Quantitative Characterization of PAHs in Burn Residue and Soot

Aug 12, 1999 - Janne Fritt-Rasmussen , Birgit Elkjær Ascanius , Per Johan Brandvik , Arne Villumsen , Erling H. Stenby. Marine Pollution Bulletin 201...
0 downloads 0 Views 149KB Size
Download PDF
Environ. Sci. Technol. 1999, 33, 3100-3109

Quantitative Characterization of PAHs in Burn Residue and Soot Samples and Differentiation of Pyrogenic PAHs from Petrogenic PAHs-The 1994 Mobile Burn Study Z H E N D I W A N G , * ,‡ M E R V F I N G A S , ‡ Y . Y . S H U , †,‡ L I S E S I G O U I N , ‡ MICHAEL LANDRIAULT,‡ AND PAT LAMBERT‡ Emergencies Science Division, ETC, Environment Canada, Ottawa, Ontario, Canada K1A 0H3 ROD TURPIN§ AND PHIL CAMPAGNA§ Environmental Response Team, U.S. EPA, Edison, New Jersey 08837 JOSEPH MULLIN| Minerals Management Service, U.S. Department of The Interior, Herndon, Virginia 22070-4817

Several mesoscale burns were conducted in 1994 in Mobile Bay, AL, to study various aspects of diesel fuel burning in situ. The target PAHs in the diesel, residue, and soot samples collected during each burn were quantitatively characterized by GC/MS. A simple model based on mass balance of individual petroleum PAHs pre- and postburn was proposed to estimate the destruction efficiencies of the total petroleum PAHs. This study demonstrates the following: (1) Distributions of PAHs in the original diesel and soot were very different. (2) The average destruction efficiencies for the total target diesel PAHs including five alkylated PAH series and other EPA priority unsubstituted PAHs were greater than 99%. (3) Using the model, 27.3 kg of the diesel PAHs were destroyed for each 1000 kg of diesel burned. These were mostly two- and three-ring PAHs and their alkylated homologues. (4) Combustion also generated trace amounts of high molecular weight fiveand six-ring PAHs as well as the four-ring benz[a]anthracene. But the total mass of these pyrogenic PAHs was found to be extremely low: only 0.016, 0.032, and 0.048 kg of the five- and six-ring PAHs were generated by combustion in the three different scenarios for each 1000 kg of diesel burned. From these points, we conclude that in situ burning is an effective measure to minimize the impact of an oil spill on the environment, greatly reducing exposure of ecosystems to the PAHs of spilled oils. A new “pyrogenic index”, Σ(other three- to six-ring PAHs)/Σ(five alkylated PAHs), is proposed (see the text for the definition) as a quantitative indicator for identification of pyrogenic PAHs and for differentiating pyrogenic and petrogenic PAHs. Also, this index is demonstrated to be a useful tool for distinguishing heavy fuels from crude oils and light refined products. This method, combined with other criteria, is expected to be applicable to such situations as oil spill 3100

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

investigations, site assessment, and apportioning of legal responsibility for pollution cleanup.

Introduction In situ burning of spilled oil as an oil spill cleanup countermeasure is gaining acceptance because of its distinct advantages over other countermeasures (1, 2). This technique can convert large quantities of oil into its primary combustion products, carbon dioxide and water. It requires less equipment and labor and can be applied in areas where many other methods cannot due to lack of response infrastructure and/or lack of alternatives. Efficiency of oil removal from water was estimated to be greater than 99% based on the results of the 1993 large-scale Newfoundland Offshore Burn Experiments (NOBE) (3). In addition, because oil burn products, mainly smoke particulates and burn residues, are only a small percentage of the spilled oil, the need to collect, store, and transport recovered fluid can be reduced to a minimum compared to those from the conventional methods. In situ combustion on water is a complex process, which is affected by many factors and surrounding dynamic conditions (4). In the last 15 years, numerous lab and tank tests and offshore burning experiments, from small to large scale, have been performed to study various aspects (such as type of oil, ignition technique, burning rate, oil thickness, and wind direction and speed) of in situ burning of crude oils and refined products (3-6). In 1994, several U.S. and Canadian government agencies jointly conducted a series of mesoscale diesel burns in Mobile Bay, AL, to study a variety of parameters that might affect diesel burning, emissions, and smoke products. The 1994 burns were conducted in a specially constructed steel pan (15.2 × 15.2 m) with an outer berm filled with water. In each test, 17.1 m3 of diesel was released and floated on 0.6 m of saltwater pumped from Mobile Bay. The fuel was ignited, and the burn lasted about 25 min. Environment Canada, in collaboration with the U.S. Environmental Protection Agency (EPA) and the U.S. Coast Guard, set up a series of instruments and samplers to monitor all suspect emissions (1) including CO2, CO, and SO2, volatile organic compounds (VOC), carbonyls, polycyclic aromatic hydrocarbons (PAHs), dioxins and dibenzofurans, metals, and particulates. In addition to ground station samplers and airborne samplers, the National Institute of Standards and Technology (NIST) developed a new smoke sampling package deployed from a helicopter to rapidly collect data on smoke and/or emissions from the burns (2). The general analytical results of these trials have been reported (1). A relatively high temperature was reached during in situ combustion, and the diesel fuel was nearly complete burned. In situ oil burning produces a visible smoke plume containing smoke particulate and other combustion products which may persist over several kilometers downwind from the burn. This fact gives rise to public health concerns related to the chemical content of the smoke, in particular the levels of PAHs in the smoke. Since the late 1980s, a number of research projects have been carried out in the field of emis* Corresponding author phone: (613)990-1597; fax: (613)-9919485; e-mail: [email protected]. † Present address: Department of Chemistry, National Kaohsiung Normal University, Taiwan. ‡ETC, Environment Canada. § U.S. EPA. | U.S. Department of The Interior. 10.1021/es990031y CCC: $18.00

