Monitoring of the Deposition of PAHs and Metals Produced by a Steel Plant in Taranto ( Italy )

A high time-resolved monitoring campaign of bulk deposition of PAHs and metals was conducted near the industrial area and at an urban background site in province of Taranto (Italy) in order to evaluate the impact of the biggest European steel plant. The deposition fluxes of the sum of detected PAHs at the industrial area ranged from 92 to 2432 ngmd. In particular the deposition fluxes of BaP, BaA, andBkFwere, on average, 10, 14, and 8 times higher than those detected at the urban background site, respectively. The same findingwas formetals.The deposition fluxes ofNi (19.8 μgmd) andAs (2.2 μgmd) at the industrial site were about 5 times higher than those at the urban background site, while the deposition fluxes of Fe (57mgmd) and Mn (1.02mgmd) about 31 times higher. Precipitation and wind speed played an important role in PAH deposition fluxes. Fe and Mn fluxes at the industrial site resulted high when wind direction favored the transport of air masses from the steel plant to the receptor site. The impact of the industrial area was also confirmed by IP/(IP + BgP), IP/BgP, and BaP/BgP diagnostic ratios.


Introduction
Atmospheric deposition has been identified as an important source of pollutants for terrestrial and aquatic environments, especially near urban and industrial areas, where a great amount of particulate polycyclic aromatic hydrocarbon (PAHs) and metal concentrations are emitted [1][2][3].These pollutants can be transported far from their sources and they may reach very remote environments.PAHs are produced as byproducts of the incomplete combustion of fossil fuels or pyrolysis of organic materials, while metals are emitted in gaseous form from combustion processes, adsorbed on fine particles, and retained in heavier ash [4][5][6][7][8].The growing interest in these pollutants is due to their toxicity for human health [9][10][11][12][13][14]. Iron and steel industries are known to be hot spots of organic micropollutants, such as PAHs, and heavy metals.Steel is mainly produced in integrated works using a series of closely linked phases: coke ovens, sintering, blast furnace, basic oxygen steelmaking, and finishing procedures.Among these processes, coke ovens and sinter plants have been identified as potential sources of metals and PAHs.Therefore, air pollution has become a serious problem for highly industrialized areas such as Taranto (Apulia region, Italy).Taranto is one of the areas identified at high risk of environmental crisis in Italy and it has been included in the list of the polluted sites of national interest (SIN) because of a wide industrial area developed close to the urban settlement, having a high population density [15,16].Although several studies were carried out in the last decade in order to evaluate pollutant concentrations in atmosphere [17][18][19][20][21], no study was performed to evaluate deposition fluxes of PAHs and metals near the steel plants of Taranto.Therefore, this work aims to present the first and high time-resolved study on atmospheric deposition fluxes and chemical imprints near the biggest steel plant in Europe.To this end, monitoring campaigns of bulk deposition were conducted near the industrial area and at an urban background site in the province of Taranto (South of Italy).In particular, sampling periods of a week at the most critical site and two weeks in the urban background site were performed.PAHs and some metals were analyzed, and

Material and Methods
2.1.Sampling Sites.Bulk deposition samples were collected at an industrial site of Taranto (South of Italy) and at an urban background site of Mottola, a city of Taranto province (see Figure 1).Taranto (40 ∘ 28  N 17 ∘ 14  E) is the third most populated city of Southern Italy and it is the seat of one of the biggest steel plants in Europe.Moreover, in Taranto, there is an important industrial area in which metallurgical, chemical, petrochemical, and cement-producing plants, military and trade harbor, and naval shipbuilding industry are located.The industrial site (Taranto) was located near the iron and steel pole of Taranto (Figure 1).Mottola (40 ∘ 38  0  N, 17 ∘ 02  0  ) is a small city of about fifteen thousand inhabitants, situated on a hill (370 meters on sea levels) in the subregion of Murgia.The monitoring site (Mottola) was located in a suburban area of the city but close to a busy regional highway, so it can be classified as urban background site (Figure 1).October 16, 2007, to February 13, 2008, with sampling periods of 7 and 15 days each month at industrial and urban background sites, respectively.A total of 16 samples at the industrial site and 8 samples at the urban background site were collected over the investigated period.Bulk particle deposition (wet + dry depositions) collections were achieved using two bulk collectors consisting of a funnel-bottle combination openly exposed at all times (LabService Analytica Srl, Italy).Collectors were made in glass for PAHs and high density polyethylene for metals.They have an aperture of 800 cm 2 placed at 1,5 meters above ground in order to avoid the sample contamination due to ground during heavy rains.Prior to the sampling, the inner surface of the funnel was washed according to UNI EN 15980:2011 and UNI EN 15841:2010 [21,22].In each site the following meteorological parameters were also monitored: ambient temperature, solar radiation, barometric pressure, rainfall, and wind speed and direction.

