Environ. sei. Technol. 1994, 28,2380-2386
Polycyclic Aromatic Hydrocarbons in U.K. Urban Air Crispin J. Halsall,t Peter J. Coleman,* Brian J. Davis,* Victoria Burnett,?Keith S. Waterhouse,t Peter Harding-Jones,§and Kevin C. Jones'pt Institute of Environmental and Biological Sciences, Lancaster University, Lancaster, LA 1 4YQ U.K., Warren Spring Laboratory, Gunnels Wood Road, Stevenage, Hertfordshire, SGI 2BX U.K., and Rechem Environmental Research, Charleston Road, Hardley, Hythe, Southampton, SO4 6ZA U.K.
Polycyclic aromatic hydrocarbon (PAH) data for the first two years (January 1991-December 1992) of a national urban air monitoring scheme in the U.K. are presented. Urban sample sites were operated in the cities of London, Manchester, and Cardiff and in the light industrial town of Stevenage. Both the particulate and vapor phases of 15 PAHs were sampled using high-volume air samplers at roof-top level (-25 m). London, the largest urban center, had the highest annual mean CPAH concentrations of 166ng/m3in 1991. Phenanthrene and fluorene dominated the total PAH at each site and were present predominantly in the vapor phase throughout the year. The heavier PAHs (MW > 250) were present on the collected particulate and showed a distinct seasonal variation (winter > summer). PAH profiles were similar at each site, even though the conurbations were different in size, indicating sources common to each site. Specific atmospheric contamination episodes, associated with particular meteorological conditions, were identified throughout the 2-year period. NO2 concentrations were obtained for the Manchester site for 1991. Weak correlations ( P I0.05) were found to exist between elevated NO2 concentrations and particulatebenzo[a]pyrene and CPAH concentrations. Benzo[alpyrene concentrations measured here are compared to data reported for inner London from the 1950s, 1960% and 1970s. Contemporary air concentrations have declined substantially over this time, perhaps by around 2 orders of magnitude.
Introduction Polycyclic aromatic hydrocarbons (PAHs)produced by combustion sources are ubiquitous in the urban atmosphere. Major sources identified include vehicular traffic, use of fuel for residential heating, industrial processes, and waste incineration (1). PAHs have been linked to adverse health effects in the human population (21, and consequently environmental monitoring of these compounds is essential. Measurements of PAHs in ambient urban air were made in the U.K. as early as the 1950s (3) while more routine monitoring has taken place in Britain and elsewhere over the last 20 years (4-10). Until recently, the majority of studies have sampled the particulate PAH phase only. This has given information largely on the multi-ringed heavier PAH but has neglected the lighter PAHs, which are prevalent in the vapor phase. Although these lighter compounds have weaker carcinogenic/mutagenic properties (2),they are most abundant in the urban atmosphere and react with other pollutants to form more toxic derivatives (11). Adsorbants of the
* Author for correspondence. + Lancaster University. Warren Spring Laboratory. 5
Rechem Environmental Research.
2380
Enwiron. Sci. Technol., Vol. 28, No. 13, 1994
vapor phase component have been used increasingly in sample programs for PAHs (10, 12,13)and were routinely employed in this study. Of growing importance is the acute exposure brought about by contamination episodes in many of our cities when certain meteorological conditions result in increased pollution loading. In this study, we identify several of these episodes and their effect on PAH concentrations in ambient city air. Data for NOz, a secondary combustion product and a pollutant of great concern in the urban environment, is obtained for one of the sample sites and compared to the PAH loading. The sampling program was part of the Toxic Organic Micro-Pollutants Survey (TOMPS) set up to establish concentrations and trends of various semivolatile organic contaminants in the U.K. atmosphere. This is envisaged as a long-term study, although the results presented here are for the first two years (1991-1992) only. PCDD/Fs and PCBs are also being monitored, and the data are presented elsewhere (14, 15).
