Transport of Gas-Phase Polycyclic Aromatic Hydrocarbons to the

Three sampling sites, Northern inlet of the lagoon (Marine), Moranzani site ... 15 m high; samples were collected when the wind blew from the southeas...
0 downloads 0 Views 179KB Size
Environ. Sci. Technol. 2004, 38, 5357-5364

Transport of Gas-Phase Polycyclic Aromatic Hydrocarbons to the Venice Lagoon A N D R E A G A M B A R O , * ,†,‡ L A U R A M A N O D O R I , † I V O M O R E T , †,‡ G A B R I E L E C A P O D A G L I O , †,‡ A N D P A O L O C E S C O N †,‡ Environmental Sciences Department, Ca’ Foscari University of Venice, 30123 Venice, Italy, and Institute for the Dynamics of Environmental Processes, CNR, 30123 Venice, Italy

Concentrations of gas-phase polycyclic aromatic hydrocarbons (PAHs) were studied over one year at two sites of the Venice lagoon (designated Marine and Industrial) and at a mainland station (designated Rural) in Italy. Average ∑PAH concentrations, calculated as sum of 16 PAHs, at Marine are about three and five times lower than those at Industrial and Rural, respectively. The seasonal trends, the temperature-PAH relationship, and principal component analysis indicate that at Industrial and Marine sites several local sources (vehicle and industrial emissions, etc.) could be the PAH sources in the warmer months, whereas in the colder months the main PAH sources could alternate between vehicle emissions and residential heating. At Rural the main PAH sources are: vehicle emissions in the spring and autumn; vehicle emissions, field burning, and wood combustion in the summer; and vehicle emissions and fuel consumption for residential heating in the winter. To evaluate the contribution from different sources to the Venice Lagoon air, horizontal fluxes of PAHs have been obtained. The estimated annual flux of PAHs is about 9 times greater at Industrial (193.5 mg m-2 y-1) than at Marine (20.6 mg m-2 y-1). These results show that study of the chemical contamination of the Venice atmosphere must take into account the PAH flux derived from marine sources as well as the continental input.

Introduction The Venice Lagoon is the largest lagoon and one of the most important wetland sites in the Mediterranean Sea. It is a unique ecosystem where the mingling of human action and natural ecology has been enduring, complete, complex, and profound. Due to its position between the upland drainage and the sea, the Lagoon of Venice has been subject to important anthropogenic inputs of nutrients and pollutants, which have increased greatly with industrial and agricultural development. There are several studies of the impact of industrial and urban activities on the quality of the water and sediment of the lagoon (1-3) but very little is known about the role of the aerosol on chemical contamination of the Venice ecosystem. This study is part of a project evaluating the transport of pollutants to the Venice atmosphere by * Corresponding author fax: +39-41-2348549; e-mail: gambaro@ unive.it. † Ca’ Foscari University of Venice. ‡ Institute for the Dynamics of Environmental Processes, CNR. 10.1021/es049084s CCC: $27.50 Published on Web 09/08/2004

 2004 American Chemical Society

aerosol to answer specific queries raised by policy makers and public administrators. This paper focuses on the aerosol concentration of polycyclic aromatic hydrocarbons (PAHs) studied over one year at two sites overlooking the Venice Lagoon and at a mainland station. PAHs are ubiquitous compounds in the environment; they are generated by incomplete combustion of organic material emitted by a wide variety of industrial processes, motor vehicles, and residential heating. Belonging to a group of compounds commonly known as persistent organic pollutants (POPs), PAHs have attracted much attention in recent years because of their inherent toxicity and ability to disperse in the environment by direct emissions to the air and consequent long-range atmospheric transport. The highest concentrations of atmospheric PAHs can be found in the urban environment, due to the increasing vehicular traffic and the poor dispersion of the atmospheric pollutants. PAHs are present in the atmosphere in both gas and particulate phases, depending on the vapor pressure of each. Lighter PAHs are found predominantly in the gas phase, while those with four or more rings are mainly bound to the particle phase. Gas/particle partitioning of PAHs has been extensively studied in different urban, remote, and coastal environments as the dominant mechanism that controls the removal of PAHs from the atmosphere via dry and wet deposition (4, 5). In many studies on PAHs, only the particulate phase was collected using filters due to its higher carcinogenic potential. In practice these studies have provided much information on the multi-ringed heavier PAHs but the lighter vaporphase PAH components have been somewhat neglected. Although these lighter compounds have weaker carcinogenic/ mutagenic properties, they are the most abundant in the urban atmosphere (6, 7) and react with other pollutants to form more toxic derivatives (8). The objectives of this study were the following: (i) to investigate the seasonal trend of PAH concentrations in the aerosol around the Venice Lagoon; (ii) to characterize the PAH emission sources and to evaluate the PAH aerosol fluxes from urban and marine sources to the Venice Lagoon atmosphere.

