Chemical Characterization Particulate

Aerosol Science 38 (2007) 1079 – 1118 www.elsevier.com/locate/jaerosci

Review

Chemical characterization of particulate emissions from diesel engines: A review M. Matti Maricq∗ Research and Advanced Engineering, Ford Motor Company, P.O. Box 2053, MD 3179, Dearborn, MI 48121, USA Received 15 June 2007; received in revised form 30 July 2007; accepted 8 August 2007

Abstract This review examines the chemical properties of particulate matter (PM) in diesel vehicle exhaust at a time when emission regulations, diesel technology development, and particle characterization techniques are all undergoing rapid change. The aim is to explore how changes in each of these areas impact the others. Particle composition is of central interest to the practical issues of health effects, climate change, source apportionment, and aerosol modeling. Thus, the emphasis here is to identify the emerging questions and examine how they can be addressed. As regulations drive down the allowed tailpipe emission levels, advances in engine and aftertreatment technology have made it possible to substantially reduce PM emissions. Besides the reduction in level, new technologies such as diesel particulate filters (DPFs) and selective catalytic reduction (SCR) can also affect the physical and chemical properties of PM. This in turn introduces new analytical demands that must address not only the issue of sensitivity, but also of specificity. New methods of aerosol chemical analysis are described that address these needs, improve our understanding of particle composition, and provide critical insight into the current issues surrounding motor vehicle PM emissions and their environmental impact. 䉷 2007 Elsevier Ltd. All rights reserved. Keywords: Diesel; PM; Particulate matter; Soot; Chemical characterization

Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Traditional diesel exhaust PM characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Sampling PM from vehicle exhaust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Overview of chemical speciation methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. EC/OC analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Motor vehicle PM composition—roadway tunnel studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Motor vehicle PM composition—laboratory studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Next generation diesels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Advances in diesel engine technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. DPF/oxidation catalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. NOx aftertreatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Fuel/lubricant modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Measurement challenges at low emission levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ∗ Tel.: +1 313 594 7527.

E-mail address: [email protected]. 0021-8502/$ - see front matter 䉷 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jaerosci.2007.08.001

1080 1081 1082 1085 1086 1086 1087 1090 1090 1090 1091 1092 1092

1080

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

4. Emerging tools for PM chemical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Single particle MS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Thermal desorption aerosol MS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Multiple laser approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. DMA/ICP-MS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Laser breakdown spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6. Advanced X-ray methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7. Neutron scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.8. Raman spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.9. Particle volatility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Chemical insight into current questions on diesel PM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. How does engine design and operation impact PM emissions? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. What is the influence of ambient temperature on diesel PM? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. How do fuel properties affect PM emissions? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4. How does lube oil affect PM emissions? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5. What are diesel nanoparticles? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6. How does exhaust aftertreatment affect PM? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7. Can PM composition help source apportionment? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.8. How do atmospheric process alter diesel exhaust? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.9. What can PM composition contribute to health effects studies? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.10. What about PM from non-road diesel engines? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1092 1093 1094 1095 1096 1096 1097 1097 1098 1099 1100 1101 1101 1102 1103 1103 1105 1107 1108 1108 1109 1110 1110 1110

1. Introduction Particulate matter (PM) has the distinction of being the only criteria pollutant regulated by the United States Environmental Protection Agency (US EPA) that is not chemically defined. Even its physical specification is limited to two aerodynamic 50% cut points of D50 = 10 and 2.5 m that delineate the PM10 and PM2.5 categories, respectively. Yet, of the criteria pollutants, PM has arguably the widest range of potential environmental impacts, including health effects, climate change, ecological effects, and visibility. Chemical composition should be important in each of these areas. Sulfate aerosols reflect sunlight, but elemental carbon (EC) particles absorb light, with opposite influences on climate change (Jacobson, 2001; Penner, Chuang, & Grant, 1998). Numerous epidemiological studies reveal a small yet persistent association between ambient PM levels and a variety of primarily cardio-pulmonary adverse health effects (HEI, 2003; Pope et al., 2004). To date, these effects appear generic of PM (HEI, 2002), but the expectation remains that the causative mechanism(s) when identified will exhibit chemical specificity. Diesel combustion is widely used in both stationary and mobile applications, especially where high power output is needed. There is presently a strong resurgence of interest in diesel technology for automobiles. One motivation is the desire for improved fuel efficiency and reduced greenhouse gas emissions (Weiss, Heywood, Drake, Schafer, & AuYeung, 2000). Another is the performance benefit of high torque at low engine speeds provided by diesel engines. But, diesel engines represent an important source of ambient particles (Lloyd & Cackette, 2001). And unless the conventional diesel’s PM and NOx emissions are brought into compliance with the increasingly strict standards being implemented, these benefits will not be realized under regulatory changes taking place over the next 5–10 years. Consequently, diesel engine emissions are currently under considerable scrutiny. This arises from two perspectives. First, there is a concerted effort from the engineering side to reduce tailpipe emissions. This includes the introduction of high pressure, common rail, direct injection systems (Guerrassi & Dupraz, 1998; Stumpp & Ricco, 1996), the use of shaped and multiple pulse fuel injection strategies (Park, Kook, & Bae, 2004), and the development of NOx and PM aftertreatment devices (Johnson, 2003). Second, there has been a parallel rapid growth in the investigation of diesel emissions (Mohr, Lehmann, & Margaria, 2003a, 2003b), the development of new methods to sample engine exhaust, and the advancement of instrumentation used to characterize the PM (Burtscher, 2005; Kittelson, 1998). The rapid changes in government air quality and emission regulations taking place over the present decade provide a primary motivation for the advances in both areas. The EPA Tier II and 2007 heavy duty diesel (HDD) regulations