Experimental Section Sampling occurs in four urban sites through the U.K., in the three major cities of London, Manchester, and Cardiff and in the light industrial town of Stevenage (-40 km north of London). Three laboratories are involved in the TOMPs program; Warren Spring Laboratory, Rechem EnvironmentalServices, and Lancaster University. PAH samples from Manchester and Cardiff were analyzed at Lancaster University while samples from London and Stevenage were analysed a t Warren Spring Laboratory. Air Sampling. Sampling of the ambient atmosphere was carried out using high-volume air samplers supplied by General Metal Works, in the city/town center at rooftop level (-25 m). PAHs were sampled every other week, resulting in data for 26 weeks being collected for each year of the program. During a sample week, the high-volume pump was timed to run for 30 min in every hour, aspirating ca. 500 m3 of air through the week. Grant Instrument %bit squirrel loggers were incorporated into each highvolume sampler via a pressure transducing unit to enable accurate determination of the volume of air aspirated each week. The particulate phase was defined as that trapped on a Whatman 10-cm GF/A filter, and the vapor phase was defined as that associated with two in line polyurethane foam (PUF) plugs. These components in the samples from Manchester and Cardiff were analyzed separately but bulked for London and Stevenage. The filter and PUFs were held in an aluminium sampling module to aid sample placement and removal and also to reduce possible handling contamination. Similar high-volume systems using PUF/filters have been deployed successfully for the monitoring of semivolatile organics in a variety of locations (16-20). On collection, the module was transferred back 0013-936X/94/0928-2380$04.50/0
0 1994 American Chemical Society
Table 1. Mean Blank Concentrations and Limits of Detection for HPLC-Fluorescence Method and Limits of Detection for GC-MSD Methoda HPLC
PAH
PUF mean blank (ng/PUF)
ACE FLU PHE ANTH FLUO/MPHE PYR BENZANTH CHRY B[blF D[ac]A/B[klF B[alP B [ghi]P COR
ND ND 7.1 (9.7) 14.6 (15.6) 1.6 (3.5) ND 1.8 (4.0) ND 5.4 (7.7) 27.2 (12.0) 11.2 (15.1) 42.5 (16.0) 34.4 (16.9)
LOD (ng/500 m3)
filt mean blank (ng/filt)
2.5
ND ND ND 2.6 (4.4) ND ND ND ND 5.3 (7.0) 10.3 (8.0) 3.8 (5.3) 14.4 (10.1) 15.4 (7.4)
11.0
36.1 61.4 12.1
21.0 13.8 47.0 28.5 63.4 56.3 90.5 85.1
LOD (ng/500 m3) 2.5 11.0 12.5 15.2 54.0 21.0 20.5 47.0 26.3 34.3 19.7 44.7 37.6
GC-MSD LOD (ng/500 m3) 1.0 1.0 1.0 1.0 1.0 1.0
1.0 1.0 1.0
1.0 1.0 1.0
NA
LOD = limits of detection. ND = not detected. NA = not analyzed. The standard deviation are shown in parentheses.