Experimental Section Sampling and Analysis. To study the atmospheric transport of PAHs to the Venice lagoon it has been necessary to adopt a sampling strategy able to differentiate between the contributions of the different sources. Because of this, the general circulation of wind has been investigated: analysis of the historical series of meteorological data from the Venice Lagoon collected from 1975 to 2002 shows that the prevailing wind direction is from the northeast (continental sources), followed by the southeasterly direction (marine sources). Three sampling sites, Northern inlet of the lagoon (Marine), Moranzani site (Industrial), and Teolo site (Rural) were selected to characterize the PAH concentrations in ambient air for this study (Figure 1). Marine was located at the Northern inlet of the lagoon (N 45°25′21.8′′ E 12°26′12.2′′) on the lighthouse, about 15 m high; samples were collected when the wind blew from the southeast, which we assume carried predominantly “marine” aerosols from the Adriatic Sea. The second sampling site (Industrial) was south of the industrial zone of Porto Marghera and the urban area of Mestre (N 45°25′38.5′′ E 12°12′47.6′′). Industrial was developed on about 2000 ha and there are more than 40 factories, representing the chemical, petroleum, and plastic industries, etc. There is a residential area and a heavy traffic highway VOL. 38, NO. 20, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5357

FIGURE 1. Map showing the locations of the sampling stations.

TABLE 1. Summary of the Performed Campaigns, with Corresponding Mean Meteorological Dataa Marine

from

to

air temp. °C

15/03/2002 17/04/2002 19/06/2002 03/07/2002 16/07/021 05/08/2002 27/08/2002 12/09/2002 09/10/2002 22/10/2002 06/11/2002 19/11/2002 28/11/2002 19/02/2003 06/03/2003 24/03/2003 30/05/2003 17/06/2003

27/03/2002 30/04/2002 03/07/2002 16/07/2002 31/07/2002 21/08/2002 10/09/2002 24/09/2002 22/10/2002 06/11/2002 19/11/2002 28/11/2002 19/12/2002 06/03/2003 24/03/2003 09/04/2003 17/06/2003 01/07/2003

10.6 15.2 24.7 24.2 24.0 23.0 22.7 19.3 17.9 13.5 13.2 12.7 8.2 5.8 8.7 8.7 25.9 24.8

campaign I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII

Industrial

RH %

rain mm h-1

wind speed m s-1

air temp. °C

73 67 68 70 69 73 69 73 81 82 79 95 76 67 62 70 66 69

0.0 0.0 0.1 0.2 0.1 0.2 0.0 0.2 0.4 0.0 0.1 0.2 0.1 0.0 0.0 0.1 0.0 0.1

2.9 3.3 3.4 3.3 3.3 3.3 3.1 3.4 4.6 5.9 6.8 6.4 2.8 1.4 1.4 3.3 2.5 1.8

Rural

RH %

rain mm h-1

wind speed m s-1

air temp. °C

RH %

rain mm h-1

wind speed m s-1

11.0

69

0.0

3.3

11.0

65

0.0

2.2

24.6 24.3 24.3 23.4 22.5 19.0 15.6

75 77 77 81 80 80 84

0.2 0.3 0.1 0.4 0.1 0.2 0.3

4.2 4.1 3.5 3 4.2 3.5 4

22.2 21.7 20.4 16.7 13.7 12.2 7.1 11.5 5.2 5.9 8.4 8.8 26.7 25.6

72 73 72 76 82 83 77 94 82 61 60 67 58 68

0.2 0.2 0.1 0.1 0.2 0.2 0.0 0.3 0.1 0.0 0.0 0.3 0.0 0.1

1.9 2.1 1.9 2.2 2.4 2.2 1.7 2.7 3.0 1.8 2.2 3.0 1.7 2.0

8.9 13.6

85 96

0.0 0.2

3.5 3.6

7.5

84

0.0

2.3

9.8 27.4

80 77

0.1 0.0

4.6 2.8

a Rain, relative humidity and temperature were obtained by averaging the hourly data over each sampling period; the mean wind speeds were calculated in the same manner, considering only winds blowing from the set direction.

