Gas / Particle Partitioning of Dioxins in Exhaust Gases from Automobiles

This study investigates distributions of polychlorinated dibenzo-p-dioxin and polychlorinated dibenzofuran (PCDD/F) congeners in the exhaust gases of gasolineand diesel-fueled vehicles. 6 sport utility vehicles (SUVs), 6 diesel passenger vehicles (DPVs), and 3 heavy duty diesel vehicle (HDDV) were examined using chassis dynamometer tests for measuring vehicular dioxin emissions. The mean PCDD/F I-TEQ emission factors were 0.101, 0.0688 and 0.912 ng I-TEQ/km for the SUVs, DPVs and HDDV, respectively. Highly chlorinated congeners dominated both gaseous and particulate phase PCDD/Fs. The major contributors of gas-phase PCDD/F I-TEQ for the SUVs, DPVs, and HDDV were 2,3,4,7,8-PeCDF, 2,3,7,8-TeCDD, and 2,3,4,7,8-PeCDF, respectively; however, 2,3,4,7,8-PeCDF was the major contributor in particulatephase PCDD/F I-TEQ of these vehicles. The particulate-phase PCDD/Fs was responsible for 78.0, 90.3 and 71.1% of total PCDD/Fs for the SUVs, DPVs, and HDDV, respectively. Therefore, the control of particulate matter is more critical than that of gaseous pollutants for reducing PCDD/F emissions from automobiles.


INTRODUCTION
In recent years, polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/Fs) has been extensively concerned due to their acute toxicity and associated adverse health effects.PCDD/F congeners with chlorine substitution in 2,3,7,8 positions are most toxic to human and the PCDD/Fs can be formed during any incomplete combustion process (Bumb et al., 1980).In the past, many studies focused on the PCDD/F emissions from various sources, such as municipal solid waste incinerators (MSWIs), metallurgical activities, crematories and temples (Wang et al., 2003a-b;Lee et al., 2004;Li et al., 2007;Wang and Chang-Chien, 2007;Hu et al., 2009a-c).However, mobile sources are almost as important as stationary sources in PCDD/F emission contribution in many countries.
The emission of PCDD/Fs from automotive has been investigated in several studies (Geueke et al., 1999;Joumard et al., 2000;US EPA, 2001;Kim et al., 2003;Weilenmann et al., 2005;Dyke et al., 2007).For example, Ryan and Gullett (2000;2002) reported that the value was 0.023 ± 0.022 ng TEQ/km for a HDDV with a high mileage engine.Geueke et al. (1999) found that the PCDD/F emission concentrations ranged within 0.93-45 and 1.2-218 pg/Nm 3 , for unleaded gasoline cars and diesel trucks, respectively.Some researchers also estimated the PCDD/F emissions of vehicles in tunnels (Oehme et al., 1991;Wevers et al., 1992).Oehme et al. (1991) observed that the PCDD/F emission factors of gasoline and diesel vehicles were 0.028-0.52 and 0.72-9.5 ng I-TEQ/km, respectively.Ryan and Gullett (2000) indicated that tunnel studies relied on indirect emission measurements and generalized assumptions (e.g., average fuel consumption and/or mileage) that can introduce uncertainties in the derivation of emission estimates.In Taiwan, the relative contribution fraction of PCDD/Fs from mobile sources has gradually raised since the execution of more stringent PCDD/F emission standards for stationary sources.However, limited data are available to assess the magnitude of automobile PCDD/F emissions.
In the exhaust of vehicles, PCDD/Fs may be gaseous or bound to particulate matter (PM).Precisely understanding the PCDD/Fs partitioning in vapor/solid phases is important for developing the PCDD/F control technologies of vehicles and modeling the transportation of PCDD/Fs in environment.Atmospheric fate of PCDD/Fs is primarily governed by their partitioning between particle and gas phases that influences the deposition, chemical reaction, and long range transport of PCDD/Fs.The vapor/solid partitioning of PCDD/Fs is also important for their movement from sources to aquatic environments and the food chain.Therefore, understanding PCDD/Fs partitioning between gas and particulate phases is necessary to conduct their dispersion modeling and risk assessments (Oh et al., 2001;Rice et al., 2008;Armitage et al., 2009;Novak et al., 2009).So far, few study reported the PCDD/F partitionings between vapor/solid phases in vehicle exhausts.To examine this important feature, this study was motivated to investigate the partitioning of PCDD/Fs between vapor/ solid phases in the exhausts of gasoline-and diesel-fueled vehicles.In this study, 6 SUVs, 6 DPVs, and 3 HDDV were tested in accordance with regulated driving cycles to estimate emission concentrations, emission factors, and gas/particulate phase distributions of PCDD/Fs in the vehicle tailpipe exhausts.This study provides essential information for developing control strategies, dispersion modeling, and risk assessments of vehicular PCDD/F emissions.

