Analysis of Polycyclic Aromatic Hydrocarbons in ... - ACS Publications

---

Analysis of Polycyclic Aromatic Hydrocarbons in Air Particulate Matter from a Lightly Industrialized Urban Area John R. Cretney Christchurch Polytechnic, Christchurch, New Zealand

Hlan K. Lee and Graeme J. Wright * Chemlstry Department, University of Canterbury, Christchurch, New Zealand

Wllllam H. Swallow and Mlchael C. Taylor Chemistry Division, D.S.I.R., Chrlstchurch, New Zealand

rn Levels of polycyclic aromatic hydrocarbons (PAH) have been measured in the air particulate matter (APM) of a lightly industrialized urban area in which the main sources of PAH are domestic fires and automobiles. The PAH were analyzed by glass capillary gas chromatography combined with mass spectrometry. Statistical analysis of the results using 26 PAH allows the relative contribution of the two sources to be estimated and shows that during periods of heavy pollution (APM > 100 pg m-3) domestic fires are the main contributors. A simpler ratio, In (1 + [benzofluoranthenes]/ [total PAH]) - In (1 [benzo[ghi]perylene]/ [totalPAH]), gives similar discrimination.

+

Introduction Historically Christchurch has had an air pollution problem in winter when climatic and topographical conditions lead to a stagnant pool of cold air over the city. Under such conditions the levels of pollutants can become quite high. During 1979, for example, the WHO recommended limit for smoke was exceeded on 22 days (1). The population of the Christchurch urban area is approximately 300 000, motor vehicles (predominantly petrol engined) number about 136000 (2),and the city is lightly industrialized. During the winter domestic fires are a popular form of heating. The aim of this study was to examine the type and concentration of polycyclic aromatic hydrocarbons (PAH) present in Christchurch air particulate matter (APM). PAH are generally formed by the incomplete combustion of organic material ( 3 , 4 ) ,usually fossil fuels, and several of these compounds have been shown to be potent carcinogens (5). The relative amounts of individual PAH vary considerably depending upon the source of the emission (6),and it was considered that the PAH profile obtained from the APM might give an indication of origin of the PAH pollution in the city. Experimental Section Samples of APM were collected over a 24-h period unless otherwise indicated, the sampling period ending at 9 a.m. on the date specified for the sample. A high volume dia0013-936X/85/0919-0397$01.50/0

phragm pump was used to draw air through filters held in two 37-mm filter holders in parallel; the holders were mounted so that the filter surface was horizontal, and the whole assembly was protected by a cover. The filters were held between 2.5 and 3 m above the ground. The meters used to measure the volume of air pulled through the filters were calibrated against standardized meters at flow rates of 1.5 and 6.0 m3 h-' and found to be accurate to within 1%. Whatman GF/B glass-fiber filters were used until Aug 16, 1979; after this date Gelman GF/A glass-fiber filters were used. Before use the filters were extracted in a Soxhlet apparatus for 8 h with acetone, dried at 200 OC for 3 h or fired at 400 "C for 1h, stored in a desiccator for 24 h, and then weighed. After sampling the filters were dried in a darkened desiccator for 24 h, reweighed, and stored in the dark in sealed polythene bags at 0 OC until required. A Gelman high volume sampler fitted with a two-speed hurricane pump and 200 mm X 250 mm filter holder mounted to hold the filter vertically 0.75 m above the ground was used to collect sufficient APM for PAH analysis over short periods (up to 2 h). Samples were collected on Gelman GF/A filter papers and treated in the same manner as the small filters. Sampling Sites. Most of the samples obtained for analysis were collected over 24-h periods at three sites used regularly by the New Zealand Department of Health Air Pollution Monitoring Unit and approved by the World Health Organization (WHO) as official monitoring sites. These sites were located in Avonside, at Bealey Avenue, and at Manchester Street. The pollution at the inner city site of Manchester Street was expected to have a major automobile contribution, whereas the other two inner suburb sites were believed to have a greater domestic component. In addition, APM samples were taken from the three city car park buildings with the high volume sampler. Soot samples representative of domestic pollution were obtained from the tops of chimneys of conventional open fireplaces, all of which had used more than one of the usual fuels: coal, coke, and wood. Analysis for PAH. Filters containing the APM were placed in a Soxhlet apparatus and an internal standard, benzo[b]chrysene, added at a level of approximately 10

