquitous pollutants, comprising hundreds of lipophilic compounds with fused aromatic rings. They may have natural or anthropogenic origins, existing in many commonly used goods. In the pyrolysis phase of the combustion process, PAHs compounds are formed by complex mechanisms involving cyclization and aromatization of carbonaceous matter at temperatures typically higher than 673 K [22,23]. The United States Environmental Protection Agency (US EPA) listed 16 PAHs compounds as priority pollutants for environmental risk assessment due to their toxic effects [24]. Table 2 presents the US EPA 16 PAHs list and the acronyms used in this work. PAHs can be divided into two classes: Low Molecular Weight PAHs (LPAHs) with two and three benzenoid rings (Nap, Acy, Ace, Flu, Phe, and Ant), and High Molecular Weight PAHs (HPAHs) with four or more benzenoid rings (Flt, Pyr, BaA, Chry, BbF, BkF, BaP, DahA, BghiP, and IcdP). The International Agency for Research on Cancer (IARC) classifies PAHs into different categories, according to carcinogenic potency (Table 1); BaP is classified in group 1-Carcinogenic to humans, and the others are 2 A and 2 B. For almost 30 years, correlations of PAHs toxicities with the most toxic BaP have been pursued [26,27]. The US EPA Provisional Guidance for Quantitative Risk Assessment of Polycyclic Aromatic Hydrocarbons (EPA/600/R-93/089, July 1993), recommends that Relative Potency Factors (RPF) be used to convert concentrations of carcinogenic PAHs to an equivalent concentration of benzo[a]pyrene when assessing the cancer risks posed by these substances from oral exposures [26]. Nisbet and Lagoy (N&L) proposed to consider a Toxic Equivalent Factor (TEF) presented in Table 2 [27]. While the US EPA considers DahA as carcinogenic as BaP, Nisbet and Lagoy attributed a factor of five to DahA, for its carcinogenicity proved to be higher than BaP in low doses. Other equivalence systems are used worldwide and still no consensus exists. However, in the EU also [28] the priority is given to these 16 PAHs, this is the reason this study is focused on these 16 compounds.

Ultrasonic Extraction
For PAH extraction from soot samples, an ultrasound bath was used. Various solvents such as dichloromethane, acetone, and acetonitrile were tested. Dichloromethane and acetone require the use of an ice bath to condense the emitted vapors. So, acetonitrile was selected for its higher boiling temperature which prevents any loss of the lightest PAHs (i.e., naphthalene, acenapthylene, acenapthtene). The time of extraction was 30 min with the use of 20 mL of acetonitrile as extraction solvent. Then, 0.45 µm filters (Acrodisc ® , Merck KGaA, Darmstadt, Germany) with glass syringes of 5 mL (Fortuna Optima ® , Poulten & Graf GmbH, Wertheim, Germany) were used for filtering soot samples before SPE extraction.

Solid Phase Extraction-SPE
One of the main difficulties of the clean-up step is the selection of the appropriate solid phase. To find the best adsorbent phase, we were guided by the literature and the work of Andrade et al. [29], who have worked on particularly sullied soot samples from the combustion of kerosene. The authors have shown that only silica-C18 is able to retain PAHs strongly enough to avoid losses during the clean-up procedure. On another hand, glass columns must be employed instead of polymer cartridges in order to avoid any interaction with acetonitrile. So silica-C18 bulk from Supelco was used for filling the clean-up glass columns. A vacuum manifold SPE 12-G from Baker allowed the conditioning of the SPE column and the elution of compounds.

Conditioning of the Cleaning Column
The column should be conditioned with an organic solvent with a polarity similar to that of the sample. This is a necessary step since the column, frit, and adsorbent may contain impurities that can interfere with the analysis, especially at low concentrations [30]. In addition, this process activates the octadecyl chains in the column which facilitates the retention of polarisable molecules (i.e., PAHs). The conditioning was carried out with 2 mL of acetonitrile and 2 mL of methanol.

Loading and Drying
In this step, 2 mL of sample were passed through the column by gravity. Then the drying is performed under a slight vacuum (900 mbar) for 30 min to avoid a loss of light PAHs. This step allows the use of solvents not miscible with acetonitrile or methanol, such as hexane, in the next elution step. Furthermore, this process facilitates the elution of the analytes of interest.

Elution of PAHs
As a result of multiple tests, we found that the best results were obtained using hexane as the elution solvent as its toxicity is lower than that of other possible organic solvents, such as toluene. The elution was carried out under vacuum (600 mbar) with 4 mL of hexane.