 1999 American Chemical Society Published on Web 08/12/1999

sions from in situ oil combustion (3-6). However, there are no quantitative data on the distribution of the petroleumcharacteristic alkylated PAHs of primary concern (that is, the alkylated naphthalene, phenanthrene, dibenzothiophene, fluorene, and chrysene series) and other EPA priority unsubstituted PAHs in the burn residue and in the smoke, on the relative amount of the total PAHs in the oil versus the amount in the smoke particulates, and on the total PAHs destroyed and new PAHs generated by combustion. This information is extremely important for spill response staff and decision-makers to determine the most suitable countermeasures after an oil spill and for assessment of the impact of in situ burn products to the environment. This paper focuses on quantitative characterization and comparison of the composition and distribution of PAHs in the diesel, residue and soot, destruction and formation of PAHs, estimation of the destruction efficiencies of the total diesel PAHs, and development of quantitative differentiation criteria to distinguish pyrogenic PAHs from petrogenic PAHs.

Experimental Section Diesel, Burn Residue, Water, and Soot Samples. The diesel was collected from the barge which delivered the fuel to the site prior to each burn. The fuel used for burns was no. 2 diesel obtained from a commercial supplier in Mobile, AL. Water samples were collected 10 cm below the surface, before and after burns, using precleaned 1 L amber bottles. All water samples were collected without headspace, placed in refrigerated coolers, and sent to laboratory for analysis. Residue samples were collected in new, clean, 250 mL wide-mouth glass jars with a Teflon lined cap (Fisher Scientific, Nepean, ON). All residue samples were collected manually by skimming the residue from the surface of the tank water. Anderson (Smyrna, GA) PM-10 high-volume sampler and the Anderson TSP high-volume sampler were used to collect smoke particles of less than or equal to 10 µm in size and the total suspended particles from 0.3 to 50 µm from the burns, respectively. Both samplers meet U.S. EPA specifications. A flow rate of 1.1-1.7 m3/min was employed. The sampling media consisted of quartz fiber filters in 8 × 10" sheets. The inside of each sampler was rinsed with hexane prior to use. After sampling was completed, the filters were wrapped in aluminum foil, placed in an envelope, and refrigerated. All filters were weighed on a precision balance before and after the burn to obtain the weight of the soot particulate. Sample Extraction and Cleanup. The microwave extraction method (7) was used to extract PAHs and other hydrocarbons from the soot samples. Three to six pairs of 36 mm (ID) PM-10 or TSP filter disks, cut from the whole PM-10 or TSP filter sheets (8 × 10"), were placed into the Teflon vessels of the MES-1000 microwave extraction system (CEM Corp, Mathews, NC) and spiked with 100 µL of deuterated surrogate mixture containing 1 µg each of four deuterated PAHs (acenaphthene-d10, phenanthrene-d10, benz[a]anthracene-d10, and perylene-d12). Sixty milliliters of a methanol/ benzene/hexane solvent mixture (2:1:9) was added to the vessels, enough to cover the filter. The vessels were closed and placed on the sample carousel inside the microwave oven for extraction. The temperature was increased to 100 °C using the preset temperature program, at which temperature the extraction was maintained for 10 min. Then the vessels were cooled to ambient room temperature. Satisfactory recoveries of PAHs from filters by using microwave extraction have been demonstrated and reported (7). The extracts were dried by filtering through anhydrous sodium sulfate and concentrated to approximately 1-2 mL by rotary evaporation. The concentrated extracts were quantitatively transferred to a preconditioned 1.5 g silica gel microcolumn topped with 1 cm anhydrous sodium sulfate for sample cleanup. Ten milliliters of a benzene/hexane mixture (1:1)

was used to elute the saturate and aromatic hydrocarbons. The eluent was then concentrated under a stream of nitrogen, spiked with internal standards terphenyl-d14, and made up to the preinjection volume (0.50 mL) for GC analysis. Aliquots of 500-1000 mL of water samples were spiked with a mixture of PAH surrogate standards and successively extracted three times with dichloromethane. The combined raw extract was concentrated, spiked with internal standard, and made up to the preinjection volume (8-10). The diesel, preburn fuel, and postburn residue samples were directly dissolved in hexane at a concentration of ∼100 mg/mL. Aliquots of the samples (∼15 mg of oil) were spiked with PAH surrogates and fractionated into saturate and aromatic fractions using a silica gel microcolumn fractionation technique (8-10). The aromatic fractions were then spiked with internal standards for GC analysis. Gas Chromatography/Mass Spectrometry (GC/MS) Analysis. For the GC/MS instrumentation, quality control, determination of relative response factors of target PAHs from authentic standards, and quantitation of the target diesel PAHs, refer to refs 8-10. To achieve improved analytical precision and accuracy for the water and soot samples which may only contain traces of PAHs, some refinements (such as more frequent and rigorous calibration checkup of instrument performance, manual integration of PAHs having low abundances, manually setting the baselines for integration of alkylated PAH groups, and using one GC/MS through the entire program) were implemented in addition to the routine quality control measures.