Analysis of Metals.
The sample was acidified on arrival at the laboratory by adding 1 mL of concentrated nitric acid per 100 mL sample to dissolve the metals bound to particles or adsorbed to the walls of the container and to prevent growth of microorganisms [22].The whole acidified sample was vacuum filtered and the filter (cellulose acetate; pore size 0.45 m) was digested in 8 mL of nitric acid and 2 mL of hydrogen peroxide solution by using a microwave system (Milestone model Ethos D) [22].The extract was analysed by inductively coupled plasma-mass spectrometry (ICP-MS, Perkin Elmer ELAN 9000).Acidified filtrates were individually analysed by ICP-MS [22].The concentrations of Fe and Mn in bulk deposition were determined as micrograms per square metre day and Cd, As, and Ni as nanograms per square metre day.

Data Treatment.
Pearson correlation analysis was applied to collected data.Pearson's correlation coefficient (Pearson's ) is a measure of the linear correlation between two variables, giving a value between +1 and −1.Total positive or negative correlation occurred between two variables when  is equal to 1 or −1, respectively.There is no correlation between variables when  is equal to 0.

Results and Discussion
Bulk deposition fluxes of PAHs and metals obtained during the monitoring campaign at the industrial and urban sites are listed in Tables 1-4 [25,26].Pearson correlation analysis was also applied to data set showing that individual PAHs were strongly correlated with each other (Pearson's  in the range: 0.75-0.99)with the exception of BaP for which lower correlation was found (Pearson's  in the range: 0.49-0.76).This finding is not in line with the results reported in our previous work which focused on particulate PAH concentrations monitored in the same industrial area [17].However Li et al. (2009) showed that deposition fluxes of PAHs were not correlated with particulate PAHs in air because they were affected by different factors during the several seasons [27].This change can be attributed to the different distribution patterns of PAHs among the several particle size fractions, different precipitation scavenging ratios, and deposition velocities of different size particles.Moreover the different reactivity of PAHs in atmospheric processes period should be also considered [27].Various studies demonstrated that BaP degrades photolytically in the atmosphere at much faster rates than its isomers or other commonly measured parent PAHs [28,29].In highly polluted areas, as the industrial area of Taranto, the high concentrations of pollutants such as ozone and/or oxides of nitrogen promote BaP degradation.
Lower PAH deposition fluxes were detected at urban background site of Mottola: average flux over the monitoring period was 211 ng m −2 d −1 which ranged between 87 and 476 ng m −2 d −1 .This site was characterized by daily mean BaP flux of 17 ng m −2 d −1 (7-38 ng m −2 d −1 ) as in Venice Lagoon site (Italy) (33.7 ng m −2 d −1 ) [30].As shown in Table 2, the highest deposition fluxes were determined for BgP confirming Mottola site as urban one [31,32].Moreover, significant positive correlations were found among investigated PAHs (average Pearson's  = 0.88) and also for BaP (Pearson's  in the range 0.76.-0.99).Bulk deposition fluxes of the sum of PAHs (∑ PAHs) measured at industrial site were averaged over a period of 15 days in order to compare the two monitored sites (see Figure 2).Deposition fluxes of ∑ PAHs near the industrial area of Taranto were on average six times higher than those at the urban background site with peak values 10 times higher during the first four monitored periods.About single PAH, it was found that deposition fluxes of BaP, BaA, and BkF at industrial site were on average substantially 10, 14, and 8 times higher than those detected in urban background site, respectively.Lower enrichments (4-6 times) were determined for other investigated PAHs.