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1081

limit PM emissions to 10 mg/mi and 10 mg/hp-h, respectively, for light-duty and heavy-duty vehicles. Europe’s EURO 5 standard is now being revised downward from 5 to 3 mg/km, and Japan has proposed to limit emissions to 5 mg/km beginning in 2009. Additional regulations apply to non-road applications. In 2004 the EPA promulgated its Clean Air Nonroad Diesel—Tier 4 Final Rule (US EPA, 2004). These regulations mandate emission reductions of greater than 90% for construction, agricultural and industrial diesel-powered equipment, consistent with the measures applied to motor vehicles. The new PM emission limits are below what can be achieved solely by engine design and, therefore, require exhaust aftertreatment. The preferred approach is the diesel particulate filter (DPF), which exists in many variations (Cutler, 2004). Most are wall-flow devices, which remove PM from exhaust by forcing it through porous walls separating adjacent channels in a monolithic substrate. All require regeneration to remove accumulated soot, or risk excessive backpressure on the engine. They differ with respect to substrate material, regeneration strategy (continuous versus forced), and use of catalysts (none, fuel borne, incorporated into substrate). Measurements at the lowered PM standards require more than just a commensurate increase in instrument sensitivity (Vouitsis, Ntziachristos, & Samaras, 2003). Below about 5 mg/km, the regulatory method for PM mass measurement faces qualitatively new challenges as effects from temperature, humidity, electrostatic charge, and gas phase adsorption become comparable to the mass collected by filter media (Chase, Duszkiewicz, Lewis, & Podsiadlik, 2005; Chase et al., 2004). Moreover, exhaust aftertreatment does not just lower emissions, it alters the chemical composition of vehicle exhaust. By trapping soot particles, the DPF reduces the available surface area for condensation of semivolatile compounds as they exit the tailpipe and, thereby, increases the propensity for nucleation (although the DPF’s large interior surface area might also provide a sink for some semivolatile compounds). Thus, while PM upstream of the DPF is primarily EC, the exhaust downstream can be dominated by semivolatile PM. NOx aftertreatment, either by urea-based selective catalytic reduction (SCR) or lean NOx trap, can further alter the chemical composition of diesel exhaust. Chemical information has an important role to play in the study of diesel emissions. It can help clarify the origin of adsorption artifacts that interfere with the regulatory filter-based PM measurement method. It can help optimize multicomponent aftertreatment systems composed of DPF, NOx reduction, and oxidation catalyst needed to meet the combined hydrocarbon, CO, NOx , and PM emission standards. It can support health effects studies, such the Advanced Collaborative Emissions Study (ACES) to evaluate the health effects of “clean diesel” technology. It can provide valuable feedback as to how effectively regulations accomplish their goals of improving air quality. And the benefits of understanding diesel PM chemistry extend further to diverse areas such as the degradation of stone buildings (Simão, Ruiz-Agudo, & Rodriguez-Navarro, 2006). We begin by reviewing the standard methods for sampling and chemical speciation of PM from conventional diesel engines. But the emphasis of the paper is on emerging chemical characterization methods. These methods constitute a diverse group, one that cannot be easily categorized into a one to one correspondence with specific particle characteristics. Instead, we explore these methods via a series of topical questions about new developments in diesel engine and aftertreatment technology, the possible ensuing changes in the character of diesel PM, and the potential environmental impact.