around the sampling site. The samples were collected when the wind blew from the northeast, which we hypothesize is dominated by “urban” and “industrial” aerosols. The last site (Rural) is a “possible” background site to compare with the industrial park. Rural was located on Monte Grande, in the Euganean Hills natural park, about 460 m above sea level and about 50 km from Venice and 14 km from Padua (N 45°21′43.0′′ E 11°40′ 22.4′′); it is surrounded by various trees and cultivated land. The samples were collected when the wind blew from the northeast which we suppose carried aerosols from no particular direct local sources. Table 1 indicates the sampling information and meteorological conditions during the sampling period. The sampling and analytical methods are described in detail elsewhere (9). Briefly, air samples were collected between March 2002 and July 2003 by high volume samplers (Tisch Environmental Inc., Village of Cleves, OH) equipped with a quartz fiber filter (QFF; size 102 mm, SKC, Eighty Four, PA) and a polyurethane foam plug (PUF; height 75 mm, diameter 65 mm, SKC) to separate the operationally defined “particulate” and “dissolved” phases, respectively. 5358

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 20, 2004

Sampling was performed over a 15-day period providing a total of 18 observations for Marine, 13 for Industrial, and 15 for Rural (Table 1). Average operative flow was 0.34 m3 min-1 and the collected volumes varied from 36 to 2250 m3 due to different meteorological conditions. Every month a calibration was carried out to check flow rates. Sample PUFs and QFFs were pre-cleaned as reported in a previous study (9) and after collection, for quantification purposes, samples were spiked with 13C-labeled phenanthrene (Cambridge Isotope Laboratories, Andover, MA) and Soxhlet extracted for 24 h using n-pentane/dichloromethane (2:1, v/v) mixture. Extract volumes were dried by anhydrous sodium sulfate, reduced to 5 mL under gentle nitrogen flow, and cleaned up by chromatographic column. The eluates were reduced to 100 µL, and 16 PAHs in 13 chromatographic peaks (acenaphthylene (Acy), acenaphthene (Ace), fluorene (F), dibenzothiophene (DbT), phenanthrene (Phe), anthracene (An), 1-methylphenanthrene (1MP), 4H-ciclobenzophenanthrene (4CbP), 2-methylphenanthrene (2MP), fluoranthene (Flt), pyrene (Py), benzo(g,h,i)fluoranthene + benzo[c]phenanthrene (BFP); and benzo[a]anthracene +

TABLE 2. Maximum, Minimum, and Average Concentrations of Gas-Phase PAHs at Each Sampling Sitea Marine

Acy Ace F DbT Phe An 1MP 4CbP 2MP Flt Py BFP BCT ΣPAH a

Industrial

Rural

max

min

avg.

max

min

avg.

max

min

avg.

1.9 0.79 2.2 0.20 3.1 0.18 0.40 0.25 0.19 1.0 0.67 0.090 0.094

b.l.d. b.l.d. b.l.d. 0.020 0.10 b.l.d. 0.077 0.012 0.045 0.060 0.006 0.004 b.l.d.

0.11 0.044 0.34 0.084 0.84 0.028 0.16 0.058 0.086 0.30 0.17 0.024 0.026

0.43 0.44 1.5 0.53 4.5 0.67 1.4 0.73 0.68 2.6 1.8 0.49 0.45

b.l.d. b.l.d. 0.083 0.078 0.96 b.l.d. 0.21 0.11 0.11 0.15 0.29 0.012 b.l.d.

0.10 0.14 0.43 0.21 2.1 0.17 0.54 0.28 0.28 1.0 0.80 0.13 0.13

0.29 0.95 3.4 0.70 10 1.1 0.86 0.38 0.34 4.6 1.0 0.20 0.18

b.l.d. b.l.d. 0.63 0.15 2.2 b.l.d. 0.21 0.14 0.097 0.75 b.l.d. b.l.d. b.l.d.

0.046 0.059 1.8 0.36 5.3 0.14 0.47 0.25 0.19 1.8 0.54 0.054 0.045

0.48

2.3

2.3

6.4

22

5.1

11

15

11

b.l.d. ) below the limit of detection.