Average deposition fluxes of Fe, Mn, Ni, As, and Cd at the industrial and the urban background sites are shown   in Tables 3 and 4. The deposition fluxes of metals near the industrial area were higher than those determined at the urban background site for all sampling periods, except for Cd.Several studies on elemental composition of PM demonstrated that Cd can be considered as a traffic marker because it is emitted by the wearing of tyres, brakes, and other parts of vehicles [33][34][35].In particular, the deposition fluxes of Ni and As at industrial site were about 5 times higher than those determined at the urban background site, while the deposition fluxes of Fe and Mn were 31 times higher.For example, the comparison between Fe deposition fluxes at two monitored sites is reported in Figure 3. Ni and As deposition fluxes at industrial site (19.8 and 2.2 g m −2 d −1 on average, resp.) were similar to those determined at other coastal sites in Tokyo, France, and China and at industrial site in Spain [36][37][38][39].The mean deposition fluxes for Ni and As at urban background site (3.7 and 0.5 g m −2 d −1 , resp.) were similar to those determined in USA by Sweet et al. [40].Manganese at the industrial site (1.02 mg m −2 d −1 on average) resulted higher than that determined in many polluted countries such as in China [38].Nowadays no information is available on Fe concentrations in bulk deposition fluxes at the industrial and the urban sites and even less at the industrial area of Taranto that was identified as an area of high environmental risk in Italy.Therefore the high-time resolved information on Fe bulk deposition obtained by this study (57 mg m −2 d −1 on average) is of great interest to the scientific community and stakeholders.In particular it was found that Fe fluxes were two orders of magnitude higher than those determined only in dry deposition in USA and five times higher than those    determined at highly industrialized area in Turkey (Izmir) [41,42].The high content of Mn and Fe in deposition fluxes at industrial site and the strong correlation between these pollutants (Pearson's correlation coefficient = 0.95) are in line with results obtained by chemical characterization of atmospheric particulate matter in the same area [16,20,24].As reported in previous papers, Mn and Fe atmospheric concentrations measured near Taranto industrial area were the highest in Apulia region.Moreover Mn and Fe were the most significant markers of mineral park emissions within at the steel plant [16,20,24].The impact of the industrial area on two sampling sites can be highlighted by the contributions of each PAH to the total deposition fluxes.At industrial site, each PAH accounted for similar percentage to the total PAH flux (about 15-20%), except for DBA which accounted for about 4% (see Figure 4).On the contrary, the deposition fluxes at the urban background site (Figure 5) were principally due to benzo[ghi]perylene (25% of the total flux) identified as marker of vehicular traffic emissions [40,41].The different percentage compositions obtained in two sites can be due to predominant contribution of the steel plant on the industrial site.This finding was also confirmed by significant correlation among PAHs and metals at the industrial area (Table 5) and no correlation at the urban background site (Table 6).This evidence highlighted that PAHs and metals at the industrial site were emitted from the common source, the steel plant.