0 1985 American Chemical Society

Envlron. Sci. Technoi., Vol. 19, No. 5, 1985

397

0 min

I

10

20

I

I

30

I

Time

Figure 1. Gas chromatographictrace of poiycyclic aromatic hydrocarbons from APM (Avonside site). For key to peak identification, see Table 1.

kg125 mg of APM. The filter was then extracted with cyclohexane for 8 h. (Soot samples were placed in the Soxhlet directly and extracted in the same way.) The cyclohexane extracts were cleaned up by the method of Grimmer and Bohnke (7) which involved washing with basic aqueous methanol (1:4), liquid-liquid extraction with dimethylformamide (DMF)-water (9:1), addition of an equal volume of water to the DMF phase and reextracting with cyclohexane, and chromatography of the residue on silica gel using cyclohexane followed by chromatography on Sephadex LH-20 using propan-2-01. Only the column fraction that contained PAH with four or more rings was further analyzed. The residue from the Sephadex LH-20 column was analyzed quantitatively by gas chromatography (HP5710 GC with FID) and qualitatively by gas chromatography-mass spectrometry (GC-MS) (HP5982A with multiple ion detector). A 38 m X 0.5 mm i.d. glass, OVlOl SCOT capillary column (Chromalytic Technology) with a split ratio of 1 0 1 using helium carrier and make-up gas was used isothermally at 260 "C. Quantification was carried out by measurement of peak area using the height and width of the peak at half-height; cutting and weighing and electronic integration (Hewlett-Packard 3390A), gave results identical within experimental error. The accuracy and reproducibility of the analytical method were established with standard PAH mixtures and authentic samples; the uncertainties in the reported PAH levels are estimated to be *lo% for the major and &20% for the minor components. A recovery of better than 90% for benzo[b]chrysene was obtained. The following components in the chromatograms were identified by comparison of their retention times and electron ionization (EI) mass spectra with reference compounds: fluoranthene, pyrene, benzo[a]anthracene, 4methylchrysene, benzov]fluoranthene, benzo[e]pyrene, benzo[a]pyrene, perylene, indeno[ 1,2,3-cd]pyrene,benzo[blchrysene, benzo[ghi]perylene, anthanthrene, coronene, and picene. Other components were identified by comparison of their relative retention times with literature values (7-11)and confirmed by comparison of their E1 mass spectra with literature data (12-14). Unambiguous assignment of PAH identities by GC-MS alone is not possible in all cases (12), and the uncertainties which re398

Environ. Scl. Technol., Vol. 19, No. 5, 1985

main in this work are indicated in Table I. The two sulfur-containing compounds, M , 234 (component 35 and part of 32) had E1 mass spectra, M:M 2 ratios, and GC retention times consistent with benzo[b]naphtho[2,1-d]thiopene and benzo[b]naphtho[2,3-d]thiophene, respectively; the former has been reported in APM (15).

+

Results and Discussion Identification of the sources of PAH adsorbed on air particulate matter has been attempted by a number of workers by using as source indicators either the ratio of selected PAH (16-18) or the PAH gas chromatographic profiles (19). The ratios most commonly used, benzo[a]pyrenejcoronene, benzo[a]pyrenejbenzo[ghi]perylene, and pyrenejbenzo[a]pyrene, all depend on the concentration of benzo[a]pyrene which has been shown in several studies (20-23) to be significantly less stable than the other PAH common in APM samples; it would seem that any source indicator which places heavy weight on benzo[a]pyrene should be avoided. Other PAH have been shown to be reactive in specific conditions (24-26), for example, under UV radiation when adsorbed on cellulose (24), and there are significant differences between the decomposition rates of several PAH. On the other hand, the long distance transport (27) of PAH and their presence in sediments far removed from possible sources (28) suggest that general PAH instability is not a major problem. Selective loss by evaporation during sampling would also distort the PAH composition, but this is probably significant only for mixtures which include two- and three-ring PAH, over long sampling periods (29). With these problems in mind, we restricted our analysis to PAH of four or more rings and decided to attempt source identification so as to minimize the contribution of any one component. To this end about 40 PAH were identified (Table I) and up to 26 quantified, all of four or more rings. A typical GC profile of these PAH in Christchurch APM is given in Figure 1. The resolution of some components (especially benzo[b]-, benzovl-, and benzo[k]fluoranthenes, which can be partially resolved at lower column temperatures) was sacrificed to give a faster analysis time; separate temperature-programmed analysis established that the carcinogenic [b] and isomers constitute about 70% of the total

v]