2. Traditional diesel exhaust PM characterization An artist’s conception of diluted and cooled diesel PM is illustrated in Fig. 1. It consists of two types of particles: (a) fractal-like agglomerates of primary particles 15–30 nm in diameter, composed of carbon and traces of metallic ash, and coated with condensed heavier end organic compounds and sulfate; (b) nucleation particles composed of condensed hydrocarbons and sulfate. The standard approach to studying the composition of this material is by off-line chemical analysis. This generally proceeds in the following steps: (1) preparation of collection substrates, (2) sampling exhaust from the vehicle, (3) extraction of PM from the substrates, and (4) chemical analysis of the recovered material. Aspects of the second step are specific to the motor vehicles, but the remaining three steps follow standard analytical chemistry protocols. Therefore, the present section begins by describing how PM is sampled from engine exhaust and then gives a brief review of the chemical analysis techniques, with more detailed descriptions left to the standard analytical chemistry literature. The non-chemically specific, but oft used thermal EC/OC method is reviewed, and the section concludes with a survey of diesel PM chemical composition studies.

1082

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

= soot

= condensed HC/SO4

= nucleation mode

= imbedded metallic ash

Fig. 1. Artist’s conception of diesel PM.

2.1. Sampling PM from vehicle exhaust Engine exhaust consists primarily of nitrogen, oxygen, water vapor, and CO2 , with minor constituents of CO, hydrocarbons, NOx , and PM. Owing to the high temperature of engine exhaust, 150.300 ◦ C, and the potential for water condensation as it cools, regulatory protocols specify that PM emissions be measured through a dilution tunnel (see e.g., CFR, 2007 and the dilution stack sampler of Hildemann, Cass, & Markowski, 1989). A chassis dynamometer facility used for emissions testing of light-duty vehicles is illustrated in Fig. 2. A heated hose transfers exhaust from the tailpipe to the dilution tunnel that runs along the side wall of the test cell (the hose in the foreground vents the exhaust when the vehicle is prepped for testing). Filtered, temperature-, and humidity-controlled dilution air is supplied through the vertical structure at the right side of the dilution tunnel. Vehicle exhaust enters the tunnel coaxially ahead of an orifice plate to promote turbulent mixing. When constant volume sampling is used for sampling is used for testing light duty vehicles over a transient drive, cycle, the dilution ratio varies, but is on average ∼ 20 : 1, as compared to ∼ 1000 : 1 in the ambient. Sampling locations are located more than 10 tunnel diameters downstream to allow complete mixing of the exhaust and dilution air. An analogous procedure is used to test engine/aftertreatment systems in isolation from a vehicle. This is preferred for heavy-duty vehicles and off-road applications where testing the vehicle (tractor, bulldozer, locomotive, etc.) becomes impractical. In this case two stages of dilution are often used to accommodate the large exhaust flows from HD engines and avoid the need for cumbersomely large dilution tunnels. Conventional PM mass measurements are made simply by passing a steady flow of diluted exhaust through a filter and recording the mass increase of the filter at the end of the test, the proportionality of sampling being ensured by constant volume sampling dilution. As pointed out by Hildemann, Markowski, and Cass (1991), this is not sufficient for chemical analysis. Care must be taken to avoid materials such as plastics, rubber compounds, and grease often used in engine and vehicle test cells, since these potentially release organic compounds that pose a significant contamination problem for chemical speciation. Secondly, no single analytical method is capable of identifying all the chemical species in diesel exhaust. Therefore, most test campaigns rely on a suite of measurements aimed at a more or less rigorous chemical characterization based on the project’s aims. Table 1 lists a variety of sampling media and chemical analyses included in such suites.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1083

Fig. 2. Chassis dynamometer facility for measuring motor vehicle emissions.

