Heavy duty (HD) diesel trucks are responsible for a disproportionate amount of BC and POA relative to light duty (LD) vehicles due to emission factors that are ∼50 and ∼15 times larger, respectively, the researchers note.

In contrast to BC, POA contains many thousands of organic compounds that are thought to be derived from unburned fuel, unburned lubricating oil, and as byproducts of incomplete combustion. Previous chassis dynamometer studies have estimated the contributions of unburned fuel and unburned lubricating oil to POA emissions for a small number of vehicles (<100 per study). These studies showed a wide range in the contribution of unburned lubricating oil to POA emissions (20−80%). A substantial fraction of POA has been shown to be semivolatile under atmospheric conditions. Following emission these organics can evaporate due to dilution to form low volatility vapors that can be oxidized to form SOA.

Characterizing the chemical composition of POA and apportioning the contribution of these sources to the total POA mass is important as the chemical composition affects the rate and composition of SOA formation following evaporation, as well as the physical properties and toxicity of the aerosol.

The number of possible hydrocarbon isomers increases exponentially with increasing carbon number, which results in a very high number of constitutional isomers at large carbon numbers. Due to the large number of hydrocarbon compounds present in POA, many of which are constitutional isomers, the vast majority of the mass cannot be resolved by conventional separation techniques … In this work, we determine the comprehensive chemical composition of POA for a much larger number of vehicles than is possible with dynamometer testing (N = 4000−7500 light duty, 50−200 medium, and 0−150 heavy duty) by analyzing filter samples collected in a highway tunnel using GC × GC/VUV-HRTOFMS. It is important to note that these data represent composites of a high number of vehicles and while they are representative of real-world fleet and driving conditions, high-emitting vehicles can make a disproportionate contribution to these composite samples. Here we use this comprehensive analysis to better determine the contributions of unburned gasoline fuel, unburned diesel fuel, lubricating oil, and combustion byproducts to motor vehicle POA compared to tracer based methods.

—Worton et al.

As in their earlier study on the formation of secondary organic aerosols (earlier post), the team collected filter samples at the Caldecott Tunnel on Highway 24 in the San Francisco Bay Area. The tunnel is 1 km long and eastbound traffic is uphill on a 4% grade—helpful for providing engine load for the purposes of emissions testing.

Averaged chemically characterized distributions of motor vehicle POA, separated into low heavy-duty (HD) (BC <15 μg m−3) and high HD influence (BC>15 μg m−3), and SAE-10W30 and SAE-15W40 lubricating oils. Distributions are shown as a function of carbon number (NC) and number of double bond equivalents (NDBE). The POA distributions are shown as mass concentrations in μg C m−3 while the lubricating oils are shown as mass fraction. Also shown is the average number of vehicles (±1 standard deviation) observed during filter collection periods separated by type (light duty, LD; medium duty, MD; and heavy duty, HD). Credit: ACS, Worton et al. Click to enlarge.

Filter samples were collected downstream of sharp cut cyclones at a flow rate of 16.7 L min⁻¹ to remove particles with diameters greater than 2.5 μm from the sampled air.

Portions of all the filters (1.1 cm²) were analyzed for total organic carbon (OC) and black carbon (BC) using a thermal optical technique. A subset of 18 filter samples were analyzed by GC × GC/HRTOFMS (gas chromatography × GC/high resolution time-of-flight mass spectrometry).

The recently developed gas chromatography mass spectrometry approach utilizes “soft” vacuum ultraviolet photoionization to achieve unprecedented chemical characterization of the motor vehicle POA emissions with a mass closure of >60%.

The researchers characterized the POA by number of carbon atoms (N_C); number of double bond equivalents (N_DBE); and degree of molecular branching.

They observed that vehicular POA mass was dominated by branched cycloalkanes with one or more rings and one or more branched alkyl side chains (≥80%). The remaining mass comprised branched alkanes (<5%); n-alkanes (<3%); single ring aromatics (<3%); PAHs (<2%; and oxygenates (∼2%). High molecular weight combustion byproducts—i.e., alkenes, oxygenates, and aromatics—were not present in significant amounts.

Based on the similar observed carbon number (volatility) and chemical composition of POA and lubricating oil, and in the absence of significant high molecular weight material formed through the combustion process, we conclude that lubricating oil is the dominant constituent of vehicular POA emissions.

… The dominance of cycloalkanes in vehicular POA emissions has implications for the atmospheric behavior of POA, as a result of the structural dependence of hydrocarbon reaction rates and SOA yields. … Taken together, these data suggest that evaporated POA components, which are likely dominated by cyclohexane and cyclopentane rings, will react slower and form less fragmentation products than if it were assumed to be purely branched alkane material. Gas phase oxidation of relevant cycloalkanes is less well studied and further work to characterize the fate of branched cyclohexanes and cyclopentanes is important to improve our understanding of the processing of atmospheric POA in regions downwind of urban centers.

—Worton et al.

Resources

• David R. Worton, Gabriel Isaacman, Drew R. Gentner, Timothy R. Dallmann, Arthur W. H. Chan, Christopher Ruehl, Thomas W. Kirchstetter, Kevin R. Wilson, Robert A. Harley, and Allen H. Goldstein (2014) “Lubricating Oil Dominates Primary Organic Aerosol Emissions from Motor Vehicles,” Environmental Science & Technology doi: 10.1021/es405375j