Results and Discussion Fingerprints of Target PAHs in Diesel, Burn Residue, and Soot Samples. Table 1 summarizes the quantitation results of target PAHs in the diesel, burn residue, and soot samples. The suite of target PAHs (11, 12) was expanded to include both the five target petroleum-characteristic alkylated PAH homologous series and the other 15 EPA priority unsubstituted PAHs (Table 1). Figure 1 depicts representative fingerprints and distinguishing features of target PAHs for the diesel, residue, and soot samples. The PAHs in the water samples were found to be extremely low, and most congeners are below the detection limit (1). Analyses of volatile compounds demonstrated that there were basically no volatile compounds remaining in the water after a burn. Compared to most crude oils, the Mobile diesel is characterized by a narrower distribution of n-alkanes (n-C8 to n-C27) but with much higher concentrations of the total n-alkanes (∼165 mg/g oil) than most crude oils. As well as containing relatively smaller quantities of BTEX (0.14, 0.36, 0.32, and 3.62 mg/g oil for benzene, toluene, ethylbenzene, and the xylene isomers, respectively) and C3-benzenes (7.53 mg/g oil), the aromatic fraction of the diesel and preburn samples contains mainly alkylated naphthalene, dibenzothiophene, fluorene, and phenanthrene homologues (56, 19, 14, and 11% of the total of five-target PAH homologues, Table 1). The alkylated chrysene series were the least abundant (only 25 µg/g oil, ∼0.1% of the total alkylated PAHs), resulting in the relative ratios of chrysene series to other four PAH series approaching zero. Among the other EPA priority PAHs, the dominance of two- and three-ring PAHs over the fourand five-ring PAHs is apparent (Tables 1 and 2). The concentrations of five- and six-ring PAHs are extremely low, and indeno(1,2,3-c,d)pyrene, dibenz[a,h]anthracene, and benzo[ghi]perylene were not detected. It was observed that “virtually all of the fuel was consumed by burning” (2), and the residue was estimated to be no more than 0.1-0.3% of the original diesel by weight. The residue was a mixture of burn products and some unburned oil. Therefore, the residue showed a mixed PAH fingerprint that retained some characteristics of the diesel oil but had VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3101

TABLE 1. PAH Quantitation Results and Diagnostic Ratios of Paired PAH Isomers and Alkylated PAH Homologues Series 94-mobile burn target PAHs

94-diesel

soot sample (µg/g particulate) 94-preburn 94-residue (n ) 3) (n ) 3) PM10-B1b PM10-B2 PM10-B3 TSP-B1 TSP-B2

TSP-B3

Concentrations (µg/g) five target alkylated PAHs naphthalenes (C0 to C4) phenanthrenes (C0 to C4) dibenzothiophenes (C0 to C3) fluorenes (C0 to C3) chrysenes (C0 to C3) sum of five target PAHs other EPA priority PAHsa biphenyl (two-ring) sum of three- to six-ring PAHs sum of five- to six-ring PAHs total of target PAHs

15146 3000 5020 3902 25 27093

16359 3222 5424 4096 28 29129

5929 5014 6870 3819 318 21950

110 90 25 12 27 264

33 41 18 5 24 121

30 148 85 10 29 302

120 78 27 22 28 275

34 99 45 17 20 215

26 109 56 7 35 233

309 107 0.97 27509

327 107 0.85 29563

63 347 92 22360

6 440 414 710

2 333 291 456

2 334 229 638

10 466 433 751

2 316 241 533

2 395 302 630

chrys/naphs chrys/phens chrys/dibens chrys/fluos C2D/C2P:C3D/C3P

0.00 0.01 0.01 0.01 1.89:1.72

0.99 0.20 0.34 2.82 0.68:0.76

0.23 0.36 1.04 1.25 0.54:0.55

0.58 0.20 0.44 1.20 0.60:0.67

1.38 0.33 0.63 4.85 0.68:0.78

36.59 6.85 0.004

Diagnostic Ratios 0.04 0.24 0.05 0.30 0.04 1.06 0.06 2.20 1.66:1.44 0.51:0.44 1.64:1.51 1.73:1.52 10.31 7.73 3.59 1.59 0.014 0.81

0.73 0.58 1.36 4.71 0.74:0.75

phen/An chry/BaA Σ(other three- to six-ring PAHs)/ Σ(five PAH series)

0.00 0.01 0.00 0.01 1.85:1.69c 1.91:1.72 1.89:1.73 32.40 6.50 0.004

4.90 1.22 1.94

4.53 1.29 0.86

7.19 2.10 0.81

3.92 1.31 1.05

4.29 1.27 1.28

a The other 15 EPA priority unsubstituted PAHs include the following: biphenyl (Bph, two-ring), acenaphthylene (Acl, three-ring), acenaphthene (Ace, three-ring), anthracene (An, three-ring), fluoranthene (Fl, four-ring), pyrene (Py, four-ring), benz[a]anthracene (BaA, four-ring), benzo[b] fluoranthene (BbF, five-ring), benzo[k]fluoranthene (BkF, five-ring), benzo[e]pyrene (BeP, five-ring), benzo[a]pyrene (BaP, five-ring), perylene (Pe, five-ring), indeno[1,2,3-c,d]pyrene (IP, six-ring), dibenz[a,h]anthracene (DA, five-ring), and benzo[ghi]perylene (BP, six-ring). b B1, B2, and B3 represent 1994 Mobile burn 1, burn 2, and burn 3, respectively c Values of C2D/C2P:C3D/C3P for three preburn and residue samples

some characteristics of additional pyrogenic PAHs. The most noticeable change in the aromatic hydrocarbon distribution was the complete loss of alkylbenzene compounds and dramatic decrease in abundances of naphthalene and its alkylated homologues relative to other PAH series. This feature is very similar to that of the highly weathered oils (10, 13-16). However, there are several other compositional changes of PAHs which are significantly different from that caused by weathering or biodegradation. First, the concentration of chrysene and its alkylated homologues increased 10-14 times in the residue samples, resulting in significant increase in the relative ratios of chrysene series to the other four alkylated PAH series (Table 1). In contrast, the concentrations of the chrysene series in the long-term weathered Arrow (10) and BIOS (13) oil samples and highly labweathered ASMB oil (45% weathered) (17) were only altered by a factor of 0.8-1.5 of the corresponding source oils. Second, additional high-number-ring unsubstituted PAHs, including indeno[1,2,3-c,d]pyrene, dibenz[a,h]anthracene, and benzo[ghi]perylene, were detected. The increase in abundances of four- to six-ring PAHs plus three-ring anthracene and their dominance over lower molecular weight two- and three-ring biphenyl, acenaphthylene, and acenaphthene was pronounced. This fact implies that these high molecular weight PAHs may be largely formed from combustion, even though their absolute amount was very small (Tables 1 and 2). Compared to the diesel and burn residues, alteration in the PAH distribution of the smoke particulates was extremely striking. First, the other EPA priority unsubstituted three- to six-ring PAHs were significantly more abundant than the five alkylated PAH homologues in the six soot samples (Table 1). Second, the relative distribution of the alkylated chrysene series in the total of the five target alkylated PAH homologues jumped to 10-20% for soot samples from 0.1% in the starting 3102