Table I. Polycyclic Aromatic Hydrocarbons Identified in the Christchurch Environment by Gas Chromatography and Gas Chromatography-Mass Spectrometry

peak no. M ,

abbreviation

PAH

1 228 benz[a]anthracene/cyclopenta[cd]pyrene

2 3 4 5 6 7 8

9 10 11 12 13 13a 13b 14 15 16 17 18 19 20 21 22

23 24 25 26 27 28 29 29a 30 31 32 33 34 35 36 37 38 39 40 41 42 43

228 252 252 252 252 266 266 266 264 264 278 276 278 278 276 276 290 292 292 292 292 302 302 302 300 300 300 300 302 302 202 202 218 202 216 228 226 226 234 242 242 242 240 240 256 268 276

chrysene benzofluoranthrenes ([b], GI, + [k] isomers) benzo[e]pyrene benzo[a]pyrene perylene methylbenzofluoranthene/methylbenzopyrene methylbenzofluoranthene/methylbenzopyrene methylbenzofluoranthene/methylbenzopyrene methylenebenzo[e]pyrene methylenebenzo[a]pyrene dibenz[aj]anthracene indeno[l,2,3-cd]pyrene dibenz[a,c]anthracene + dibenz[a,h]anthracene picene benzo[ghi]perylene anthanthrene methylbenzo[ghi]perylene, other methyl derivatives of PAH of MI 276 methyldibenzanthracene methyldibenzanthracene methyldibenzanthracene methyldibenzanthracene dibenzofluoranthene dibenzofluoranthene dibenzofluoranthene cyclopentabenzo[ghi]perylene cyclopenta[eflbenzo[ghi]perylene cyclopenta[bc] benzo[ghi]perylene coronene dibenzopyrene dibenzopyrene fluoranthene benzacenaphthylene ethylcyclopenta[deflphenanthrene pyrene benzofluorenes + methylpyrene/methylfluoranthenes benzo[c]phenanthrene(+M, 234, benzo[b]naphtho[2,1-d]thiophene) benzo[ghi]fluoranthene unknown benzonaphthothiophene methylbenzo[c]phenanthrene methylchrysene/methylbenz[a]anthracene methylchrysene/methylbenz[a]anthracene dibenzo[def,i]fluorene or 4H-cyclopenta[deflchrysene 4H-cyclopenta[defl triphenylene or 4H-benzo[fg]pyrene (+MI 254, binaphthyl) ethylchrysenes + ethylbenz[a]anthracene methylbinaphthyl dibenzo[b,ghi]fluoranthene,dibenzo[b,rnno]fluoranthene,dibenzo[a,ghi]fluoranthene, indeno[ 1,2,3-cd]fluoranthene,cyclopenta[cd]perylene, acenaphth[ 1,2-a]acenaphthylene + phenanthro[ 10,1,2,3-cdeflfluorene 278 dibenz[a,i]anthracene + pentacene

benzofluoranthene in any sample. Since Christchurch has only a very small level of industrial activity, we assumed that the two major sources of PAH in air pollution would be domestic fires and motor vehicles (mainly petrol engined). Samples of soot from the top section of domestic chimneys (DS) were used as reference samples of domestic PAH. Obtaining reference automobile samples proved more difficult. It is known that aerosol lead from vehicles burning leaded gasoline (as all New Zealand gasoline-engined vehicles do) rapidly settles so that the level of airborne lead falls rapidly with distance from the road (30). Sampling the vehicle exhausts directly, or from a 1.25-m3chamber connected to the exhaust by a 3-m pipe, gave lead levels in the collected particulate of up to 50% by weight, and such samples may well not be representative of vehicle-generated pollution after it has been distributed into the general pollution. For this reason, we used APM samples taken from the city’s three car park buildings (CPB); sampling was carried out over 2 h

