Environ. Sci. Technol. 2001, 35, 2264-2267
Temporal Trends, Temperature Dependence, and Relative Reactivity of Atmospheric Polycyclic Aromatic Hydrocarbons MARWAN DIMASHKI, LEE H. LIM, ROY M. HARRISON, AND STUART HARRAD* Division of Environmental Health and Risk Management, School of Geography and Environmental Sciences, University of Birmingham, Birmingham B15 2TT, United Kingdom
Atmospheric polycyclic aromatic hydrocarbon (PAH) concentrations have declined appreciably between 1992 and 1997 at an urban background site in Birmingham, U.K. In contrast, no decline is evident in the city center between 1994 and 1997. Although most PAHs display statistically significant negative correlation with temperature, so does NOx, for which traffic (a nonseasonal activity) is the major source, and for which the negative correlation with temperature reflects seasonal boundary layer depth variations. When concentrations of PAHs divided by NOx were plotted against temperature, no significant relationship was detected for any PAH, except fluorene, phenanthrene, and fluoranthene. For these PAHs, the relationship was positive, suggesting volatilization from surfaces may be appreciable. For samples collected simultaneously at the city center and urban background sites, greater negative temperature-dependence was observed at the latter location. Although this may be partly due to the fact that the enhanced reactivity of PAHs at higher temperatures exerts a greater influence at sites more distant from emissions; the dichotomy in temperature-dependent behavior and temporal trends may also be due to city center concentrations being “buffered” by volatilization from surfaces to a greater extent than those at the urban background site.
Introduction Polycyclic aromatic hydrocarbons (PAHs) are regarded as priority pollutants by both the United States Environmental Protection Agency and the European Community. The U.K. government’s Expert Panel on Air Quality Standards (EPAQS) has recently set a U.K. air quality standard for benzo[a]pyrene (B[a]P) of 0.25 ng m-3. This standard is particularly challenging given that mean concentrations of particulate phase B[a]P alone at 2 sites in Birmingham for the period JulyDecember 1997, were 0.53 and 0.8 ng m-3 (1). These Birmingham results are not anomalous when compared to those from other major U.K. cities, where B[a]P concentrations between 1991 and 1995 were in the range of 0.3-1.8 ng m-3 (2). Against this backdrop, the need to monitor temporal trends in concentrations of atmospheric PAHs has intensified, with particular reference to the changing pollution climate in U.K. cities in recent years. Although there is * Corresponding author phone: +44 121 414 7298; fax: +44 121 414 3078; e-mail:
[email protected]. 2264
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
inconsistency between inventories for the U.K. as a whole (3, 4), traffic emissions have remained the principal source of PAHs in Birmingham over the period 1992-1996/97 (1, 5). This is despite significant changes in the size and composition of the U.K. motor fleet; in particular the introduction on 1/1/93 of a mandatory requirement for all new gasoline vehicles to be fitted with catalytic converters, together with an increase in traffic volume and U.K. new market penetration of diesel domestic vehicles. In addition to the U.K. government’s Toxic Organic Micropollutants (TOMPS) program, which has been monitoring atmospheric PAH concentrations at a number of (predominantly urban) locations since 1991 (2, 6), PAHs have been measured at a number of locations in Birmingham between 1992 and 1997 (1, 5, 7). Birmingham is the second largest city in the U.K., and is part of the West Midlands conurbation that has a population of ca. 2.5 million. The data presented here thus constitute an important addition to the U.K. database on atmospheric PAHs. This paper discusses atmospheric PAH concentrations recorded in Birmingham by our research group over the period 1992-1997. In particular, we examine the data for (a) evidence of a decline in concentrations similar to those reported by other authors (2, 6, 7), and (b) the causes of any temporal changes in concentrations.