Sampling period Fe
The high time-resolved monitoring campaign enabled us to study the impact of meteoclimatic parameters on daily deposition fluxes of PAHs and metals.The PAH fluxes were positively correlated with rainfall during the sampling period, but significant inverse correlation was found with wind speed [27,43,44].In fact the highest PAH fluxes were registered during the periods characterized by wind calm and/or high rain frequency (>10%).As concerns metals, it was found that when wind direction favored the transport of air masses  from the steel plant to receptor site, the highest Fe and Mn fluxes were detected (Figure 6).Moreover, high fluxes were registered even during calm period.This evidence can be explained considering the proximity of receptor site to the industrial source.At the end, several diagnostic ratios (IP/(IP + BgP), IP/BgP, and BaP/BgP) were calculated in order to identify the industrial area as the main emission source.PAH diagnostic ratios have recently come into common use as a tool for identifying and assessing pollution emission sources [17,28,[45][46][47][48][49].The use of diagnostic ratios is based on the principle  that some PAHs are emitted in reasonably regular proportions and the paired compounds are diluted to a similar extent during transport, and thus their subsequent ratios remain constant between the source and receptor.Diagnostic ratios were used to identify pyrogenic or petrogenic sources, diesel or gasoline emission, fuel or biomass combustion, and traffic related sources [17,28,[45][46][47][48][49].Values of IP/(IP + BgP), IP/BgP, and BaP/BgP were 0.44, 0.78, and 1.13, respectively, in agreement with those reported in literature for coke combustion emissions [17,[45][46][47][48][49][50].Lower diagnostic ratio values were found for urban background site (0.60 for all ratios) confirming the combustion processes of gasoline and diesel as the most important source of PAHs in this site [49,50].

Conclusions
A high time-resolved monitoring campaign of bulk depositions of PAHs and metals was conducted at an industrial site (sampling period of one week) and at an urban site (sampling period of two weeks) in province of Taranto from October 16, 2007, to February 13, 2008.The analysis of deposition fluxes and the study of the impact of meteoclimatic parameters enabled the evaluation of the impact of the biggest European steel plant on the surrounding area.It was found that average deposition fluxes of BaP (147.6 ng m −2 d −1 ), BaA (162.3 ng m −2 d −1 ), and BkF (140.9 ng m −2 d −1 ) at the industrial site were on average 10, 14, and 8 times higher than those detected in urban sites, respectively.The impact of the industrial was also confirmed by diagnostic ratios as IP/(IP + BgP), IP/BgP, and BaP/BgP.The same findings were obtained for most metals.and Mn (1.02 mg m −2 d −1 ) at industrial site were 6, 5, 31, and 31 times higher than those determined at urban site, respectively.Therefore, this work has enabled us to obtain useful information about deposition fluxes of micropollutants near the biggest European steel plant, information not available in the literature up to now.

Figure 1 :
Figure 1: Map of the investigated area.
. The deposition fluxes of the sum of investigated PAHs (∑ PAHs) at the industrial area ranged from 92 to 2432 ng m −2 d −1 with an average flux of 1012 ng m −2 d −1 .Moreover, it was found that the daily mean flux of BaP was 148 ng m −2 d −1 with a range from 11 to 303 ng m −2 d −1 .BaP fluxes are comparable to those reported by Esen et al. in an industrial site of Turkey or monitored during the winter in China (220 ng m −2 d −1 ) ΣPAHs urban background site

d − 1 5 Table 2 : 3 :
and the measurement uncertainty (SD) of each PAH for the industrial site.The sum of detected PAHs (∑ PAHs) during sampling periods is listed in the last column.Bulk deposition flux ng m −2 d −1 and the measurement uncertainty (SD) of each PAH for the urban background site.The sum of the PAHs detected (∑ PAHs) during sampling periods is listed in the last column.Bulk deposition flux and the measurement uncertainty (SD) of Fe and Mn g m −2 d −1 and of Ni, As, and Cd ng m −2 d −1 for the industrial site.

Figure 4 :
Figure 4: Percentage contributions of each PAH to the total deposition fluxes at the industrial site.

Figure 5 :
Figure 5: Percentage contributions of each PAH to the total deposition fluxes at the urban background site.

Figure 6 :
Figure 6: Fe deposition fluxes (g m −2 d −1 ), wind speed (m/sec), and probability of wind direction (WDP) from North-North-West (NNW) in percentage at the industrial site.

Table 4 :
Bulk deposition flux and the measurement uncertainty (SD) of Fe and Mn g m −2 d −1 and of Ni, As, and Cd ng m −2 d −1 with standard deviation for the urban background site.

Table 5 :
Correlation coefficients among PAHs and metals collected at the industrial site.

Table 6 :
Correlation coefficients among PAHs and metals collected at the urban background site.