BaA/CYC Chr BF BeP BaP Pe MeBF, MeBP MethBeP MethBaP DBajA IP BPe An MeBPe etc. MeDBA DBF DBF, CBPe CBPe co DBP

by using the high volume equipment, and the collected APM showed lead levels of 6-14%. The analytical results for the street sampling sites, car park buildings, and domestic soot samples are given in Tables 11-VI. These data show that Christchurch can suffer quite high PAH levels, although the conditions conducive to such high levels do not usually last for more than a week. Comparison of Tables V and VI shows many of the expected differences between domestic and automobile PAH. The methylbenzofluoranthenes and methylbenzopyrenes (components 7-9, Table I) are present in greater amounts in domestic soot than in the car park samples, whereas the methylenebenzopyrenes (components 10 and 11)are much more prominent in the CPB samples (10, 31). Again, the soot samples contain small amounts of dibenzofluoranthenes (components 21 and 22) and cyclopentabenzo[ghi]perylenes (components 24 and 25), while the car park buildings show only the perylenes, in much larger amounts. Cyclopenta[cd]pyrene (component 1,M, Environ. Sci. Technol., Vol. 19, No. 5, 1985

399

Table 11. PAH Concentrations over Selected 24-h Periods at Avonside ([APM] > 100 pg/m3)

date, daylmonth wind total APM, mg APM, rglm' PAH, ng/m3 peak no." 1 2

3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

817 calm 30.8 373 871

917 calm 9.6 107 184

1017 calm 24.0 275 584

1117 calm 12.2 132 271

1817 NE 13.2 141 349

2017 calm 15.9 183 440

2217 calm 21.0 238 535

2317 calm 24.4 301 799

89 55 105 28 71

9 7 26 9 19 4 4 9 8 7 7 9 11 4 3 9 2 4 3

37 26 76 22 57 15

16 9 35 14 23 6 6

19 10 46 17 29 7 7 14

61 27 48 17 36 9 9 23 14

62 35 74

88 56 105 41 66 14

individual PAH,b ng/m3 BaA Chr BF BeP BaP Pe MeBF, MeBP MeBF, MeBP MeBF, MeBP MethBeP DBajA IP BPe An MeBPe etc. MeDBA MeDBA MeDBA MeDBA DBF DBF DBF, CBPe CBPe CBPe co DBP

21

23 30 46 38 20 31 33 18 8 25 7 13 4 31 68 30 28 23 8 17

11

26 18 18 22 31 37 15 7 23 11 12 6 26 60 9 3

11

3 6 3 1

1

4 2

9 6

11

9 7 9 16 18 6 3

11

10

12 11

12

20 22

17 21 9 3 14 3 8 3

10

5 13 4 6 5 15 37 9 5 4 9 3

10

2 5 3 10 28 6 5 3 8 3

22

70 10 10 23 17 13 16 26 28

11

36 23 19 23 43 37

12

12 10 22 7 11 6 35 74 18 13 8 13 8

6 16 4 10 2

21

11

46 7 7 5 5 4

35 10 9 4 6 4

"PAH corresponding to peaks 11 and 28 (see Table I) were not detected. *See Table I for an explanation of abbreviations. Table 111. PAH Concentrations over Selected 24-h Periods at Bealey Avenue ([APM] > 100 pg/ma)

date, daylmonth wind total APM, mg APM, pg/m3 PAH, ng/m3 peak no." 1

2 3 4 5 6 7 8 9 10 12

13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

817 calm 17.9 316 569

917 calm 11.8 115 251

1017 calm 23.7 274 728

1117 calm 12.7 134 249

1817 NE 10.8 114 225

2217 calm 10.1 230 697

2317 calm 25.1 287 786

35 25 75 23 55 13

11

74 42 82 31 57 15

24 16 35 10 19 6 6 7 8 6

19 9 30 10 20 4 3 8 6 3 6

11

11

18 6 2

16 6

10 1

6

78 71 85 18 53 9 9 26 20 18 20 34 38 16 5 18

80 46 107 32 72 14 15 35 32 20 22 35 37 19 8 25

1

2

3

3

2 11

2

1

37 66 20 15 9 17 7

8

2 10 25 6 4 3 7 5

10 7 29 61 21

individual PAH,b ng/m3 BaA Chr BF BeP BaP Pe MeBF, MeBP MeBF, MeBP MeBF, MeBP MethBeP DBajA IP BPe An MeBPe etc. MeDBA MeDBA MeDBA MeDBA DBF DBF DBF, CBPe CBPe CBPe co DBP