Experimental Section Sampling Campaign Details. Monitoring of both vapor and particulate phase PAHs was conducted at four locations and over several time periods (Campaigns I-IV) . In all instances, the duration of each sample was 24 h. Additional details of Campaigns I, II, and IV, have been reported previously (1, 5). The campaigns may be grouped according to the locations at which they were conducted: campaigns I and II were carried out at an urban background site, whereas monitoring during campaigns III and IV was conducted at a heavily trafficked city center locale. Campaign I. Monitoring was conducted at a roof-top location (15 m above ground level) on the University of Birmingham campus situated ca. 300 m from the nearest busy road and ca. 3 km from the city center. Fifty-five samples were collected over the periods 2/2/92-2/28/92 and 7/27/ 92-8/23/92. Campaign II. In total, 19 air samples were taken on a weekly basis between 7/15/97 and 12/9/97 at a ground level location on the University of Birmingham campus located ca. 500 m from the nearest busy road, and ca. 200 m from the site studied in campaign I. Campaign III. Monitoring of PAHs was conducted at a roof top location (10 m above ground level) in the center of Birmingham. The site was ca. 70 m from the nearest major road, and ca. 600 m from the location studied in campaign IV. In all, 35 samples were taken over the periods 7/27/94 (n ) 10) and 11/9/95-2/19/96 (n ) 25). Campaign IV. This sampling site was a roof top location (8.7 m above ground level) in the center of Birmingham ca. 7 m from a busy road. 79 samples were collected on a weekly basis over the period 5/1/96-12/9/97. Note also that sampling at this site was conducted simultaneously with that during campaign II on nineteen 24-h periods between 7/15/ 97 and 12/9/97. Air Sampling. In each campaign, samples were collected using Graseby-Andersen high-volume air samplers. Particulate-associated PAHs were trapped on Teflon-coated glass fiber filter (GFF) papers, with the vapor phase fraction collected on polyurethane foam (PUF) plugs. Sampling flow10.1021/es000232y CCC: $20.00
2001 American Chemical Society Published on Web 05/04/2001
TABLE 1. Arithmetic Mean Atmospheric Concentrations (Vapor + Particle Phases) of PAHs Recorded in Campaigns I-IV PAH
Campaign I (1992)
Campaign II (1997)
Campaign III (1994-1996)
Campaign IV (1996-1997)
London (1992)
London (1997)
acenaphthylene acenaphthene fluorene phenanthrene anthracene fluoranthene benz[a]anthracene chrysene benzo[b+j+k]fluoranthene benzo[a]pyrene indeno[123,cd]pyrene benzo[ghi]perylene dibenzo[ah+ac]anthracene ΣPAH
9.1 8.9 10.4 14.0 2.6 7.3 3.0 3.6 2.0 0.53 1.2 1.4 0.45 64.5
1.7 0.67 4.4 11.9 1.2 4.5 0.39 0.66 1.1 0.50 0.57 0.75 0.11 28.5
1.8 1.2 6.3 20.9 2.4 5.7 0.60 0.82 1.4 0.50 0.71 1.21 0.15 43.7
2.2 1.1 7.1 25.9 2.9 8.8 0.72 1.2 1.7 0.71 0.91 1.4 0.18 54.8
4.2 2.3 13.5 86.0 5.0 7.5 0.75 1.4 2.2 0.6 1.7 4.5 0.35 130
2.1 1.1 5.1 19.8 1.4 4.8 0.49 1.0 1.1 0.29 0.54 0.94 0.08 38.7
rates were maintained at ca. 0.6 to 0.8 m3 min-1 yielding individual sample volumes of ca. 1000 m3. Analytical Procedures. Detailed accounts of the analytical protocols employed in campaign I have been reported elsewhere (5); those for campaigns II and IV can be found in reference 1; and those for campaign III are identical to those employed to determine PAHs in Damascus, Syria (8). In all cases, GFFs and PUFs were Soxhlet extracted for 24 h with dichloromethane. Following solvent removal, extracts were purified using solid-phase extraction cartridges, and concentrated prior to instrumental analysis. In campaigns I and III, analysis was conducted using HPLC with combined UV/fluorescence detection, whereas in campaigns II and IV, PAHs were determined via GC/MS operated in selected ion monitoring mode. Given the differing objectives of each campaign, the exact suite of PAHs that was monitored varied. Consequently, for the purposes of comparison, data are reported here for those PAHs that were common to each campaign. These are listed in Table 1. Note that because of the differing capacities of HPLC and GC to resolve benzo[b], benzo[j], and benzo[k]fluoranthene, concentrations quoted here are for the sum of these 3 isomers. In addition, benzo[e]pyrene (B[e]P) was monitored in campaigns II and IV. Quality Control and Quality Assurance. Details of the QA/QC procedures employed during each campaign may be found elsewhere (1, 5, 8). In summary however, analytical accuracy and precision - together with inter-campaign comparability - were assessed by conducting replicate analyses of appropriate standard reference materials (i.e., SRMs 1649 (Urban Dust) and 1941a (Marine Sediment)). Regular field and laboratory blanks were also analyzed, and PAH losses were evaluated either by analysis of spiked samples (campaigns I and III), or by the use of deuterated internal standards (campaigns II and IV).
Results and Discussion Temporal Trends in Concentrations. Table 1 compares the arithmetic mean concentrations of ΣPAH and individual PAHs detected in each campaign. To help place Birmingham concentrations in context, those recorded in central London as part of the TOMPS network in both 1992 and 1997 are included (9). Although the 4 campaigns discussed here were not specifically designed to monitor temporal trends, we believe that there are sufficient data to detect any appreciable changes at both city center and urban background locations. Between 1992 and 1997, atmospheric concentrations of ΣPAH on the University of Birmingham campus declined by 56% (the range for individual PAHs is 5-92%). These data are consistent with those recorded between 1/1/92 and 12/ 31/97 in London.