Vehicles
This study took into account the sale volumes, mileages, and engine years of tested SUVs, DPVs, and HDDV before conducting experiments.Therefore, we tested six 2002-2005 top five sale SUVs, six 2005-2006 top three DPVs, and three 1994-1999 best sale HDDV, compliant to the Phase II HDDV emission regulations.The mileages of tested SUVs and DPVs ranged from 20000 to 62500 km and 12000 to 70000 km, respectively.The basic information concerning the tested vehicles is listed in Table 1.
The FTP cycle sequence includes tests of expressway, congested urban, and uncongested urban driving patterns.For the vehicle exhaust gas emission measurements, the FTP-75, NEDC, and FTP transient cycle test procedures are also the current Taiwan standard emission driving patterns used for SUVs, DPVs, and HDDV, respectively.Thus, they were employed in this study.

Sampling
All tested engines were fueled with the same batch commercial diesel fuel or unleaded gasoline in order to disable the interference resulted from fuel variations.The sampling procedures followed the USEPA modified Method 23.The sampling train adopted in this study is comparable with that specified by the USEPA modified Method 5. Gas-phase samples were collected by XAD-2 resins while particulate-phase ones were simultaneously sampled using fiberglass filters.All the exhaust gas samples were collected isokinetically under the aforesaid test cycles and implemented for a complete test period.
High-resolution gas chomatograph/high-resolution mass spectrometer (HRGC/HRMS) was used for PCDD/Fs analyses.The HRGC (Hewlett-Packard 6970 Series gas chromatograph, CA) was equipped with a DB-5 fused silica capillary column (L = 60 m, i.d.= 0.25 mm, film thickness = 0.25 μm) (J&W Scientific, CA) with a splitless injection.The oven temperature program was set according to the following: beginning at 150°C (held for 1 min), followed at 30 °C/min to 220°C (held for 5 min), then at 1.5 °C/min to 240°C (held for 5 min), and finally at 1.5 °C/min to 310°C (held for 20 min).The HRMS (Micromass Autospec Ultima, Manchester, UK) mass spectrometer was equipped with a positive electron impact (EI+) source.The analysis mode of the selected ion monitoring (SIM) was used with resolving power at 10000.The electron energy and source temperature were specified at 35 eV and 250°C, respectively.