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

diesel. Also, the dominance of parent chrysene over its alkylsubstituted homologues and the decrease in relative abundances with increasing level of alkylation in the chrysene series (that is, in the order of C0- > C1- > C2- > C3-) is very pronounced (Figure 1). This is a typical characteristic of pyrogenic PAHs generated in combustion of organic materials including wood, coal, and refined petroleum products (16, 18-21). In contrast, petroleum alkylated PAH homologous series including the alkylated chrysenes show the characteristic bell-like distribution profiles, which are readily modified to the distribution profile of C0- < C1- < C2- < C3- by weathering or degradation. However, it is noted that such distribution profiles of C0- > C1- > C2- > C3- were not obvious for the other four two- and three-ring alkylated PAH series, which indicates that these lower molecular weight PAHs in the smoke particulates may largely come from volatilized and uncombusted diesel PAHs, rather than newly generated. Third, the high molecular weight four- to six-ring parent PAHs plus three-ring anthracene are remarkably abundant relative to the other lower molecular weight PAHs. This results in significant decrease in relative ratios of two pairs of isomers, phenanthrene to anthracene (MW ) 178) and chrysene to benz[a]anthracene (MW ) 228), to 4-7 and 1-2 from 37 and 7, respectively (Table 1). These PAHs were mostly generated from incomplete combustion of the diesel through thermal aromatization of free-radicals pyrolyzed from precursor hydrocarbons (18, 21-23). As an example, Figure 2 compares extracted ion chromatograms at m/z 178, 228, 252, and 276 for the diesel, residue, and soot samples, clearly illustrating the changes in the relative distribution of target PAHs and demonstrating the formation of pyrogenic PAHs from three-ring anthracene to six-ring benzo[ghi]perylene by combustion. For comparison purposes, the values of the double ratio of C2D/C2P:C3D/C3P are also presented in Table 1. This

FIGURE 1. Representative PAH fingerprints and distinguishing features of different distribution of PAHs for the diesel residue and soot samples. N, P, D, F, and C represent naphthalene, phenanthrene, dibenzothiophene, fluorene, and chrysene, respectively; 0-4 represent carbon number of alkyl groups in the alkylated PAH homologous series. The abbreviations from Bph to BP represent the other 15 EPA priority unsubstituted PAHs (refer to Table 1 for the full names of these PAHs). For comparison, the fingerprints of the other three- to six-ring PAHs have been enlarged and shown in the left insets. Note that, for clarity, different Y-axis scales are applied to the fingerprints of the soot samples. ratio, as an important fingerprinting criterion, has been heavily used to correlate the 1989 EXXON VALDEZ spilled oil sediment samples to the source oil (11, 19, 20). However, unlike the EXXON VALDEZ spilled oil, the double ratio determined in the Mobile Burn Experiments demonstrated the order of preburn samples > postburn residue . soot

samples (Table 1 and Figure 3). This fact indicate that the burn situation is greatly different from the slow natural weathering process. In the slow natural weathering process, weathering may occur at comparable rates for each level of alkylation of dibenzothiophene and phenanthrene series. But under the long-term heavy weathering/degradation VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3103

TABLE 2. Estimation of Destruction Efficiencies of Target PAHs for 1994 Mobile Burn Diesel target PAHs

Ar ring nos.

diesel oil (Cf,PAH) (µg/g)

residues (Cr,PAH) (µg/g, n ) 3)

soota (Cp,PAH) (µg/g, n ) 6)

EPAHb (%) (scenario A)

EPAHb (%) (scenario B)

EPAHb (%) (scenario C)

Alkylated PAHs naphthalene C0-N C1-N C2-N C3-N C4-N sum phenanthrene C0-P C1-P C2-P C3-P C4-P sum dibenzothiophene C0-D C1-D C2-D C3-D sum fluorene C0-F C1-F C2-F C3-F sum chrysene C0-C C1-C C2-C C3-C sum total biphenyl (Bph) acenaphthylene (Acl) acenaphthene (Ace) anthracene (An) fluoranthene (Fl) pyrene (Py) benz[a]anthracene (BaA) benzo[b]fluoranthene (BbF) benzo[k]fluoranthene (BkF) benzo[e]pyrene (BeP) benzo[a]pyrene (BaP) perylene (Pe) indeno[1,2,3-c,d]pyrene (IP) dibenz[a,h]anthracene (DA) benzo[ghi]perylene (BP) total of three- to six-ring PAHs total of five- to six-ring PAHs total of target PAHs