Table 1 Engine exhaust sampling and chemical analysis methods Species

Sampling medium

Analysis method

Particulate matter Total mass Size selective cyclones for PM10 and PM2.5

Teflon or Teflon impregnated glass fiber (TIGF) filter

Gravimetric

Soluble organic fraction

Teflon filter or TIGF

Elemental/organic carbon Metals and elements Al, Si, P, S, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, As, Se, Br, Rb, Sr, Y, Zr, Mo, Pd, …

Pre-fired quartz filter Teflon filter

Weight loss after extraction with dichloromethane and drying Thermal/optical reflectance (TOR) Inductively coupled plasma mass spectrometry, X-ray fluorescence

Inorganic ions and acids − 2− 3− + NO− 2 , NO3 , SO4 , PO4 , NH4 , HNO2 , HNO3 , H2 SO4

Teflon or pre-fired quartz filter and water impingers

Water extraction and ion chromatography

Semivolatile organic compounds heavy hydrocarbons, PAH, hopanes/steranes Nitro-PAH

Dioxins/furans

XAD-coated annular denuder—Teflon or TIGF filter—PUF/XAD cartridge Teflon or TIGF filter followed by PUF/XAD cartridge Teflon or TIGF filter followed by PUF/XAD cartridge Large area Zefluor or TIGF filter

Extraction, HPLC separation, high resolution GC–MS Extraction, HPLC separation, negative ion chemical ionization GC–MS Extraction, conversion to silyl or methyl ester derivatives, GC–MS Extraction, high resolution GC–MS

Gaseous compounds VOC (C1 –C12 ) Carbonyls Alcohols CO, NO, NO2 , N2 O, SO2

Tedlar bags, SUMA canisters DNPH impingers Water impingers Tedlar bags or continuous

GC–MS or GC-FID HPLC GCMS or GC-FID NDIR and/or FTIR

Polar organic compounds

1084

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Fig. 3. Sampling trains to collect PM for chemical analysis (reprinted with permission from Schauer, J. J., Kleeman, M. J., Cass, G. R., & Simoneit, B. R. T. Environ. Sci. Technol. 33, 1578. Copyright 1999 American Chemical Society).

These sampling methods have evolved to suit the chemical species of interest. Fig. 3 illustrates three sampling train variations taken from the work by Schauer, Kleeman, Cass, and Simoneit (1999b) on diesel exhaust, but used by others as well (Zielinska, Sagebiel, Arnott et al., 2004; Zielinska, Sagebiel, McDonald, Whitney, & Lawson, 2004). At the upstream end of each, a cyclone separator removes large particles from the sample stream (typically > PM10 or > PM2.5 ). These are usually debris shaken loose from the walls of the exhaust system, sampling system, or other similar sporadic sources, which unchecked interfere with the intended PM emission measurements. The sampling train labeled “Filter/PUF” adds polyurethane foam (PUF) cartridges to capture gaseous species downstream of the PM sampling filter. It provides a means to sample semivolatile material, such as PAHs, that can coexist in both particulate and gas phases, and would otherwise likely be underestimated by filter collection alone. A drawback of the “Filter/PUF” sample train is that it does not distinguish between particle bound and gaseous species. This is overcome by adding an XAD-coated annular denuder upstream of the filter, as illustrated by the “Denuder/Filter/PUF” train in Fig. 3. By its rapid diffusion, gaseous material is adsorbed by the denuder, but particles and particle bound organic material pass through and are trapped by filters. Removal of the gaseous components from the sample alters the gas/particle phase equilibrium; thus, some particle bound semivolatile organics may blow off the filters during sampling. These are reclaimed by the PUF cartridges at the downstream end of the train. Subsequent chemical analyses of the material collected with this sampling train not only tell us about the chemical composition of the exhaust, but also about the gas–particle partitioning. The third sampling train, “Particulate and gaseous emissions”, is aimed at collecting PM, carbonyls, organic acids, and gas phase hydrocarbons. Two quartz filters in series collect PM for thermal analysis of elemental versus organic carbon (OC), the back-up filter providing an estimate of the positive vapor phase adsorption artifact. The Teflon followed by KOH impregnated quartz filters are used for gravimetric determination of PM mass (Teflon), X-ray fluorescence (XRF) analysis of trace elements (Teflon), and gas phase organic acids (KOH impregnated). The final Teflon filter provides a duplicate mass determination and allows inorganic ion measurement by ion chromatography, atomic absorption spectroscopy, and colorimetry. The downstream canister and DNPH cartridges are used for gas phase analysis of volatile hydrocarbons and carbonyls. The preceding sampling procedures provide no size information regarding the collected PM, other than the upper size cut-off of the cyclone. This can be remedied by replacing filters with a cascade impactor, which collects size selected particles onto aluminum foil substrates (Geller et al., 2002; Kleeman, Schauer, & Cass, 2000; Riddle, Robert, Jakober, Hannigan, & Kleeman, 2007; Zielinska, Sagebiel, Arnott et al., 2004). With chemical analysis in mind, the grease coating often used to reduce particle bounce is sometimes omitted to avoid interference. Owing to the nature