to the laboratory where the filter and PUFs were removed separately. The filter was dried in a desiccator for 24 h and weighed. Weight differences were used to determine the total suspended particle concentration (TSP). The filter and PUFs were wrapped separately in aluminium foil and stored in a fridge at 4 "C prior to analysis. Chemical Analysis. Both laboratories involved in the sampling and analysis of PAHs followed their own protocol for sample preparation and analysis. Basically this involved Soxhlet extraction with hexane, chromatographic separation and solvent exchange on a Florisil column, and analysis of the appropriate extract using techniques detailed in Clayton et al. (21). PAHs were analyzed either by HPLC with fluorescence detection (Lancaster) or by GC-MSD (Warren Spring Laboratory). The following compounds were determined and make up the CPAH content reported: acenaphthene (ACE),fluorene (FLU), phenanthrene (PHE),anthracene (ANTH), fluoranthrene (FLUO), 1-methylphenanthrene (MPHE),pyrene (PYR),benzanthracene (BENZANTH), chrysene (CHRY),benzo[blfluoranthene (B[blF),dibenz[acl anthracene (D[acl A), benzo[kl fluoranthene (B[kl F), benzo[a]pyrene (B[alP), benzo[ghilperylene (B[ghilP), and coronene (COR). Chromatographic separation was not possible for FLUOiMPHE and D[aclA/B[klF using HPLC with fluorescence detection, so these co-elutants were quantified together. Quality Control and Assurance. A strict regime of quality control was operated in each laboratory. Before the onset of the sampling program PAH recovery studies were undertaken to verify the method. This involved spiking six separate PUF plugs and six filters with a working standard containing all the PAH compounds. These were extracted and purified in the same way as the samples. Average recoveries were determined for each compound from both the PUFs and the filters. Sample data were corrected for the method recoveries and also for the recovery of a sampling spike. The sampling spike was dibenz[ahlanthracene (D[ahlA),which was applied to the filter of air samples prior to their exposure in the field. The range of D[ahlA recoveries for the air samples was between 70 and 100%. A working standard was made up from a stock solution prepared from solid PAHs purchased from Aldrich (U.K.), Promochem (U.K.), and Greyhound (U.K.). Each compound was quantified against a series of calibration
Table 2. Mean Concentrations of PAH Analyzed in Six Batches of NBS1649 Urban Reference Dusts
PAH
qg/g (SD)
ACE FLU PHE ANTH FLUO/MPHE PYR BENZANTH CHRY B[blF D[aclA/B[klF B[alP D[ahlA B[ghi]P COR
0.27 (0.17) 0.36 (0.09) 5.05 (0.85) 0.41 (0.08) 12.89 (1.67) 6.56 (0.32) 2.38 (0.29) 3.59 (0.21) 5.74 (0.45) 5.17 (0.15) 3.30 (0.23) 0.56 (0.05) 4.22 (0.82) 4.08 (0.59)
Cert (GC/LC)
Cert (LC)
4.2-4.8 6.6-7.6' 2.3-2.9
2.4-3.4 3.4-5.6
6.5-7.5 5.9-6.7 2.4-2.8 3.4-3.8 5.9-6.5 1.9-2.1c 2.3-3.0 0.3-0.5 4.6-5.8
HPLC-fluorescence method. * Cert FLUO only. CCert B[k]F only. (I
standards made up from the stock solution. To verify the calibration, a confirmation standard containing 1.25 ng/ mL of each PAH was run at intervals within sample batches. Blanks were run with each sample batch. Limits of detection derived from blank values are presented in Table 1 for the HPLC-fluorescence and GC-MS methods. As a quality assurance procedure, certified reference material (CRM) 1649 Urban Reference Dust (supplied by the US. National Institute for Standards and Technology)was run for the selected PAHs in