2 2 2 2 2

232.1 1025.2 4599.2 6305.4 2984.1 15146

42.1 111.2 1109.8 2745.4 1920.4 5929

10.4 8.7 20.7 12.5 6.5 59

99.8 99.9 99.9 99.9 99.9 99.9

99.1 99.9 99.9 99.9 99.9 99.9

98.8 99.8 99.8 99.8 99.8 99.8

3 3 3 3 3

253.6 909.6 1066.8 570.5 199.3 3000

339.9 1133.5 1741.2 1196.5 602.8 5014

11.5 12.0 28.8 28.6 13.3 94

99.6 99.8 99.7 99.5 99.4 99.7

99.3 99.5 99.4 99.1 98.7 99.3

98.9 99.3 99.1 98.6 98.0 99.0

3 3 3 3

511.4 1507.1 2019.6 982.2 5020

460.8 1691.7 2934.6 1783.2 6870

1.4 3.4 18.1 19.9 43

99.9 99.9 99.8 99.7 99.8

99.8 99.7 99.6 99.4 99.6

99.7 99.5 99.4 99.1 99.4

3 3 3 3

241.4 1002.3 1421.7 1236.9 3902

144.0 802.1 1374.3 1498.3 3819

1.9 1.7 2.7 6.0 12

99.9 99.9 99.9 99.8 99.9

99.8 99.8 99.8 99.7 99.8

99.7 99.7 99.7 99.6 99.7

4 4 4 4

6.8 9.1 7.5 1.9 25 27094

48.4 90.4 122.4 57.1 318 21950

15.3 5.8 4.1 2.1 27 235

88.0 96.0 96.0 91.0 93.0 99.8

76.0 90.0 91.0 83.0 86.0 99.7

64.0 85.0 87.0 74.0 80.0 99.6

2 3 3 3 4 4 4 5 5 5 5 5 6 5 6

309.2 14.3 74.1 6.9 2.6 7.3 1.0 0.2 0.1 0.4 0.1 0.2 NDc NDc NDc 107 1.0 27510

Other EPA Priority PAHs 62.9 62.7 30.0 35.2 44.2 69.0 14.8 9.3 12.1 13.1 17.1 4.7 14.5 1.3 19.4 347 92 22360

4.0 2.8 0.5 2.2 21.1 24.3 11.0 29.0 73.4 34.0 35.6 9.0 54.6 6.9 76.0 380 319 620

99.9 99.0 99.9 97.9 57.0 82.0 43.0 (7×)d (28×) (4×) (36×) (3×) e e e 82 (16× by wt) 99.8

99.8 97.9 99.8 95.8 15.0 65.0 (1.1×) (14×) (56×) (8×) (72×) (6×) e e e 64 (32× by wt) 99.6

99.7 96.8 99.7 93.7 (1.2×) 47.0 (1.7×) (21×) (85×) (12×) (108×) (9×) e e e 46 (48× by wt) 99.4

a The concentration of target PAHs were averaged from six soot samples. b Three scenarios were chosen to estimate the destruction efficiencies of PAHs using eq 5: scenario A, soot ) 5%, residue ) 0.1%; scenario B, soot ) 10%, residue ) 0.2%, and scenario C, soot ) 15%, residue ) 0.3% of the original diesel by weight. c ND: under the detection limit. d Eight five- and six-ring PAHs were largely generated by combustion of the diesel. e Newly generated.

conditions (10, 13) or the burn situation, the loss rate of sulfur-containing dibenzothiophenes was obviously greater than the loss rate of the phenanthrene series. The plot of double ratios in Figure 3 can be best described by a linear equation “Y ) 0.8292X + 0.1345 (r 2 ) 0.99)’’. This plot may be useful for tracing the relative distribution of PAHs of the diesel burn products in the future experiments. Estimation of Destruction Efficiencies of the Diesel PAHs. The PAH destruction efficiency estimates are based on differences of initial and remaining PAHs in the burn residue and soot. Therefore, the best estimation of destruction efficiencies of PAHs in the in situ burn diesel largely depends 3104

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

on “true values” of soot produced. Since the late 1980s, several groups have published data in the literature on soot production from in situ oil fires (3-6). However, these data do not agree and often vary from each other by as much as an order of magnitude. A common approach used in the past to estimate the soot production is the carbon balance method. The major assumption of this method is that all carbons resulting from the fire are in the smoke plume. The carbon balance method may be somewhat applicable for very small burns where sampling is performed directly in the chimney and where the gases do not have a chance to escape from the smoke. However, in situ observations and extensive data

FIGURE 2. Comparison of extracted ion chromatograms at m/z 178, 228, 252, and 276 for the diesel (preburn MB-17), burn residue (postburn MB-16), and soot sample (TSP-B3), illustrating changes in the relative distribution of unsubstituted PAH isomers and demonstrating the formation of pyrogenic PAHs from three-ring anthracene (An) to six-ring indeno[1,2,3-c,d]pyrene (IP) and benzo[ghi]perylene (BP) due to combustion. Refer to Table 1 for the full names of these EPA priority unsubstituted PAHs shown in Figure 2. Note that, for clarity, different Y-axis scales are applied to the extracted ion chromatograms. obtained from a three-dimensional array of dozens of carbon dioxide meters, sampling stations, and helicopter sampling packages from the 1993 NOBE experiments (3, 24) and from the 1994 Mobile Burn Experiments (1, 25) demonstrated that there was a significant gas separation between the plume and the surface, and the ground concentrations of carbon dioxide were always greater than those found in the air and in the plume. Therefore, a single carbon balance measurement in the plume may overestimate soot production of in situ burns. Very recently, two new methods of estimating soot production, integration of the soot concentration by volume under the plume and integration of soot deposition weight over the area under the plume, have been proposed (25). The NOBE and Mobile experimental data were fed into

the model to estimate the soot production. For the Mobile diesel, the average value of soot production obtained was 8.6% (25). Walton and co-workers (26) reported the soot production yields of 9.3-13.7% of the mass of fuel burned estimated by the carbon balance method. The soot samples were collected by using a newly designed helicoptertransported sampling package (26). It should be understood that current measurement and estimation techniques are still fraught with numerous difficulties; therefore, “true experimental values” may not be forthcoming for a period of time. The destruction efficiency of the individual PAHs, EPAH, is defined as the mass of target PAHs destroyed in burns, mB,PAH, over the mass of the same PAHs in the starting fuel, VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3105