High Pressure Liquid Chromatography-HPLC
Chromatographic experiments were performed by using an HPLC system Flexar from Perkin Elmer ® . The instrument integrates a UV/Vis photodiode array detector and a spectrofluorimetric detector. A pre-column SUPELCO LC-18 (5 µm, 200 mm length × 4 mm ID, Supelco, Bellefonte, PA, USA) was positioned before the analytical column a SUPELCOSIL LC-PAH (5 µm, 250 mm length × 4.6 mm ID, Supelco, Bellefonte, PA, USA). Acetonitrile and water were used as eluent components at a flow rate of 1 mL/min. The program started with 60% acetonitrile for 10 min. Linear gradient elution from 60% to 100% acetonitrile during 30 min was applied, followed by isocratic elution with acetonitrile for 12 min to rinse the column. To remove possible contaminants left behind, each run was concluded with a conditioning step (60% Acetonitrile/40%water) for 3 min. Then, 20 µL of the eluates were injected onto the column, utilizing the sample injector. The oven temperature was maintained at 30 • C throughout the analysis. The fluorescence excitation and emission wavelengths were changed during the chromatographic separation to have the highest fluorescence intensity. The programmed wavelengths were determined using an important concentrate standard of 16 PAHs at 10 ppm to estimate the retention time of each PAH. Using these preliminary results, the wavelengths settings were determined and presented in Appendix A. For the quantification, a mixture of 16 PAHs at 10 µg/mL of each component in acetonitrile (Supelco, Bellefonte, PA, USA) from Sigma Aldrich was used for preparing standards by dilution. Components in the mixture were: acenapthene, acenaphthylene, anthracene, benz(a)anthracene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(ghi)peyrlene, benzo(a)pyrene, chrysene, dibenzo(ah)anthracene, fluoranthene, fluorene, indeno(1,2,3-cd)pyrene, naphthalene, phenanthrene, and pyrene. The calibration range was carried out using five known standards with concentrations between 5 to 50 ppb. Limits of detection (LOD) were calculated using a signal-to-noise ratio of 3 are shown in Table 3. For the whole procedure of extraction and analysis, standard deviations within the range of 3-7% were measured. To give a whole overview of the analytical method, Figure 1 summarizes the main experimental conditions of the four steps.

Soot and PAH Production
For the first time, the rate of production of soot of each fuel was estimated and compared to the rate of production of all the PAHs in order to quantify the rate of PAHs trapped in the soot. Figure 2 presents the rate of soot and PAH production of the five biomass fuels.