Average PCDD/F Concentrations and Emission Factors in Exhaust Gases
The mean PCDD/F I-TEQ concentrations in exhaust gases of the SUVs, DPVs, and HDDV are listed in Table 2, which were 0.0544 (RSD = 70.3%),0.0337 (88.6%), and 0.0724 ng I-TEQ/Nm 3 (56.7%),respectively.The total PCDD/F I-TEQ concentrations of SUVs and HDDV were 1.6 and 2.1 times h ighe r tha n that of DPVs.The PCDD/ PCDF I-TEQ ratios were all less than 1, indicating that the toxicity from PCDFs dominated the total PCDD/F toxicity in the exhaust gases of vehicles.
The PCDD/F I-TEQ concentrations obtained in this study were higher than those of HDDVs (1.1-9.7 pg I-TEQ/Nm 3 ) reported by Geueke et al. (1999) and a light duty diesel engine (6.4-14.5 pg I-TEQ/Nm 3 ) (Kim et al., 2003).Hagenmaier et al. (1990) observed that the PCDD/F I-TEQ concentration was 141.5 pg I-TEQ/Nm 3 for passenger cars fueled with leaded gasoline.
The fuel consumption had not been measured in this study, so the PCDD/F emission factors of the vehicles obtained in this study were only reported on the basis of mileage.For the tested SUVs, DPVs, and HDDV, the calculated mean PCDD/F emission factors were 2.92, 0.884 and 13.6 ng/km, respectively, with corresponding mean I-TEQ emission factors of 0.101, 0.0688 and 0.912 ng I-TEQ/km, respectively.Thus, the HDDV emitted more PCDD/Fs than the other tested vehicles.The PCDD/F I-TEQ emission factors of SUVs and DPVs in this study were greater than those of unleaded gasoline vehicles and a diesel car (0.0007-0.005 and 0.0024 ng I-TEQ/km, respectively) reported by Hagenmaier et al. (1990) and that of a diesel truck in tunnel sampling (0.172 ng I-TEQ/km) performed by Gertler et al. (1996).The PCDD/F I-TEQ emission factor of HDDV in this study is comparable to the data (0.663-1.3 ng I-TEQ/km) of HDDVs in dynamometer tests obtained by Lew et al. (1993).The parameters such as the models, mileages, and engine years had been evaluated their relation with PCDD/Fs by conducting principal component analysis and spearman correlation.However no relation had been found between them, maybe because there were some more influential factors, such as maintenance condition, replacement period of lubricant, and the types of catalyst convertors, that had not been considered in this study.

Partitioning of Gas/Particulate Phase PCDD/Fs
Table 3 shows the PCDD/F gas/particulate phase distributions in the exhaust gases of tested vehicles.The percentage of a gaseous (or particulate) phase PCDD/F congener is equal to its concentration × 100% divided by the total (gas + particulate) PCDD/F concentration.The particulate-phase PCDD/Fs accounted for 78.0, 90.3 and 71.1% of total PCDD/Fs for the SUVs, DPVs and HDDV, respectively.These vehicles emitted more particulatephase PCDFs than particulate-phase PCDDs.For the SUVs, the top PCDD/F contributor was particulate-phase OCDD (27.3%), and the highly chlorinated (hexa-to octa-) PCDD/Fs was responsible for more than 90.1% of total PCDD/Fs.Similarly, the particulate-phase OCDD was the major contributor of PCDD/Fs for the DPVs and HDDV, and the highly chlorinated PCDD/Fs accounting for 83.6 and 84.2% of total PCDD/Fs, respectively.Note that the PCDD/Fs were almost dominated by particulate-phase PCDD/Fs except for 2,3,7,8-TeCDD.

CONCLUSIONS
In this study, 6 SUVs, 6 DPVs and 3 HDDV were tested (in accordance with standard driving cycles) for vehicular gas-and particle-phase PCDD/F emissions.The DPVs emitted lower PCDD/F concentrations than the other tested vehicles.Highly chlorinated congeners dominated gas-and particulate-phase PCDD/Fs in the exhausts of vehicles.Particulate-phase PCDD/Fs accounted for 78.0, 90.3 and 71.1% of total PCDD/Fs for the SUVs, DPVs and HDDV, respectively; furthermore, the major contributor of PCDD/Fs was the particulate-phase OCDD.The particulatephase PCDD/Fs also contributed more PCDD/F I-TEQ emissions than the gas-phase ones in the exhaust gases.

Fig. 1 .
Fig. 1.Congener profiles of gas/particulate phase PCDD/Fs in the exhaust gases of vehicle.

Fig. 2 .
Fig. 2. Partitioning of gas/particulate phase PCDD/F I-TEQs in the exhaust gases of vehicles.

Table 1 .
Basic information concerning the tested vehicles.

Table 2 .
Mean PCDD/F concentrations and emission factors of vehicles.

Table 3 .
Percentages of gas/particulate (G/P) phase PCDD/F concentrations in the exhaust gases of vehicles.