12

27 22 15 15 27 35 12 3 18 5 18 4 25 56 11

16 7 10 5

13 37 14 25 3 3 6 10 7 8 14 17 2 2 11

3 6 4 9 29 8 2 1

4 2

11

24 22 17 22 35 49 24 6 24 5 15

11

21

7 3 2 5 3

2

12 11 18 8

OPAH corresponding to peaks 11 and 28 (see Table I) were not detected. bSee Table I for an explanation of abbreviations.

400

Environ. Sci. Technol., Vol. 19, No. 5, 1985

4 31 80 18 11

7 15 8

Table IV. PAH Concentrations over Selected 24-h Periods at Manchester Street ([APM]

date, day/month wind total APM, mg APM, wg/m3 PAH, ng/m3 peak noa 1 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

318

10.5 141 110

10.7 147 236

11.0 154 246

16 7 11 4 8 3 1 3 1 3 3 6 14 3 1 2

9 6 22 17 15 3 3 7 14 8 5 9 16 4 10 8 1 3 1 11 21 11 5 5 16 6

20 8 28 14 18 4 3 9 10 6 6 13 28 7 4 5 1 2 1 9 16 5 11 2 14 4

SW NE-.SW

718

1118

15.5 217 371

10.1 138 136

9.6 155 161

36 15 45 19 32 9 10 18 17 11 12 20 24 11 5 12 2 5 5 12 27

8 4 18 6 10 3 2 5 6 3 4 12 11 1 1 5 1

618

2717

418

SW-NE

calm

NE-SW

1418

> 100 pg/m3) 1618

calm

SW NE-SW

1718

1818

SW SW-NE

2218

2318

calm

NE

9.9 134 125

15.4 213 358

9.6 120 120

12.3 158 123

13.2 166 291

13.8 174 241

6 4 14 8 11 2 2 4 4 3 3 9 12 5 3 4 1 1

29 15 45 19 38 9 7 15 8 9 14 21 33 10 4 9 2 4 2 10 26 5 5 5 9 5

6 3 13 8 11 3 2 5 3 3 4 8 14 4

5 3 16 9 12 2 2 4 3 1 5 8 17 3 1 3 1 2 1 3 6 2 2 2 7 3

18 9 29 18 20 5 4 13 11 9 9 24 23 7 4 13 2 6 6 9 21 11 5 4 8 3

17 7 21 8 17 5 4 8

individual PAH,* ng/m3 BaA Chr BF BeP BaP Pe MeBF, MeBP MeBF, MeBP MeBF, MeBP MethBeP DBajA IP BPe An MeBPe etc. MeDBA MeDBA MeDBA MeDBA DBF DBF DBF, CBPe CBPe CBPe eo DBP

2 1 3 6 3 1 1 6 1

3 2 5 11 5 1 2 6 1

7 4 2 7 4

10 4 17 9 13 2 1 5 9 4 5 10 19 6 5 6 2 2 2 4 9 3 2 2 7 2

4 8 5 2 2 5 3

1 2 1 1 2 6 3 2 3 9 3

8 6 8 14 25 7 5 9 2 4 3 10 20 7 5 4 10 7

OPAH corresponding to peaks 11and 28 (see Table I) were not detected. bSee Table I for an explanation of abbreviations. Table V. PAH Concentrations in Domestic Open-Fire Soot

sample fuel

2

3

4

5

6

7

8

9

10

11

12

coal/wood

coal/coke

coal

demo1 wood

wood

demo1 wood

fencing wood

coal

coal

coal

coal

coal

99

249

644

330

113

108

324

588

18

1272

162

883

peak noa

PAH, ppm individual PAH,b ppm

1 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

BaA Chr BF BeP BaP Pe MeBF, MeBP MeBF, MeBP MeBF, MeBP MethBeP DBajA IP BPe An MeBPe etc. MeDBA MeDBA MeDBA MeDBA DBF DBF DBF, CBPe CBPe CBPe co DBP