FIGURE 3. Double ratio plot of C2D/C2P to C3D/C3P in the diesel, residue, and soot samples. The plot shows the double ratio in the order of the starting oil and preburn samples > postburn residue . soot samples, demonstrating the loss rate of alkylated dibenzothiophene series was significantly greater than the loss rate of alkylated phenanthrene series under burn situation. The plot of double ratios can be best described by a linear equation of Y ) 0.8292X + 0.1345. mf,PAH. The amount of PAHs burned, which is equal to the difference between the mass of target PAHs in the starting oil, mf,PAH, and in the smoke particulate, mp,PAH, and in the residue, mr,PAH, is

EPAH ) mB,PAH/mf,PAH ) (mf,PAH - mp,PAH - mr,PAH)/mf,PAH ) 1- mP,PAH/mf,PAH - mr,PAH/mf,PAH

(1)

Two assumptions are made in the calculation. The first is that the target PAHs in the smoke particulate and in the residue are predominant ones during burns. The gas analysis results from both the 1991 (27) and 1994 (1) Mobile in situ burning experiments have demonstrated that volatility losses of unburned PAHs were negligible. The second assumption is that the samples were collected over a suitable time period to average out natural fluctuations in the fire and plume; therefore, the PAHs in the samples represent the averaged “true” PAH distribution. The mass of target PAHs in smoke particulate is equal to the concentrations of target PAHs in soot samples, CP,PAH, multiplying the mass of the smoke particulate, which can be readily obtained by the fuel mass (mf) multiplied by the soot production yield (YS ) mp/mf). The mass of the PAHs in the starting fuel can be simply obtained by the concentrations of the target PAHs in the fuel, Cf,PAH, multiplying the fuel mass mf.

mp,PAH ) mp × CP,PAH ) mf × YS × CP,PAH

(2)

mf,PAH ) mf × Cf,PAH

(3)

The mass of the PAHs in the residue can be readily obtained by the same method.

mr,PAH ) mr × Cr,PAH

(4)

Combining eqs 1-4 yields

EPAH ) 1 - (mf × YS × CP,PAH)/(mf × Cf,PAH) (mr × Cr,PAH)/(mf × Cf,PAH) ) 1 - (CP,PAH/Cf,PAH) × YS - Cr,PAH/Cf,PAH × mr/mf (5) Because there were many variables which could affect soot and residue production and collection, three scenarios with a range of yield from 5% to 15% of soot production were 3106

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

chosen to estimate destruction efficiencies of target PAHs in the diesel. The calculated results using eq 5 are presented in Table 2. The three scenarios used were as follows: scenario A with soot production of 5% and residue production of 0.1%, scenario B with soot production of 10% and residue production of 0.2%, and scenario C with soot production of 15% and residue production of 0.3% to the starting fuel by weight. Compared with the published estimated data for soot production yields of 8.6-13.7%, these three scenarios can be considered to represent a reasonable range for the diesel soot and residue production under the present burn conditions. From Table 2, the following observations are apparent: (1) The destruction efficiencies of PAH congeners differ greatly. (2) The total destruction efficiencies of target diesel PAHs including five alkylated PAH series and other EPA priority PAHs (27 510 µg per gram of oil) were greater than 99% for all three scenarios, primarily due to the high destruction efficiencies (>99%) of the predominant two- and three- ring alkylated PAH series. (3) Among the five alkylated PAH series, the four-ring chrysene series had the lowest net destruction efficiency (80-93%). (4) The total destruction efficiencies of the other three- to six-ring PAHs were determined to be only 82, 64, and 46% for scenarios A, B, and C, respectively. (5) The individual destruction efficiencies of the other three- to six-ring PAHs dramatically decreased with increasing ring number. For example, the destruction efficiency of three three-ring PAHs was around 98% at scenario B but rapidly decreased to 15% and 65% for four-ring fluoranthene and pyrene. (6) Eight five- and six-ring unsubstituted PAHs from benzo[a]anthracene to benzo[ghi]perylene plus four-ring benz[a]anthracene were largely generated from combustion of diesel. The mass of these PAHs in the soot were estimated to be 3-36 times, 6-70 times, and 9-108 times of that in the starting fuel for scenarios A, B, and C, respectively. It is important to note, however, that even though these high ring PAHs were in much higher abundance in the burn products than in the starting fuel, the total mass of these PAHs generated by combustion was very small. By estimation using our model, 27.3 kg of the diesel PAHs, mainly two- and three-ring PAHs and their alkylated homologues, were destroyed when 1000 kg of diesel was burned. At the same time only 0.016, 0.032, and 0.048 kg of the five- and six-ring PAHs were generated during burning for the three different scenarios. Therefore, these results should not be simply misinterpreted as “burning makes PAHs”.