By comparison, ΣPAH concentrations recorded in campaign IV are 41% greater (the range for individual PAHs is 10-79%) than those in campaign III. Whereas this apparent temporal increase is likely to be at least partly attributable to the closer proximity to traffic of the campaign IV site, data from these closely co-located city center sites reveal no evidence of a decline in PAH concentrations. Possible Reasons for Different Temporal Trends in Concentrations at City Center and Urban Background Locations. Assuming that the data in Table 1 are an accurate reflection of PAH trends at the 2 sites, it is pertinent to ask how this apparent dichotomy in temporal trends can be explained. Response to the Introduction of Catalytic Converters. A possible explanation is that there was a “step” decrease in atmospheric PAH concentrations at the beginning of 1993, as a result of U.K. legislation that introduced mandatory catalytic converters on new vehicles on 1/1/93. This is consistent with the decreases observed at all TOMPS sites over the period 1993-1994/5 (2). Such a “step” decrease would be discernible over the time period covered by the campus monitoring (1992-97), but not in the city center (1994-97). Although we cannot unequivocally dismiss this possible explanation, we do not believe it to be credible, given that the beginning of 1993 saw only the introduction of three-way catalysts on new gasoline vehicles and oxidation catalysts on some new light duty diesels, but no retrofitting of the existing fleet. Hence, any reduction in PAH emissions from the vehicle fleet in early 1993 arising from this advance in emissions control would not be large. Instead, the gradually increasing market penetration of catalyst-equipped vehicles since 1993 would be expected to elicit a similarly gradual decrease in atmospheric PAH. Different Source Contributions at the Two Locations. Although our previous work has shown traffic to be the principal source of atmospheric PAHs at both the city center and the University campus (1, 5), the relative contribution of this source is lower on campus (1). As a result, the relative contributions of other possible sources such as long-range transport, space heating, and volatilization of previously deposited PAHs from environmental surfaces will be different. Variable temporal trends in emissions from different sources could account for the observed dichotomy between temporal trends in concentrations at the 2 locations. As the seasonal behavior of PAH emissions from space heating, volatilization from surfaces, and traffic would be expected to vary (i.e., negative, positive, and no correlation with temperature, respectively) we examined the temperature-dependence of PAH concentrations to gain insights into the relative contribution of each source. As ambient temperature was recorded during campaign IV, we examined VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2265
TABLE 2. Correlation Coefficients (R) and Statistical Significance Levels (p) of the Relationships between Atmospheric PAH Concentrations (Sum of Vapor and Particle Phases) and Temperature Recorded in Campaign IV noncorrected
NOx-corrected
PAH
R
p
R
p
acenaphthylene acenaphthene fluorene phenanthrene anthracene fluoranthene benz[a]anthracene chrysene benzo[b+j+k]fluoranthene benzo[a]pyrene indeno[123,cd]pyrene benzo[ghi]perylene dibenzo[ah+ac]anthracene
-0.37 -0.26 -0.05 0.26 -0.18 -0.07 -0.36 -0.27 -0.22 -0.33 -0.32 -0.23 -0.32
0.1 0.1
the temperature-dependence of PAH concentrations monitored in this campaign. This comparison was accomplished by plotting concentrations (sum of both phases) of individual PAHs against the mean air temperature recorded for each day when sampling was conducted. Table 2 shows the correlation coefficients and statistical significance levels obtained for individual PAHs. There is a significant negative correlation (p < 0.01) between temperature and concentrations of most individual PAHs. This is consistent with previous observations regarding the seasonal behavior of atmospheric PAHs in U.K. cities (2, 6). However, there is a well-established seasonal variation in boundary layer depth. We therefore investigated the extent to which this could account for the apparent temperature-dependence of most atmospheric PAHs. To do so, we first examined the relationship between temperature and 24-h averaged concentrations of NOx measured at an automatic urban network monitoring station located ca. 600 m from the city center site in campaign IV. NOx was selected for this purpose because in the U.K. West Midlands conurbation (of which Birmingham is the major city) 85% of NOx arises from a nonseasonal source: i.e., traffic (10). Hence, the relationship between NOx and temperature will largely reflect seasonal variations in boundary layer depth, as opposed to source strength. In fact, there is a negative relationship between NOx and temperature that is statistically significant at the 99% level (R ) -0.42). Given