chrysene + triphenylene (BCT)) were analyzed by a HewlettPackard 6890 GC equipped with an MS detector (HP 5973). Quantification was performed by comparing the area of the PAH chromatographic peaks with those of 13C labeled phenanthrene; results were corrected by individual response factors periodically evaluated. Results obtained in this and preliminary studies (9) show that “particle-associated” PAHs are negligible in Venice Lagoon aerosol samples because they represent less than 5% of the total PAH concentration and are often below the limit of detection. Forthis reason, only PAH concentrations in the “dissolved” phase will be reported in this discussion. Intensive quality control was performed by evaluating method accuracy and repeatability and by analyzing different kinds of blanks (solvent, pre-sampling, and field) as reported in previous work (9). The limit of detection (LOD, ng) was defined as the arithmetical mean of field blank mass plus 3 standard deviations (10, 11). LODs for PAHs ranged from 46 ng found for Phe to 0.5 ng found for BCT. In general, PAH masses in PUF samples were higher than LOD values, however, for some compounds values below the limit of detection (b. l. d.) were reported. The accuracy and repeatability of gas-phase PAH determinations were evaluated by analyzing spiked PUF plugs: the average (6 repetitions) of the PAH concentration sums (∑PAH) was within (10% of the known value and most PAH compounds fell within (20%. The relative standard deviation was 18% for ∑PAH, ranging from 6% (1MP) to 33% (Ace). Data Analysis. Regression analysis and principal component analysis (PCA) are the chemometric methods mainly used for the data analysis. Regression analysis is a well-known technique and has been carried out by using the classical least-squares method. PCA is a multivariate statistical method used to reduce the dimensionality of a data set. Applied to a data matrix of the samples it allows, by a mathematical procedure, the transformation of the original variables, that present a definite degree of correlation, into a set of uncorrelated orthogonal variables, called principal components. The first of these new variables (the first principal component) accounts for as much of the variability in the data as possible, and each of the succeeding components accounts for as much of the remaining variability as possible. Often, the first two or three principal components account for a large proportion of the variance present in the original data and preserve as well as possible the original data structure. A scatterplot, obtained with these principal components, often provides information on the sample features. The data processing was performed using the

Statgraphics Plus 5.1 (Manugistics, Inc., Rockville, MD) and Systat 10.2.05 (Systat Software, Inc., Richmond, CA) software packages.

Results and Discussion Maximum, minimum, and average concentrations of gasphase PAHs for Industrial, Marine, and Rural sampling sites are displayed in Table 2. The lowest annual average ∑PAH (sum of the 16 identified PAH) concentrations were detected at the Marine site (2.3 ( 2.6 ng m-3), with values of about three and five times lower than those obtained at the Industrial (6.4 ( 4.0 ng m-3) and Rural (11.4 ( 5.4 ng m-3) sites, respectively. The highest values of ∑PAH found at Rural site are due to fire burning and wood combustion that occur in the summer period at the natural park. The concentrations of single PAH compounds, falling in the pg m-3, are observed to differ by about 1 order of magnitude from one compound to another. The results indicate the 3-ring PAHs as those mainly present in the gas phase and they accounted for 50% at the Marine and Industrial sites, and 70% at the Rural site. Considering the single PAH compound concentrations, the highest annual average abundance was observed for Phe and Flt, which contributed up to 50% of ∑PAH concentration. At the Marine and Industrial sites methylated phenanthrene (1MP and 2MP) and Py were also important contributors (abundance up to 10%). The abundance of the An and the two coeluting groups of PAH (FP and BCT), on the other hand, was always less than 3% at the three sampling stations. Comparison with literature data shows that the observed annual average ∑PAH (sum of all identified PAH) concentrations at the Rural and Industrial sites are very similar to values reported for a semirural area with strong marine influence in the U.K. by Lee and Jones (12), while for the Marine site the concentration is comparable with values representing the continental background concentration of the area near Lake Superior (10). Generally, the average values obtained at the three investigated stations are lower than those reported for other Italian towns (13, 14). Possanzini et al. (15) collected daytime air samples at a location inside the largest city garden of Rome from November 2002 to April 2003. They found that the gas-phase concentrations decreased with the size of the PAH, ranging from 687 ( 580 ng m-3 (naphthalene) to 0.3 ( 0.1 ng m-3 (benzo[a]pyrene). Lodovici et al. (14) reported that the concentrations of samples collected in a highly urbanized area of Florence with intense vehicular traffic were about 7 ng m-3 and generally 1 order of magnitude greater than those collected at a suburban hill site. Several authors VOL. 38, NO. 20, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5359

FIGURE 2. Temporal trend of ΣPAH concentrations at the three sampling sites. Average temperatures over each sampling campaign (‚‚‚‚‚‚) are reported. report that Phe was more abundant in PAH air composition. Gardner et at. (16) reported the PAH mixture composition at two adjacent sites in the U.K., observing that Phe was the most abundant compound averaging from 44 to 49%, followed by Ace + F (12-14%), Flt (13-14%), and Py (13%); An was about 5% of total PAH burden. Odabasi et al. (11) collected air samples near the center of Chicago from June to October 1995. They reported that the gas-phase ∑PAH concentrations were dominated by more volatile compounds such as Phe (46%); other consistently important PAHs were Ace (18%), F (17%), Ftl (10%), and Py (6%). An was about 3% while the other PAHs with more than 3 rings were each less than 1% of the total PAH mass. Temporal Trends. Yearly variability in ∑PAH concentration expressed as coefficient of variation (CV) is 115% at the Marine site, and 64% and 45% at the Industrial and Rural sites, respectively. In particular, at Marine Ace and Acy varied considerably (more than 400%), followed by F and An (about 170%); at Industrial Ace, Acy, BFP, and BCT varied by about 115%; and at Rural Ace, Acy, and An varied by up to 180%. Temporal concentration profiles of ∑PAH and mean air temperatures over the sampling period are shown in Figure 2. The temporal trend at the Rural site corresponds to seasonal temperature changes, while at the Marine and Industrial sites they are similar to each other and opposite to the temperature profiles. At Marine and Industrial sites the highest concen5360