TABLE 3. PAH Quantitation Results and Pyrogenic Index Values for over 60 Oils and Refined Products, Artificially Weathered Oil Series, Highly Weathered Spilled Oil Samples, Biodegraded Oil Samples, and NOBE Burn Samples

oil and refined products oils (40 oils) lube oil heavy fuels (nine fuels) Bunker C (4 Bunker C) oil contaminated birds (1995) tarballs (BC and CA, 1996) tarballs (NF, 1997) artificially weathered oils ASMB oil series (0-45%) California oil series (0-15%) 25-year-old Nipisi spill samples biodegraded oils ASMB series 3 North Slope series 3 Cook Inlet series jet fuel B series diesel no. 2 series Bunker C/diesel mixture series 1993 NOBE burn samples starting oil and preburn samples residues 1994 Mobile burn samples starting diesel and preburn samples residues soots

sum of alkylated PAH series (µg/g)

sum of other three- to six-ring PAHsa (µg/g)

4000-45000 352 2000-33000 12000-32000 9700-15500 3800-6300 12500-14700

25-160 0 30-700 550-860 300-540 80-130 380-430

7-20 11-25 11-19 13-18 7-10

1 to identify contamination sources of combustion processes. In this study, we found that the relative ratios of the sum of the concentrations of the other EPA priority unsubstituted three- to six-ring PAHs (see Tables 1 and 2 for full names of these PAHs) to the sum of the concentrations of the five target alkylated PAH homologues, Σ(other three- to six-ring PAHs)/Σ(five alkylated PAHs), for the soot samples were significantly different from that for crude oils and refined products and can be positively used to differentiate the pyrogenic and petrogenic PAHs. We define this ratio as “pyrogenic index”. The pyrogenic index values were determined to be 0.004 and 0.009-0.019 for the diesel and residues, but in a range of 0.8-2.0 for the six soot samples. Table 3 summarizes ranges of the PAH quantitation results and the pyrogenic index values for various oil-related samples. For comparison purposes, the ratios of phenanthrene/ anthracene are also listed in Table 3. Compared to other diagnostic ratios obtained from individual compounds, this index ratio has its own distinct advantages: (1) As discussed above, petrogenic and pyrogenic PAHs are characterized by dominance of five alkylated PAH homologous series and by dominance of unsubstituted high-molecular-weight PAHs, respectively; therefore, determination of the changes in this ratio more truly reflects the difference in the PAH distribution between these two sets of hydrocarbons. (2) Determination of these two sets of PAHs has become conventional measurement for many environmental labs, and this ratio can offer better accuracy with less uncertainty than those relative ratios determined from individual PAH compounds. (3) This ratio shows great consistency from sample to sample and is subject to little interference from the concentration fluctuation of individual components within the PAH series. Also, long-term natural weathering (for example, the highly weathered 25-year-old Nipisi spill samples) and biodegradation (the biodegraded ASMB oil series and nine biodegraded Alaska oil series) only slightly alter the values of this ratio (Table 3), but the ratio will be dramatically altered by combustion. Therefore, this index ratio can be used as a general and effective criterion to unambiguously differentiate pyrogenic PAHs and petrogenic PAHs. Figure 4 depicts the pyrogenic index versus the relative ratios of phenanthrene to anthracene for over 60 oils and refined products (including jet fuel, diesel, lube oil, Bunker C, and heavy fuel) analyzed. As Figure 4 shows, the pyrogenic index exclusively falls in the range of 0.01-0.05 for oils and refined products, while it dramatically increased to a range of 0.8-2.0 for the six 1994 Mobile burn soot samples. The difference in the magnitude of the data is very significant. It is also seen from Figure 4 that the jet fuel, diesel, and most crudes show the pyrogenic index ratios smaller than 0.01 with ratios of phenanthrene/anthracene being very scattered. But heavy oils (such as Cold Lake Bitumen and Orimulsion) and heavy fuels (such as IFO-180, A-02, IF-30, and Bunker C type fuels) show significantly higher ratios (falling in the range of 0.01-0.05, clusters 1 and 2), indicating that this index ratio can be also used as a screening tool to distinguish heavy oils and heavy fuels from most crude oils and light petroleum products. For example, as Figure 4 shows, the ratios for the unknown tarball samples collected from the coasts of British Columbia (BC) and California (CA) in 1996 and Newfoundland in 1997 also fell in the ratio range for Bunker C type fuels, implying that these tarballs might be from a source of heavy fuels. A comprehensive study using GC/MS and isotopic techniques has revealed that the tarball samples from BC and CA were chemically similar, and both were originated from bunker type fuels (30). It is noted that some heavy fuels (such as IF-30 and A-02) and unknown tarball samples collected from the Newfound3108

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

land shorelines in 1997 show the ratios of phenanthrene to anthracene smaller than 10. In particular, the nine sets of the biodegraded Alaskan oils show the ratios of phenanthrene to anthracene also to be smaller than 10, due to the preferential degradation of phenanthrene over anthracene by the defined inoculum (31). An incorrect conclusion could be derived if this ratio was used alone for identifying the contamination source of PAHs. In contrast, however, the pyrogenic index ratio only shows slight changes regardless of the type of oils and whether oils were weathered or biodegraded. Hence, this quantitative ratio, combined with other qualitative criteria, can be used to unambiguously differentiate sources of PAHs.