this, we examined the relationships between temperature and concentrations of individual PAHs divided by the corresponding 24-h averaged NOx concentration. By doing so, we hypothesized that allowance would be made for the temperaturerelated variations in boundary layer depth, and that we would obtain a clearer picture of the temperature-dependence of atmospheric PAH concentrations. As shown in Table 2, in the vast majority of cases, this “NOx-correction” results in there being no correlation with temperature for many PAHs, which is consistent with the major source of such PAHs being nonseasonal - i.e., traffic. Furthermore, statistically significant positive relationships with temperature are revealed for fluorene, phenanthrene, and fluoranthene. Combined, these data strongly suggest that space heating (which in Birmingham over the period covered by the 4 campaigns has been predominantly via combustion of natural gas, from which PAH emissions are low) is not an appreciable source of atmospheric PAHs in the city center. Instead, the positive relationships with temperature observed for fluorene, phenanthrene, and fluoranthene suggest that volatilization from environmental surfaces such as roads, soil, and vegetation may constitute an appreciable source of these PAHs in the center of Birmingham. Previous studies have presented 2266
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
TABLE 3. Correlation Coefficients (R) and Statistical Significance Levels (p) of the Relationships between NOx-Corrected Atmospheric PAH Concentrations (Sum of Vapor and Particle Phases) and Temperature Recorded for Samples Collected Simultaneously at City Center and Campus Locations city center
campus
PAH
R
p
R
p
acenaphthylene acenaphthene fluorene phenanthrene anthracene fluoranthene benz[a]anthracene chrysene benzo[b+j+k]fluoranthene benzo[a]pyrene indeno[123,cd]pyrene benzo[ghi]perylene dibenzo[ah+ac]anthracene
-0.89 -0.20 0.43 0.50 0.34 0.24 -0.53 -0.23 -0.37 -0.63 -0.61 -0.27 -0.49
0.1 0.1 0.1 >0.1 0.1 >0.1 >0.1 acenaphthylene > anthracene > phenanthrene ∼ fluoranthene > fluorene. Given that data could only be compared for a limited number of PAHs, there is broad agreement between our field observations and those based on laboratory rate constant data. This study analyzes recent temporal trends in atmospheric concentrations of PAHs at both city center and urban background locations in Birmingham, U.K. In common with other studies, an appreciable decline in concentrations has occurred at the urban background site. In contrast, there is no indication of a decreasing concentration trend in the city center. Evidence is presented that volatilization from environmental surfaces is “buffering” contemporary atmospheric concentrations of some PAHs in the city center, but not on the University campus. This suggests that as primary emission control strategies exert an increasing effect, the influence of such secondary emissions may become more apparent, and may slow the temporal decrease in atmospheric PAH concentrations.
Acknowledgments The authors are grateful to the Government of Brunei for providing a scholarship to Lee Hoon Lim, to the Higher Institute of Applied Sciences and Technology, Damascus, Syria for funding Marwan Dimashki, and to the West Midlands Regional Office of the National Health Service Executive for GC/MS instrumentation.
Literature Cited (1) Lim, L. H.; Harrison, R. M.; Harrad, S. Environ. Sci. Technol. 1999, 33, 3538-3542. (2) Coleman, P. J.; Lee, R. G. M.; Alcock, R. E.; Jones, K. C. Environ. Sci. Technol. 1997, 31, 2120-2124. (3) Wild, S. R.; Jones, K. C. Environ. Poll. 1995, 88, 91-108. (4) Salway, A. G.; Eggleton, H. S.; Goodwin, J. W. L.; Murrells, T. P. UK Emissions of Air Pollutants 1970-1994, National Atmospheric Emissions Inventory Report Reference AEAT/RAMP/ 20090001/R/003; 1996; pp 45-47. (5) Harrison, R. M.; Smith, D. J. T.; Luhana, L. Environ. Sci. Technol. 1996, 30, 825-832. (6) Halsall, C. J.; Coleman, P. J.; Davis, B. J.; Burnett, V.; Waterhouse, K. S.; Harding-Jones, P.; Jones, K. C. Environ. Sci. Technol. 1994, 28, 2380-2386. (7) Smith, D. J. T.; Harrison, R. M. Atmos. Environ. 1996, 30, 25132525. (8) Dimashki, M.; Smith, D. J. T.; Harrison, R. M. Polycyclic Aromat. Compds. 1996, 9, 201-208. (9) AEA Technology. UK National Air Quality Information Archive. http://www.aeat.co.uk/netcen/airqual/data/nonauto/TOMPs.html. (10) London Research Centre. West Midlands Atmospheric Emissions Inventory; London Research Centre (now Greater London Authority): London; September 1996. (11) Gustafson, K. E.; Dickhut, R. M. Environ. Sci. Technol. 1997, 31, 140-147. (12) Lee, R. G. M.; Hung, H.; Mackay, D.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 2172-2179. (13) Atkinson, R.; Arey, J. Environ. Health Perspect. 1994, 102, 117126.
Received for review October 6, 2000. Revised manuscript received March 13, 2001. Accepted March 15, 2001. ES000232Y
VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2267