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 20, 2004

trations occur in winter but high values are also found in autumn and spring while the lowest occur in summer. In colder months ∑PAH concentrations range from 1.6 to 11 ng m-3 at Marine and from 5.1 to 15 ng m-3 at Industrial, while summer range values are 0.46-0.82 ng m-3 and 2.3-3.8 ng m-3 respectively. At Rural it can be observed that the highest concentrations occur in summer (11-21 ng m-3), while the lowest occur in late autumn and early winter (6.3-10 ng m-3). It is noteworthy that at the Rural site the ratio of average winter PAH concentration to average summer concentration (W/S) is lower (0.5) than at the Industrial (4.8) and Marine (13.8) sites. Several authors have found the highest PAH concentrations during winter season, although opposite trends were not excluded since the seasonal behavior of concentrations is influenced both by human emissions and meteorology changes, which can increase photolysis and thermal decomposition or dispersion conditions (17). Do¨rr et al. (18) observed highest gas-phase PAH concentrations in winter at several German towns. Smith and Harrison (19) reported a W/S ratio of 5.5 for vapor-phase samples collected in the urban area of Birmingham (U.K.), and Odabasi et al. (11) found similar fluctuations of temperature and PAH concentrations in samples collected between June and October 1995 near Lake Michigan. The low-molecular-weight compounds such as Phe and Flt dominate the ∑PAH concentrations during the study in all stations. The PAHs considered in this study show the same temporal evolution as the corresponding ∑PAH concentrations in all stations with the exception of BFP at the Rural site, which shows higher concentrations in the colder period and lower concentrations in the warmer months. A possible interpretation of these differences in the trends between the stations may be linked to seasonal variability of the main source of the PAH. The seasonal concentration changes reflect the changes in the contributions to PAH levels of combustion sources such as residential heating, industrial emissions, fire, and vehicle traffic but they are also the result of the higher atmospheric reactivity of PAH during the summer months. In particular, the high PAH concentrations during the colder months at the three sites suggest that fossil fuel usage for residential heating increased with the simultaneous presence of lower atmospheric mixing heights, lower temperatures, and decreased photolytic oxidation. By contrast, low PAH concentrations at the Industrial and Marine sites in the warmer months suggest that the PAHs generated by combustion sources such as traffic vehicles and industrial emission are subject to transformation or removal by photochemical reactions. These observations are also valid for the Rural site, but during the summer uncontrolled fires also affect the area by producing elevated amounts of PAHs (20), and volatilization from soil and vegetation is eased due to temperature increasing (12, 21). To confirm these hypotheses, the temperature/PAH relationships have been investigated and PCA has been used. The PAH ratios (e.g., Phe/An, Ftl/Py, and methylphenanthrenes/Phe) have been developed and used for the source identification of PAHs (22, 12). Hwang et al. (23), instead of using a single ratio, coupled two or more indicators to compare different sources. Moreover, Yunker et al. (24), to minimize confounding factors, restricted the calculated ratios to PAHs within a given molecular mass. Several PAH ratios and couples of two or more indicators (e.g., An to An plus Phe ratio, Flt to Flt plus Py ratio) have been studied in this work but no noteworthy results have been obtained. Temperature-PAH Relationship. Several authors have suggested that the air concentrations of semivolatile organic compounds have a close relationship with atmospheric temperature (25, 26), but the study of this relationship between PAH atmospheric concentrations and temperature is complicated due to the presence of confounding

FIGURE 4. Plot of component weights of the first two PCs considering the three sampling sites simultaneously (Marine, b; Industrial, 1; Rural, O)