this program; the resulting values were in good agreement with the certified values given for this material (Table 2). Results and Discussion Annual Concentrations. Annual average PAH concentrations and ranges are summarized in Table 3 for the four urban sites. The mean annual CPAH (C of 15 compounds) ranged from 59 in Cardiff (1992) to 166 ng/ m3 in London (1991). London and Manchester had the highest mean concentrations throughout the 2 years, followed by Stevenage and Cardiff. Stevenage, the smallest urban site, had a mean CPAH concentration greater than that in Cardiff; Cardiff is the only center to be located near to the coast. Comparative studies in other urban areas using similar sampling techniques (filter with PUF backup) and a long-term sampling program provide a range of concentrations similar to this study. Keller and Envlron. Sci. Technol., Vol. 28, No. 13,
1994
2381
Table 3. Mean and Range of PAH Air Concentrations (ng/m3) at Four Urban Sites for 1991 and 1992 London
Stevenage
Manchester
Cardiff
PAH
1991
1992
1991
1992"
1991
1992
1991
1992
ACE FLU PHE ANTH FLUO FLUOiMPHEb PYR BENZANTH CHRY B[blF D[ac]A D[aclA/B[klF* B[UF B[alP B[ghilP COR CPAH
4.72 (1.34-22.6) 31.6 (6.31-161) 82.0 (9.04-396) 6.7 (0.80-39.8) 13.3 (3.12-62.4)
2.11 (0.79-5.77) 13.4 (3.44-40.2) 76.1 (23.6-492) 5.01 (1.20-9.54) 7.42 (2.28-16.4)
3.61 (0.81-14.1) 20.2 (5.30-85.8) 43.7 (8.46-196) 3.52 (0.65-14.3) 7.55 (2.03-24.1)
1.87 (0.20-4.16) 14.5 (8.43-28.2) 38.2 (17.6-57.8) 3.78 (1.31-5.75) 5.85 (2.28-12.1)
3.57 (0.08-29.9) 25.4 (0.07-111) 55.8 (4.88-173) 5.37 (0.20-22.0)
1.53 (0.05-7.70) 15.9 (0.11-71.5) 35.7 (2.25-97.7) 2.03 (ND'-9.00)
2.05 (0.6-8.54) 16.9 (0.28-117) 42.1 (2.25-174) 2.75 (0.61-21.9)
3.66 (0.16-31.1) 8.20 (1.70-18.6) 26.0 (5.42-81.6) 1.82 (ND-10.6)
12.3 (2.18-64.4) 1.82 (0.23-18.5) 3.05 (0.53-24.1) 1.66 (0.31-14.8) 0.35 (ND-5.00)
6.77 (2.11-15.8) 0.80 (0.16-4.31) 1.48 (0.41-5.64) 1.13 (0.31-3.62) 0.33 (ND-1.35)
5.72 (1.42-20.5) 1.11(0.14-10.2) 2.08 (0.31-14.2) 1.20 (0.20-8.99) 0.34 (ND-4.66)
5.49 (3.33-10.8) 1.11 (0.22-3.81) 1.67 (0.35-4.79) 1.28 (0.39-3.34) 0.44 (ND-1.35)
17.6 (0.93-76.4) 12.3 (1.69-48.7) 2.57 (0.12-11.8) 3.00 (0.37-10.3) 1.56 (0.31-5.46)
7.57 (2.39-21.5) 5.50 (1.76-24.2) 0.79 (0.14-5.64) 1.35 (0.13-4.84) 1.04 (0.13-7.74)
12.3 (3.23-79.8) 7.85 (1.00-52.4) 1.54 (0.14-13.3) 2.48 (0.17-24.3) 1.65 (0.13-12.6)
7.05 (0.16-37.8) 4.44 (0.33-22.0) 1.02 (ND-9.15) 1.69 (ND-9.19) 1.53 (ND-8.73)
1.78 (0.24-16.5) 1.06 (ND-10.2) 5.33 (0.40-85.3) NAd 166
1.02 (0.23-4.15) 0.56 (ND-3.15) 4.40 (0.49-16.8) NA 121
1.27 (0.15-10.5) 0.65 (ND-5.18) 3.26 (0.19-59.6) NA 94
1.32 (0.34-3.76) 0.63 (ND-2.36) 4.22 (0.15-12.2) NA 80
1.99 (0.27-8.23) 1.04 (0.21-7.71) 1.75 (0.23-15.0) 1.50 (ND-10.7) 1.82 (0.29-8.77) 2.46 (0.52-8.49) 1.18 (0.28-3.72) 135
1.20 (0.18-12.5) 1.81 (0.36-16.4) 0.72 (0.17-6.92) 76
1.73 (0.18-13.7) 2.00 (0.36-16.4) 0.79 (0.21-6.92) 96
0.58 (ND-5.64) 1.20 (ND-9.71) 0.40 (ND-2.48) 59
0 Sample collection ceased April 1992. b HPLC method co-elution, for Manchester and Cardiff samples only. ND = not detected. not analyzed.