Literature Cited (1) Fingas, M.; Ackerman, F.; Lambert, P.; Li, K.; Wang, Z. D.; Nelson, R.; Goldthorp, M.; Mullin, J.; Hannon, L.; Wang, D.; Steenkammer, A.; Schuetz, S.; Turpin, R.; Campagna, P.; Graham, L.; Hiltabrand, R. In Proceedings of The 19th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar; Environment Canada: Ottawa, ON, 1996; pp 907-978. (2) Walton, W. D.; Twilley, W. H.; Putorti, A. D.; Hiltabrand, R. R. In Proceedings of The 18th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar; Environment Canada: Ottawa, ON, 1995; pp 1053-1064. (3) Fingas, M.; Ackerman, F.; Li, K.; Lambert, P.; Wang, Z. D.; Bissonnette, M. C.; Campagna, P.; Boileau, P.; Laroche, N.; Jokuty, P.; Nelson, R.; Turpin, R.; Trespalacios, M.; Halley, G.; Belanger, J.; Pare, J.; Vanderkooy, N.; Tennyson, E. J.; Aurand, D.; Hiltabrand, R. In Proceedings of The 17th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar; Environment Canada: Ottawa, ON, 1994; pp 1099-1163. (4) Walavalkar, A. Y.; Kulkarni, A. K. In Proceedings of The 19th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar; Environment Canada: Ottawa, ON, 1996; pp 1081-1103. (5) Benner, B. A., Jr.; Bryner, N. P.; Wise, S. A.; Mulholland, G. W.; Lao, R. C.; Fingas, M. Environ. Sci. Technol. 1990, 24, 14181427. (6) Evans, D. D. In In Situ Burning Oil Spill Workshop Proceedings; NIST SP 867; U.S. Department of Commerce: Gaithersburg, MD, 1994; pp 29-37. (7) Shu, Y. Y.; Lao, R. C.; Belanger, J.; Fingas, M.; Pare, J. Proceedings of The 12th Technical Seminar on Chemical Spills (TSOCS); Environment Canada: Ottawa, ON, 1995; pp 165-171. (8) Wang, Z. D.; Fingas, M.; Li, K. J. Chromatogr. Sci. 1994, 32, 361-366. (9) Wang, Z. D.; Fingas, M.; Li, K. J. Chromatogr. Sci. 1994, 32, 367-382. (10) Wang, Z. D.; Fingas, M.; Sergy, G. Environ. Sci. Technol. 1994, 28, 1733-1746. (11) Page, D. S.; Boehm, P. D.; Douglas, G. S.; Bence, A. E. In Exxon Valdez Oil Spill: Fate and Effects in Alaska Waters; Wells, P. G., Butler, J. N., Hughes, J. S., Eds.; American Society for Testing and Materials (ASTM): Philadelphia, PA, 1995; pp 41-83. (12) Wang, Z. D.; Fingas, M. J. Chromatogr. 1997, 774, 51-78. (13) Wang, Z. D.; Fingas, M.; Sergy, G. Environ. Sci. Technol. 1995, 29, 2622-2631. (14) Teal. J. M.; Farrington, J. W.; Burst, K. A.; Stegeman, J. J.; Tripp, B. W.; Woodin, B.; Phinney, C. Mar. Pollut. Bull. 1992, 24, 607614. (15) Farran, A.; Grimelt, J.; Albaiges, J.; Bottelo, A. V.; Macko, S. A. Mar. Pollut. Bull. 1987, 18, 284-289. (16) Bence, A. E.; Kvenvolden, K. A.; Kennicutt II, M. C. Org. Geochem. 1996, 24, 7-42. (17) Wang, Z. D.; Fingas, M. J. Microcolumn Separations 1995, 7, 617-639. (18) Bjøeseth, A. In Handbook of Polycyclic Aromatic Hydrocarbons; Bjøeseth, A., Ramdahl, T., Eds.; Marcel Dekker: New York, 1985; pp 1-20. (19) Boehm, P. D.; Douglas, G. S.; Burns, W. A.; Mankiewicz, P. J.; Page, D. S.; Bence, A. E. Mar. Pollut. Bull. 1997, 34, 599613. (20) Page, D. S.; Boehm, P. D.; Douglas, G. S.; Bence, A. E.; Burns, W. A.; Mankiewicz, P. J. Environ. Toxicol. Chem. 1996, 15, 12661281. (21) Brassell, S. C.; Eglinton, G. In Analytical Techniques in Environmental Chemistry; Albaiges, J., Ed.; Pergamon Press: Oxford, 1980; pp 1-22.

(22) Badger, G. M. The Chemical Basis of Carcinogenic Activity; Charles C. Thomas: Springfield, IL, 1962. (23) Gonzalez-Vila, F. J.; Lopez, J. L.; Martin, F.; del Rio, J. C. Fresenius J. Anal. Chem. 1991, 339, 750-753. (24) Fingas, M.; Ackerman, F.; Lambert, P.; Li, K.; Wang, Z. D.; Mullin, J.; Hannon, L.; Wang, D.; Steenkammer, A.; Hiltabrand, R.; Turpin, R.; Campagna, P. In Proceedings of The 18th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar; Environment Canada: Ottawa, ON, 1995; pp 915-995. (25) Fingas, M.; Li, K.; Ackerman, F.; Wang, Z. D.; Lambert, P.; Gamble, L.; Trespalacios, M.; Schuetz, S.; Turpin, R.; Campagna, P. In Proceedings of The 19th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar; Environment Canada: Ottawa, ON, 1996; pp 999-1032. (26) Walton, W. D.; McElroy, J.; Twilley, W. H.; Hiltabrand, R. R. In Proceedings of The 17th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar; Environment Canada: Ottawa, ON, 1994; pp 735-764.

(27) Fingas, M.; Li, K.; Ackerman, F.; Campagna, P.; Turpin, R.; Getty, S.; Soleki, M.; Trespalacios, M.; Wang, Z. D.; Mullin, J.; Tennyson, E. Spill Sci. Technol. Bull. 1996, 3, 123-137. (28) Theobald, N.; Rave, A.; Jerzycki-Brandes, K. Fresenius J. Anal. Chem. 1995, 353, 83-87. (29) Benlahcen, K. T.; Chaoui, A.; Budzinski, H.; Bellocq, J.; Garrigues, P. Mar. Pollut. Bull. 1997, 34, 298-305. (30) Wang, Z. D.; Fingas, M.; Landriault, M.; Sigouin, L.; Castle, B.; Hostetter, D.; Zhang, D.; Spenser, B. J. High Resol. Chromatogr. 1998, 21, 383-395. (31) Foght, J.; Semple, K.; Westlake, D. W. S.; Blenkinsopp, S.; Sergy, G.; Wang, Z. D.; Fingas, M. J. Industrial Microbiol. Biotechnol. 1998, 21, 322-330.

Received for review January 11, 1999. Revised manuscript received May 3, 1999. Accepted June 2, 1999. ES990031Y

VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3109