FIGURE 3. ln P(PAH) as a function of inverse ambient temperature at the three sampling sites. factors, such as their degradation by atmospheric reactions and the seasonal and diurnal variability of the emission (27, 28). The analysis of this relationship may be performed using Clausius-Clapeyron diagrams of the natural logarithm of a compound’s partial pressure versus the inverse of temperature; the strength of the linear regression and its slope give information about the kind of sources involved. The regression between ln P(PAH) and T-1 was calculated for the samples at each of the sampling sites and plotted in Figure 3. A low correlation between temperature and total PAH gasphase partial pressure is found at all sampling sites even though Marine and Rural have a correlation of r >0.65. These data indicate that the sampling sites are affected by different kinds of sources, or by their different combination. Generally, the lack of temperature dependency may be explained by supposing that the air concentrations are governed by primary emission, and volatilization transfer processes are of minor importance (27); conversely, steeper slopes indicate that long-range transport influences ambient concentrations (29). This interpretation confirms the hypothesis that the Industrial site is influenced directly by local sources (industrial emissions, vehicles, and several kinds of combustion) while at the Rural site the quite strong cor-

relation and the negative slope suggest that volatilization from secondary sources (soil and vegetation), and/or long range transport govern the air concentrations even if temporally local PAH sources (fire, vehicles of tourists, etc.), principally in the warmer months, cannot be excluded. Principal Component Analysis. The correlation present in the trends can be highlighted by analyzing the standardized data matrix by PCA considering the single PAH concentrations as variables and the samplings as objects; the three sampling sites have been considered simultaneously. The variance explained from the first three principal components is of 85% (56%, 22%, 7%) and the scatterplot obtained from the first two PCs is reported in Figure 4. It can be observed that the samples of the three stations are separated with the Rural samples going from center to right, the Industrial samples going from center upward, and the Marine samples going from center to left. This suggests that the sites are subjected to different PAH sources and that, as discussed above, the temporal PAH trends at the sites are different. Regarding the sources, the PCA has been carried out for each station separately. Two significant principal components have generally been obtained and discussed. The cumulative explained variances are about 86%, 87%, and 65% for the Marine, Industrial, and Rural stations, respectively. Figure 5 shows the scatterplots and the plots of component weights obtained for the first two PCs in all their elaborations. It can be observed that PC1 is always connected to gasphase concentrations of PAHs and PC2 to the sources. At the Rural site three groups are present: (i) the samples of the spring and autumn months are linked to Ace and BCT, and the main sources could be vehicle emissions (10, 30); (ii) the samples from the warmer months are linked to F, DbT, Phe, 2-MP, and Flt, the main sources of which could be field burning and wood combustion (31) with a contribution from vehicle emissions; (iii) the samples from the colder months are linked to Acy, 1-MP, and 4CbP, the main sources of which could be residential heating with a contribution from vehicle emissions. PCA at the Industrial and Marine sites shows two groups: (i) the first is representative of the warmer months; the samples show lower PAH concentrations and not any linked to specific PAH compounds; the sources of PAHs could be the combination of several local sources (vehicle emissions, industry, etc.); (ii) the second group is scattered and is representative of the colder months; the samples show higher PAH concentrations and they are linked with all PAH compounds; the main PAH sources could alternate between residential heating and vehicle emissions. These findings are VOL. 38, NO. 20, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5361

FIGURE 5. Scatterplots and plots of the component weights of the first two PCs for (a) Marine; (b) Rural; and (c) Industrial. 5362

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 20, 2004

less evident at Marine sites, probably due to sea influence. Wania et al. (25) suggested that the concentration of semivolatile organic compounds (SOCs) in air masses which have moved large distances over sea is likely to have a small or nonexistent seasonal signal because marine climates exhibit minor annual temperature variations. Moreover SOCs in air masses have enough time to establish equilibrium conditions with the seawater surface during the passage over the water, where no point sources exist to disrupt the established conditions. PAH Fluxes. To evaluate the contribution to the Venice Lagoon air from different sources, horizontal fluxes of PAHs have been evaluated by considering the product of their concentration in the aerosol and the average wind speed over each sampling period. The Marine site shows the lowest mean horizontal flux (7.56 ng m-2 s-1) and the Rural site shows the highest (21.72 ng m-2 s-1), which is very close to the value obtained for the Industrial site (21.19 ng m-2 s-1). The seasonal trend of instantaneous fluxes reflects the trend of PAH concentrations previously described: at the Marine and Industrial sites they are higher during the colder months, while at the Rural site the highest concentrations occur over the summer seasons. To achieve an effective evaluation of the role of different sources, the horizontal fluxes have been calculated by taking account of the frequency of the wind blowing from the direction set up at every station. At every station, the estimated value of annual flux has been obtained as a sum of daily fluxes over a year, from June 1, 2002 to May 31, 2003. The estimated annual flux is about 9 times greater at the Industrial site (193.5 mg m-2 y-1) than at the Marine site (20.6 mg m-2 y-1); at both stations the highest contribution (79% and 76% respectively) to annual flux is present during the coolest months, considered from October to April. These results show that study of the chemical contamination of the Venice atmosphere has to take into account the PAH flux derived from marine sources. Moreover, Rossini et al. (32) evaluated the atmospheric input of organic pollutants to the Venice lagoon by bulk deposition sampling. In particular, they found that in the area near the industrial plants of Porto Marghera the annual load of PAHs was 389 µg m-2 y-1 from July 1998 to July 1999. In first approximation, the comparison of the atmospheric horizontal flux of PAH at the Industrial site with the previously quoted values of deposition show that at the most about 2‰ of PAHs carried by aerosol reach the lagoon ecosystem.