7 13 21 6 11 3 4 3 1 2 5 8 5 1 1 1

40 22 37 13 22 5 5 8 3 7 7 14 7 3 2 6 1 1 1 6 20

97 187 117 28 30 12 14 12 8 6 13 21 15 3 5 6 2 4 4 12 20 11 8 4 2 3

53 72 81 18 41 7 5 4 2 4 8 12 9 4

15 21 20 7 8 4 4 3 4 2 3 5 4 1 1 2

12 20 28 8 10 4 3 3 2 3 2 5 2 1 1 1 1

45 76 52 15 31 6 5 5 11 4 7 12 7 2 4 5 1 2 2 7 13 4 3 2

101 152 88 19 41 11 9 20 7 8 12 16 10 4 5 10 2 7 5 15 27 6 5 2 2 4

2 2 6 2 3

96 64 186 39 242 13 14 38 52 16 15 77 29 12 38 103 29 11 4

7

49 44 63 26 58 14 19 28 51 12 34 27 24 9 29 103 42 15 8

1

1 2 1 1 1 1

5 4 2 4 2

1 2 2 2 1 1 1

1 1 1 4 1 1

1 1

I

2

1

1 1

151 8 6

7 21 5 13 1 2 4 4 3 4 9 14 4 6 31 6 2 13 4

trace trace

trace 10 19

174 5

2

18 31

aPAH corresponding to peak 11 (see Table I) was not detected. bSee Table I for an exulanation of abbreviations.

226) has been shown (32-34) to be an indicator of petrol-engined vehicle pollution, and our results are in

agreement with this; mass spectrometric analysis of component 1 showed little evidence of this PAH in the soot Environ. Sci. Technoi., Vol. 19, No. 5, 1985

401

Table VI. PAH Concentrations in Car Park Building (CPB) APM peak no."

individual PAH,* ng/m3

Lichfield Street CPB 1 2 3

1 2 3 4 5 6 7 8

BaA/CYCc Chr BF BeP BaP Pe MeBF, MeBP MeBF, MeBP MeBF, MeBP MethBeP MethBaP DBajA IP BPe An MeDBA CBPe CBPe co

56 22 19 23 6 1 2 2 12 5 12 31 98 17 5 11 5 47

9

10 11 12 13 14 15 19 24 25 26

1

PAH

374

2 4 3 1

4

Oxford Street CPB 5

68 21 19 24 10 1

5 6 5 7 2

10

9

12 4 10 29 86 31 8 9 7 64

4 2 4 11 26 6 1 3 3 16

58

403

101

1

6

4 1 4 12 28 9 2 4 3 23

138 34 28 34 18 2 2 1 12 5 19 36 124 46 9 7 3 48

127

566

8 6 9 3 1

1 2 1 3 6 20 2 2 1

Manchester Street CPB 7 8 9 143 41 30 50 11 3 1 2 25 24 58 153 59 15 13 12 117

195 44 43 57 16 1 2 2 27 9 30 70 213 82 22 18 21 156

18 13 9 12 3 2 1 1 13 1 6 17 58 10 5 3 3 28

766

1008

203

9

'PAH corresponding to peaks 16-18, 20-23, and 28 (see Table I) were not detected. bSee Table I for an explanation of abbreviations. Major component.

samples whereas it comprised almost the whole of this peak in the CPB samples. However, cyclopenta[cd]pyrene is particularly unstable and difficult to recover (34),and only such large differencesare likely to be significant. Coronene (component 26) is a major constituent of the car park samples (35)but negligible in soot, and benzo[ghi]perylene (component 14) is important in car park samples and almost absent in soot (36-38). However, these qualitative differences become of little value in assessing the sources of the mixed samples from Avonside, Bealey Avenue, and Manchester Street because the GC profiles from these sites are almost indistinguishable from each other or from the soot samples. We therefore undertook a multivariate statistical analysis of the data in Tables 11-VI. Data from the CPB and DS samples were analyzed first, and the results of this analysis were applied to the remaining samples. Inspection of the data for DS (Table V) and CPB (Table VI) samples shows that several individual PAH were either not detected or present in very small amounts. Any such PAH baving a level