Bidleman (22) reported a mean CPAH concentration of 56 ng/m3 in Columbia, SC. In Denver, CO, the mean ambient air concentration was 93 ng/m3 over a winter period (13). Jaklin and Krenmayr (23) reported a mean CPAH of 412 ng/m3 in Vienna, Austria. Sampling in Vienna took place at ground level near a busy road whereas the sampling in Columbia, Denver, and this study took place at an elevation of 10-30 m. Thrane and Mikalsen (24)found an approximate 0.3-0.5-fold decrease in CPAH concentration from ground level to a height of 25 m in Oslo, Norway. At all the sites, the lighter three- and four-ring compounds (MW < 200) were the most abundant, notably PHE and FLU. This again is consistent with other studies of urban air (8, 10). Apart from FLUO and PYR, all the other PAH compounds measured had mean concentrations an order of magnitude lower than FLU and PHE a t each of the sample sites. B[a]P concentrations ranged from ND-2.35 in Stevenage to 0.18-13.7 ng/m3 in Cardiff for the two sampling years. The German Federal Environment Agency has proposed a guideline limit of 10 ng/m3 for the annual B[a]P concentration (10). None of the urban sample sites in this study had an annual mean near this limit. Manchester had the highest mean concentration in 1991 of 1.82 ng/m3. Vapor and Particulate Phases. PAHs will partition between the particulate and vapor phases, depending on the prevailing physical conditions within the atmosphere and the particular physical characteristics of the compounds themselves. The factors affecting PAH distribution within the atmosphere have been examined by various workers (16, 25, 26). Attention has been drawn to the sampling artifacts inherent in vapor:particle sampling with high-volume samplers (27-29) essentially due to 'blow off of particulate PAH to the PUF plug. In this study blow off of a high molecular weight PAH, dibenz[a,hlanthracene, spiked directly onto the filter prior to sampling was not apparent. It is more likely that the low MW PAHs will be underestimated in the particulate phase due to the blow off artifact. The contribution to the mean CPAH concentration by the vapor phase component exceeded the particulate phase contribution by factors of 6.5 and 4.6 for Manchester and Cardiff, respectively. PHE, the most predominant PAH 2382
Environ. Sci. Technol., Vol. 28, No. 13, 1994
NA =
in the atmosphere, exists almost exclusively in the vapor phase. Conversely, the five-, six- and seven-ring PAHs (MW 250-300) were primarily associated with the particulate on the filter (>70% of B[alP, D[aclA, B[klF, B[ghilP and COR were determined on the particulate phase). The four-ring PAHs (MW 200-250) PYR, BENZANTH, and CHRY were distributed more equally between the two phases. PYR was found >BO% in the vapor phase while CHRY is approximately distributed equally between the two phases (-57% particulate). Figure la,b shows the CPAH throughout both sampling years in Manchester and Cardiff. The particulate and vapor phase contributions are represented for each sample week. Total suspended particulate (TSP) concentrations plotted against B [alp and particulate-CPAH concentrations for the Manchester data showed weak correlations of rz = 0.37 and 0.24 (P 5 0.05), respectively. This was consistent with the work of Harkov et al. (5) and Steinmetzer et al. (7), who found no correlation between filter-collected PAH and TSP. Reasons for this weak correlation are probably due to the complex nature of urban aerosol. Particle size and composition (i.e.,carbon content) will determine PAH sorption (26, 30, 31). Seasonal Variation. CPAH (vapor plus particulate) concentrations did not vary significantly between seasons (Figure 1). There was no statistically significant difference in CPAH levels between seasons in London (ANOVA P < 0.05; F = 1.2 a t 3,38 d o , for example. However, there are seasonal differences in the concentrations of some individual compounds. B[a]P in London and PHE concentrations in Manchester are represented in Figure 2a,b, respectively. B[a]P shows a marked fluctuation, with summer concentrations lo-fold lower than those of winter. Reduction in the proportion of the high molecular weight PAHs during the summer has been attributed to several factors, including less residential fuel combustion for heating (32) and greater photolytic and thermal decomposition in the warmer summer months (33). The presence of NOz, OH radicals, and O3 will influence the rate of PAH degradation in urban environments (11, 34, 35). In contrast, the concentrations of lighter vapor phase dominated compounds such as PHE tend to remain more consistent through the year (Figure 2b). Summer values may be maintained by enhanced degassing from soil and/