Acknowledgments This work was supported by the Consortium for Coordination of Research Activities concerning the Venice Lagoon System (CORILA) under the project “Role of aerosol and secondary pollution in the chemical contamination of the lagoon of Venice” and by the National Research Council of Italy (CNR). We thank the ARPAV-Meteorological Centre of Teolo (Padua) and Ente Zona Industriale di Porto Marghera (EZI) for supplying meteorological data and support during sampling activities. We are also grateful to Valter Zampieri and Silvia De Pieri for technical support.

Literature Cited (1) Moret, I.; Piazza, R.; Benedetti, M.; Gambaro, A.; Barbante, C.; Cescon, P. Determination of PCBs in Venice lagoon sediments. Chemosphere 2001, 43, 559-565. (2) Scarponi, G.; Turetta, C.; Capodaglio, G.; Toscano, G.; Barbante, C.; Moret, I.; Cescon, P. Chemometric studies in the Lagoon of Venice, Italy. J. Chem. Inf. Comput. Sci. 1998, 38, 552-562. (3) Fattore, E.; Benfenati, E.; Mariani, G.; Cools, E.; Vezzoli, G.; Fanelli, R. Analysis of organic micropollutants in sediment samples of the Venice Lagoon, Italy. Water Air Soil Pollut. 1997, 99, 237-244.

(4) Wania, F.; Axelman, J.; Broman, D. A review of processes involved in the exchange of persistent organic pollutants across the airsea interface. Environ. Pollut. 1998, 102, 3-23. (5) Cousins, I. T.; Beck, A. J.; Jones, K. C. Measuring and modelling the vertical distribution of semi-volatile organic compounds in soils: PCB and PAH soil core data. Sci. Total Environ. 1999, 228, 5-24. (6) Baek, S. O.; Goldstone, M. E.; Kirk, P. W. W.; Lester, J. N.; Perry, R. Concentrations of particulate and gaseous polycyclic aromatic hydrocarbons in London air following a reduction in the lead content of petrol in the United Kingdom. Sci. Total Environ. 1992, 111, 169-199. (7) Gambaro, A.; Piazza, R.; Manodori, L.; Ferrari, S.; Moret, I.; Cescon, P. Atmospheric input of organic pollutants to the Venice Lagoon ecosystem. Organohalogen Compd. 2002, 57, 269-272. (8) Park, S. S.; Kim, Y. J.; Kang, C. H. Atmospheric polycylic aromatic hydrocarbons in Seoul, Korea. Atmos. Environ. 2002, 36, 29172924. (9) Gambaro, A.; Manodori, L.; Moret, I., Capodaglio, G.; Cescon, P. Determination of polychlorobiphenyls and polycyclic aromatic hydrocarbons in the atmospheric aerosol of the Venice Lagoon′. Anal. Bioanal. Chem. 2004, 378, 1806-1814. (10) Simcik, M. F.; Zhang, H.; Eisenreich, S. J.; Franz, T. P. Urban Contamination of the Chicago/coastal Lake Michigan atmosphere by PCBs and PAHs during AEOLOS. Environ. Sci. Technol. 1997, 31, 2141-2147. (11) Odabasi, M.; Vardar, N.; Sofuoglu, A.; Tasdemir, Y.; Holsen, T. M. Polycyclic aromatic hydrocarbons (PAHs) in Chicago air. Sci. Total Environ. 1999, 277, 57-67. (12) Lee, R. G. M.; Jones, K. C. The influence of meteorology and air masses on daily atmospheric PCB and PAH concentrations at a UK location. Environ. Sci. Technol. 1999, 33, 705-712. (13) Lodovici, M.; Akpan, V.; Canalini, C.; Zappa, C.; Dolora, P. Polycyclic aromatic hydrocarbons in Laurus nobilis leaves as a measure of air pollution in urban and rural sites of Tuscany. Chemosphere 1998, 36, 1703-1712. (14) Lodovici, M.; Venturini, M.; Marini, E.; Grechi, D.; Dolora, P. Polycyclic aromatic hydrocarbons air levels in Florence, Italy, and their correlation with other air pollutants. Chemosphere 2003, 50, 377-382. (15) Possanzini, M.; Di Palo, V.; Gigliucci, P.; Tomasi Sciano`, M. C.; Cucinato, A. Determination of phase-distributed PAH in Rome ambient air by denuder/GC-MS method. Atmos. Environ. 2004, 38, 1727-1734. (16) Gardner, B.; Hewitt, C. N.; Jones, K. C. PAHs in air adjacent to two inland water bodies. Environ. Sci. Technol. 1995, 29, 24052413. (17) Mastral, A. M.; Calle´n, M. S. A review on polycyclic aromatic hydrocarbon (PAH) emissions from energy generation. Environ. Sci. Technol. 2000, 34, 3051-3057. (18) Do¨rr, G.; Hippelein, M.; Kaupp, H.; Hutzinger, O. Baseline contamination assessment for a new resource recovery facility in Germany: part VI: levels and profiles of polycyclic aromatic hydrocarbons (PAH) in ambient air. Chemosphere 1996, 33, 1569-1578. (19) Smith, D. J. T.; Harrison, R. M. Concentrations, trends and vehicle source profile of polynuclear aromatic hydrocarbons in the U.K. atmosphere. Atmos. Environ. 1996, 30, 2513-2525. (20) Morawska, L.; Zhang, J. J. Combustion sources of particles. 1. Health relevance and source signatures. Chemosphere 2002, 49, 1045-1058. (21) Prevedouros, K.; Brorstro¨m-Lunde´n, E.; Halsall, K. J.; Jones, K. C.; Lee, R. G. M.; Sweetman, A. J. Seasonal and long-term trends in atmospheric PAH concentrations: evidence and implications. Environ. Pollut. 2004, 128, 17-27. (22) Nielsen, T. Traffic contribution of polycyclic aromatic hydrocarbons in the center of a large city. Atmos. Environ. 1996, 30, 3481-3490. (23) Hwang, H.-M.; Wade, T. L.; Sericano, J. L. Concentrations and source characterization of polycyclic aromatic hydrocarbons in pine needles from Korea, Mexico, and United States. Atmos. Environ. 2003, 37, 2259-2267. (24) Yunker, M. B.; Macdonald, R. W.; Vingarzan, R.; Mitchell, R. H.; Goyette, D.; Sylvestre, S. PAHs in the Fraser River basin: a critical appraisal of PAH ratios as indicators of PAH source and composition. Org. Geochem. 2002, 33, 489-515. (25) Wania, F.; Haugen, J.-E.; Lei, Y. D.; MacKayl, D. Temperature dependence of atmospheric concentrations of semivolatile organic compounds. Environ. Sci. Technol. 1998, 32, 10131021. VOL. 38, NO. 20, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5363