Soot and PAH Production
For the first time, the rate of production of soot of each fuel was estimated and compared to the rate of production of all the PAHs in order to quantify the rate of PAHs trapped in the soot. Figure 2 presents the rate of soot and PAH production of the five biomass fuels.  Figure 1 highlights that the Pistacias Lentiscus is the species that produces the most significant amount of soot with more than 1% of its dry mass. On the other hand is the Pinus Pinaster, although emitting small quantities of soot, it contains higher amounts of PAHs than the Pistacias Lentiscus. For Rosmarinus officinialis, 74% of its soot contains the 16 PAHs; whereas for Pinus Pinster, the total amount of the PAHs only represents 4% of the soot. There appears to be no correlation between the amount of soot produced and the content of PAHs trapped inside. In addition, the molar H/C ratio was calculated in order to correlate the soot production with the combustion process. As expected, the higher the ratio, the better the combustion and, therefore, the soot production is lower. The identification and abundances of the 16 PAHs of interest extracted from the various biomass soots are given in Table 4.   Figure 1 highlights that the Pistacias Lentiscus is the species that produces the most significant amount of soot with more than 1% of its dry mass. On the other hand is the Pinus Pinaster, although emitting small quantities of soot, it contains higher amounts of PAHs than the Pistacias Lentiscus. For Rosmarinus officinialis, 74% of its soot contains the 16 PAHs; whereas for Pinus Pinster, the total amount of the PAHs only represents 4% of the soot. There appears to be no correlation between the amount of soot produced and the content of PAHs trapped inside. In addition, the molar H/C ratio was calculated in order to correlate the soot production with the combustion process. As expected, the higher the ratio, the better the combustion and, therefore, the soot production is lower. The identification and abundances of the 16 PAHs of interest extracted from the various biomass soots are given in Table 4.
The biomass soot was dominated by low molecular weight. For all species, the major PAH are acenaphtylene, acenapthene, and phenanthrene followed by benzo[a]anthracene, chrysene and pyrene in lesser amounts. It is important to note that benzo[a]anthracene which exhibits mutagenic and genotoxic potential is present in significant quantities. The formation mechanisms of PAHs during the combustion of biofuels are not fully understood, but fluorene and phenanthrene are the most abundant PAHs from cellulose pyrolysis [22], which explains the high phenanthrene fractions of the particle phase PAHs observed in this study. Oros and Simoneit [31] also found a large amount of phenanthrene when studying samples from burning pine. Alvès et al. [19] reported a significant concentration of phenanthrene and pyrene when they analyzed PAH content in particulate matters emitted during prescribed wildfires. The same predominant compounds have been found by Lee et al. [32] when they quantified PAHs in organic carbon extracted from particles obtained from prescribed burning. Nevertheless, the concentrations are less important probably due to the collection of particles being carried out a few meters from the burning site. Looking at a similar study performed with boilers, Szatylowcz and Skoczko [16] found a total production of the 16 PAHs of 230.44 mg/kg (equivalent to ng/mg) for mixed firewood, whereas 285.75 ng/mg were found for Pinus Pinaster, the more productive species. Thus, even under different experimental conditions, the rate of PAHs trapped in soot are in the same range. In order to appreciate the variability of the PAH composition according to the different plant species, Figure 3 represents the percentage of PAHs according to the number of rings for each biomass fuel.  According to this graphic, one can note that the PAH composition is dominated by the three-ring PAHs. On the other hand, although the two rings only represent naphthalene, it is well-present compared to the five-ring group. In general, it would seem that the influence of interspecificity on the distribution of PAHs in the soot is less. Similar According to this graphic, one can note that the PAH composition is dominated by the three-ring PAHs. On the other hand, although the two rings only represent naphthalene, it is well-present compared to the five-ring group. In general, it would seem that the influence of interspecificity on the distribution of PAHs in the soot is less. Similar results have been found by Struppe et al. [33] when desorbing pyrolysis wood samples. They obtained a maximal concentration of two rings compared to the PAHs between 2 to 4 rings.

Diagnostic Ratios
PAHs in biomass burning emissions are generally the same as those from other anthropogenic combustion emissions. The diagnostic ratios for PAHs were used to investigate PAH origin in the atmosphere [19,34]. The ratios of phenanthrene to phenanthrene plus anthracene (Phe/(Phe + Ant)), fluoranthene to fluoranthene plus pyrene (Flu/(Flu + Pyr)), indeno [1,2,3-cd] Table 5, and compared with other sources. The ratio of Phe/(Phe +Ant) ranges from 0.68 to 0.82 in agreement with other works on shrubs [20,34,35] except for AU, which presents a very high ratio of 0.93 as also found by Vicente when studying particulate matters emitted from wildfires [36]. The Flu/(Flu + Pyr) ratios exhibit an average of 0.53 in the same way as other studies on wood combustion [19,34,35]. For IcdP/(IcdP + BghiP) for some fuels, this ratio could not be estimated due to the absence of indeno [1,2,3-cd]pyrene. This is the reason why the average was not calculated. In the other case, the results present a ratio of around 0.4 specific to lignite combustion emissions [35]. The ratio BFs/BghiP varies from species to species but the same diversity of results has been reported by Vicente et al.

Conclusions
Polycyclic Aromatic Hydrocarbons (PAHs) are associated with health hazardous effects, and wildfires are major sources of their presence in atmospheric aerosols and especially in soot. This work reports on the chemical characterization of PAHs trapped in the soot released by five biomass species under diffusion flame combustion. An analytical method developed for kerosene pool fire was adapted to biomass soot emission. This method includes three steps: ultrasonic extraction, solid phase extraction, and HPLC quantification with reproducible results. When quantifying the 16 PAHs present in biomass soot, an important variability of concentrations was observed depending on the species ranging from 285.75 ng/mg of soot for Pinus Pinaster to 14.35 ng/mg for Rosmarinus Officinalis. Even if the concentrations are very disparate, it appears that PAHs with three rings contributed largely to the total concentration of the 16 PAHs. Various diagnostic ratios have been determined and compared with those of similar studies looking at the toxicological impact of wildfires. The obtained results are in agreement with these previous ones. After this chemical characterization of PAHs, it would be interesting to carry out further investigations to evaluate the health risk for firefighters. Funding: This research received no external funding.