-
300 I
I
1
200
rn
Ii
E
-2 100
0
I
1992
II
Ilr Sample week 1991192
600
500
400 rr,
200
100
0
d ~ c o O m ~ w ~ 0 N d ~ c o 0 N d w c o 0 N ~ ~ c o ~ Na m * w ON c *oQ O c o~O N Y F w w N * w 4 4 3 3 - m ~ ~ ~ ~ m m m m ~ m d~ e ~ ~ ~ ~ ~ ~ w ~ w ~ h h h
Sample week 1991/2 Flgure 1. CPAH concentrations (vapor
+ particulate phases) for the sampling period 1991-1992:
or vegetation surfaces during times of the year when soils are warmer (I, 36). Spatial Differences in PAH Profile. At all of the sites, sampling was carried out at roof-top level where the air could be considered well mixed and away fromlocalized sources. Principle components analysis (PCA-SYSTAT Version 5.0) was applied to the data set to compare profiles between the three city sites of London, Manchester, and Cardiff and to assess the extent of any spatial variation. No distinctions were found between the individual cities. This indicates that in ambient air the PAH pattern is broadly similar at each city site. Steinmetzer et al. (7) reported a similarity in PAH profile between two Bavarian cities. Figure 3a,b shows the percentage contribution each compound makes to the mean CPAH at each of the four sites. As indicated by the PCA results, the relative proportions of each PAH at the four sites are similar, particularly the lighter three- to four-ring PAH represented in Figure 3a. The higher MW PAHs are represented in Figure 3b; here, the Y-axis is greatly expanded to
(a) Manchester site; (b) Cardiff site.
compensate for the small contribution made by the heavier compounds to the overall CPAH. CHRY and BlghiIP make the largest contribution of these seven PAH, with the B[ghilP contribution being 100% greater at the London and Stevenagesites when compared to Manchester and Cardiff. This phenomenon was not depicted in the PCA results, possibly due to the masking effects of the more abundant gas phase compounds. Contamination Episodes. Data from the Manchester site were examined with a view to understanding unusually high incidents. Elevated concentrations were noted in the sample weeks 5,47, and 101 (Figure la). These weeks were Feb 13-20,1991, December 5-11,1991, and December 16-23,1992, respectively. On each occasion, Manchester experienced stable weather conditions corresponding to high pressure and low wind speeds (37). The incident in early December 1991 (week 47) also resulted in elevated PAH concentrations in Cardiff (Figure lb, week 46), London, and Stevenage with CPAH concentrations of 568, 822, and 374 ng/m3, respectively. Static, foggy air conditions prevailed in each city at this time, and Czuczwa et Environ. Sci. Technol., Vol. 28,
No. 13, 1994 2383
h
h
c
o
SAMPLE WEEK 1991192 I30
b PHE ng/m3
1991
1992
........0........
Mean Temp
Oc
w
z
SAMPLE NEEK 1991/92
Flgure 2. PAH concentrations and temperature fluctuations at the London and Manchester sites throughout 1991 and 1992: (a) B[a]P (London); (b) PHE (Manchester).
al. (38) have shown the importance of such conditions with elevated concentrations of PCDD/Fs during fog episodes at an urban site in Switzerland. Two major contaminants found in the urban atmosphere, O3 and NO,, can react with PAH, the latter to form carcinogenic nitro derivatives (11). NO2 concentrations were obtained for the Manchester site during 1991 on dates corresponding to the PAH samples. NO2 has been implicated in the degradation of reactive PAH on particles (341, while the NOS radical has been found to react with the lighter vapor phase PAHs to form nitro derivatives (11, 35). Figure 4 shows the relationship between Manchester's CPAH concentrations with corresponding concentrations of NO2 (r2 = 0.177, P I 0.05). Although statistically significant, the correlation is a weak one. Even when particulate B[a]P concentrations were compared, a weak relationship was established (r2= 0.232, P I 0.05). Possibly a negative correlation might have been expected if increased NO2 concentrations degrade particulate-bound PAH such as B[a]P. However, the correct sampling duration with the appropriate atmospheric conditions would be crucial to observe this phenonema. Correlations between PAH and NO, concentrations have 2384