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1085

of diesel PM, soot with condensed semivolatile organic material, this omission should have minimal impact on the recorded size distributions. For a variety of reasons, increased PM collection rates may be needed: (1) emission levels are low (e.g., DPF equipped diesel); (2) the species of interest exist in low concentration (e.g., nitro-PAHs, dioxins, and furans); or (3) size selected PM samples are desired. This can be accomplished by increasing the sample flow rate through the collection medium, or raising the PM concentration in the sample. The former utilizes high volume sampling and correspondingly large area filters (e.g., 20 × 25 cm). Demokritou, Kavouras, Ferguson, and Koutrakis (2002) extended this concept to cascade impactors by designing one capable of operating at 900 L/min. They used PUF substrates to collect mg–gram quantities of PM, levels that would ordinarily overload foil substrates. The alternative is to increase particle concentration. This is accomplished using virtual impactors (Sioutas et al., 1997). The sample flow is split into two parts, with the major fraction directed through a sharp bend. Due to their inertia, particles preferentially follow the minor flow, leading to an enhanced concentration. This works well for micron size particles, but the enhancement factors fall off in the ultrafine regime as particle diameter decreases below 100 nm. As diesel particles are primarily found in the 10–500 nm range, this would distort their size distribution. Sioutas, Kim, and Chang (1999) designed an ultrafine concentrator to overcome this size limitation. In it particles are first grown to micron size using a saturation–condensation system; their concentration is then enhanced with a virtual impactor; and finally the original particles are recovered using a diffusion drier. Geller et al. (2002) combined this approach with a nano MOUDI cascade impactor to examine size selected chemical composition of ambient ultrafine particles, but it could be used as well for source sampling of diesel emissions. 2.2. Overview of chemical speciation methods Table 1 lists the main chemical characterization techniques used in motor vehicle PM studies, but this list is not allinclusive. The major classes of chemical analyses include: (1) elemental analysis, (2) inorganic ions, (3) hydrocarbons, and (4) polar organic compounds. Elemental analysis is usually accomplished using X-ray fluorescence (XRF) or inductively coupled plasma mass spectrometry (ICP-MS). The PM is collected by Teflon filters that are preleached using HCl or HNO3 and then flushed with water. Kellog and Winberry (1999) describe the application of XRF to analysis of ambient PM for 44 chemical elements lying in the second or lower rows of the periodic table. The detection limits vary by element, but are typically in the range of ng/cm2 when using Teflo filters. ICP-MS has superior sensitivity for many trace metals and is well suited for metal speciation, but the samples must first be processed by microwave-assisted acid digestion, as described in the roadway tunnel measurements of Lough et al. (2005). Inorganic ions, especially SO−2 4 , are important constituents of motor vehicle PM. Samples for inorganic analysis are collected onto Teflon filters. These are extracted with deionized-distilled water/alcohol mixture, and analyzed using ion chromatography (Small, 1989). The ions are identified by their retention time and are quantified by comparison to reference standards. There are a number of relatively recent papers on the organic chemical speciation of PM that include detailed descriptions of the methods, e.g., Schauer, Kleeman, Cass, and Simoneit (1999a, 1999b), Lev-On and Zielinska (2004), Zielinska, Sagebiel, Arnott et al. (2004), Zielinska, Sagebiel, McDonald et al. (2004). Prior to use, the components of the sampling trains, filters, PUFs and XAD cartridges, are cleaned by extraction with high purity solvents (examples include dichloromethane, hexane, methanol, and acetone). After testing, but prior to extraction, they are spiked with a mixture of deuterated internal standards (e.g., n-decane-d12, n-hexanoic acid-d11, benzaldehyde-d6, naphthalene-d8, biphenyl-d10, pyrene-d12, etc.), which are used to gauge the efficiency at which material can be reclaimed for analysis from the collection substrates. There are a variety of methods available for extraction, including Soxhlet, ultrasonic, and microwave-assisted. Piñeiro-Iglesias, López-Mahía, Vázquez-Blanco, Muniategui-Lorenzo, and Prada-Rodríquez (2002) compare these methods for the extraction of PAH and discuss their variability both with respect to standard reference materials and ambient samples. The extracted organic material is usually speciated using gas chromatography/mass spectrometry (GC/MS). Individual species are distinguished by their retention times, identified by the mass spectrum, and quantified by the total ion count relative to those from reference standards. Species identification can become difficult because of the many complex structures present in diesel PM. Instruments, such as the triple-quadrupole mass spectrometer, provide MS/MS capability to assist identifying compounds of similar molecular weight, for example as used by Henderson et al. (1983, 1984) and Bechtold et al. (1984) to investigate the possible role of nitroaromatic compounds and PAHs in diesel soot