(26) Wania, F.; Mackay, D. Tracking the distribution of persistent organic pollutants. Environ. Sci. Technol. 1996, 30, 390A-396A. (27) Dimashki, M.; Lim, L. H.; Harrison, R. M.; Harrad, S. Temporal trends, temperature dependence, and relative reactivity of atmospheric polycyclic aromatic hydrocarbons. Environ. Sci. Technol. 2001, 35, 2264-2267 (28) Sofuoglu, A.; Odabasi, M.; Tasdemir, Y.; Khalili, N. R.; Holsen, T. M. Temperature dependence of gas-phase polycyclic aromatic hydrocabron and organochlorine pesticide concentrations in Chicago air. Atmos. Environ. 2001, 35, 6503-6510. (29) Hoff, R. M.; Brice, K. A.; Halsall, C. J. Nonlinearity in the slopes of Clausius-Clapeyron plots for SVOCs. Environ. Sci. Technol. 1998, 32, 1793-1798. (30) Pengchai, P.; Nakajima, F.; Furumai, H. Estimation of origins of polycyclic aromatic hydrocarbons in size-fractionated road

5364

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 20, 2004

dust in Tokio with multivariate analysis. Diffuse Pollution & Basin Management, 7th International Conference, Dublin, Ireland, 2003, 4/48-4/54. (31) Khalili, N. R.; Scheff, P. A.; Holsen, T. M. PAH source fingerprints for coke ovens, diesel and gasoline engines, highway tunnels, and wood combustion emissions. Atmos. Environ. 1995, 29, 533-542. (32) Rossini, P.; De Lazzari, A.; Guerzoni, S.; Molinaroli, E.; Rampazzo, G.; Zancanaro, A. Atmospheric input of organic pollutants to the Venice Lagoon. Ann. Chim. 2001, 91, 491-501.

Received for review June 17, 2004. Revised manuscript received July 29, 2004. Accepted July 29, 2004. ES049084S