Envlron. Sci. Technol., Vol. 28, No. 13, 1994
been reported by Colmsjo et al. (39). Strong correlations were found between PAH and NO when sampled a t street level with heavy traffic. However, a t a more open site the correlation was weak. Futhermore, concentrations of NO2 showed little or no correlation with the PAH; possibly due to NO2 forming as a secondary reaction of NO, where time, dilution, and environmental factors play an important role in dispersing the contaminants in the ambient atmosphere (39). Clearly the presence of high NO2 concentrations will not necessarily result in low PAH concentrations (i.e., there appears to be no rapid degradation). An important point to note is that the reaction of PAH with NO2 is initiated by reaction with the OH radical; therefore, a high concentration of NO2 would not necessarilyresult in rapid degradation of gas phase PAH. Certainly in the contamination episodes, when NO2 concentrations were high, reductions in certain compounds such as PHE was not evident. Some gas phase PAHs are reactive in the atmosphere. For example, Kwok et al. (35)have reported on detailed, controlled experiments of the degradation of PHE in the prescence of NO3 and OH radicals. They showed the dominant losses to be by daytime reaction with OH radicals, with a lifetime of -1 day, and by
a
BENLANTH
CHliY
UlhlF
Dl'lCIA,"I*,F
"I",,,
0,11',11'
coli
FlgureS. PercentagecamributiofindbMual compoundstoilm ZPAH at the four sample sites: (a)ACE and PYR; (b) BENZANTH and COR. Note the Y-axis scaling differences between (a)and (b). COR was not analyzed at London or Stevenage. B
XXI
WH
*
........ .......
reported prior to the infamous smogs of the mid-1950s. In 1952, Waller (3) reported the measurement of 3,4benzopyrene (or B[alP) in 'town air' sampled in central London and elsewhere, using fluorescence detection. He reported the mean B[alP concentration for air sampled at County Hall in 1947-1951 to be 4.6 pg/lOO m3 or 460 ng/m3. Further measurements made by Commina and Hampton (40) through the 1960s and 1970s elsewhere in central London led them to conclude that B[alP concentrations had declined to "about one tenth" of Waller's values in 25 years. Concentrations detected at St. Bartholomews Hospital, for example, averaged 26 ng/m3 in 1962-1963 and 5 ng/m3in 1972-1973. Our data suggest this decline in B[alP has continued still further-to an average of 0.8 ng/m3 in 1991-1992. The highest central London value measured during our study was 10.2 ng/m3 compared to 3300ng/m3reportedfor'foggydays'byWaller(3). Clearly, even allowing for improvements in analytical techniques and the inevitable spatial differences between the sites used by Waller (3),Commins and Hampton (40), and ourselves, thesedata provide good evidencethat air quality in London has improved considerably with respect to this particular aromatic hydrocarbon over the last 45 years or so-perhaps by 2 orders of magnitude. Coal burning for domestic space heating was widely practiced in London up to the 1950s. These numerous diffuse and inefficient combustion sources are likely to have generated a much greater PAH burden to the atmosphere than the modern gas and electric domestic heating appliances that are used most extensively in modern homes. In contrast, this general trend in improved B[alP concentrations in inner London has occurred over a time when vehicle use (with gasoline and diesel consumption)has increased enormouslyin the U.K. as awhole.
ns,ni
NO1ppb
Acknowledgments The TOMPs program is funded by the Department of the Environment, Air Quality Division. and administered by the Warren Spring Laboratory. Literature Cited Wild, S. R.; Jones, K. C. Enuiron. Pollut. 1995,88 (1). in press. (2) International Agency for Research on Cancer. Overall evaluations of carcinogenicity: an updating of IARC monograph volumes 1 4 2 .Monogr. Eual. CarcinogenicRisk Chem. Humans 1987, Suppl. 7. (3) Waller, R. E. Br. J. Cancer 1952, 6, 8. (4) Gordon, R. J.; Bryan, R. J. Enuiron. Sci. Technol. 1973,11,
(1)
\*\IICEWBLI
V"JI
Flgure 4. ZPAH and NO2 concentrations in Manchester air for 1991. nighttime reaction with NO3 radicals, with a lifetime of 5 h. The calculated overall lifetime was then