1086

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

mutagenicity. It may also be difficult to identify some species of interest because they are present only in trace amounts. Nitro-PAHs are one such class of compounds of interest due to their potential health effects. Prior to GC/MS analysis these are isolated from the diesel extract using a combination of solid phase extraction and liquid chromatography, and then detected with high sensitivity using negative ion chemical ionization MS (Bezabeh, Bamford, Schantz, & Wise, 2003; Newton et al., 1982). Alternatively, direct screening of filter-collected nitro-PAHs is possible by laser desorption/ionization time-of-flight MS for samples with too little extractable material (Bezabeh, 1999). Direct GC/MS analysis of polar organic compounds is hampered by the long retention times of these compounds. A common method to handle organic acids is by diazomethane derivitization to convert them to their methyl ester analogues, which are amenable to analysis by GC/MS (Schauer et al., 1999a). Prior spiking with a deuterated acid is used to ensure that the reaction goes to completion. In a similar vein, Kawamura, Ng, and Kaplan (1985) analyzed volatile organic acids from motor vehicle exhaust by derivatizing them to p-bromophenacyl esters. Another procedure gaining use is to convert organic compounds with hydroxyl and carboxyl groups to the corresponding trimethylsilyl ethers and esters using the reagent bis-(trimethylsilyl)trifluoracetamide. Nolte, Schauer, Cass, and Simoneit (2002) used this approach to search for polar organic compounds that might serve as tracers for specific emission sources. 2.3. EC/OC analysis Thermal analysis of filter-collected PM does not provide chemical speciation, but the technique is widely used in ambient and source sampling studies to distinguish OC from EC, the latter being loosely equivalent to black carbon and soot. Basically, the method oxidizes carbon in the sample to CO2 , and quantifies the latter either directly with non-dispersive infrared detection, or by conversion to CH4 for flame ionization detection (Cadle, Groblicki, & Stroup, 1980; Huntzicker, Johnson, Shah, & Cary, 1982). The oxidation occurs in two stages. First, the sample is heated in an inert atmosphere, typically in stages to > 600 ◦ C, to volatilize organic material, which is subsequently oxidized and registered as OC. Second, the remaining sample is again heated, but this time in an oxygen environment, and the evolved CO2 is attributed to EC. For diesel exhaust PM, Japar et al. (1984) found the thermal method to be in good agreement with the soluble organic fraction (SOF) determined by solvent extraction, but the EC/OC split is now recognized to be operationally defined. The problem is that a certain fraction of OC can be pyrolized during the first stage OC analysis. This can be monitored as a darkening of the filter by the reflectance (or transmittance) of a He–Ne laser beam. Then the initial portion of the second stage, until the filter returns to its initial reflectance (transmission), is assigned to OC instead of EC. But the implementation of this correction, and the temperature levels specified during the two heating stages, differ from one instrument to another and lead to operational differences in the EC/OC split. The two principal protocols are the Interagency Monitoring of Protected Visual Environments (IMPROVE) (Chow et al., 1993) and NIOSH (1996). The differences in these methods are examined in depth by Chow, Watson, Crow, Lowenthal, and Merrifield (2001). It is possible to automate the thermal EC/OC method, but not with sufficient time resolution for transient emission measurements. To overcome this, Moosmüller et al. (2001) explored the possibility of using a combination of realtime instruments to achieve the same purpose. They tested five instruments including a photoacoustic instrument, a nephelometer (TSI DustTrak), an aethalometer (Anderson), a smoke meter (light extinction), and a tapered element oscillating microbalance (TEOM) to measure diesel PM over the federal test procedure (FTP) drive cycle. Photoacoustic detection was found to correlate well with the thermal/optical EC mass, as also concluded by Arnott et al. (2005). Fast response OC was not possible directly, but could be obtained as a difference between TEOM and photoacoustic measurements if the emission levels are moderate to high. 2.4. Motor vehicle PM composition—roadway tunnel studies The combination of the Arab oil embargo in 1973 and the 1977 Clean Air Act Amendment contributed to a flurry of diesel PM emissions research. A second wave of interest arose in the mid-1990s with the reported associations of daily mortality and morbidity to ambient PM levels. A complete review of this literature is beyond the scope of this paper; however, it is instructive to survey some of the work done over this time span. A comprehensive study of real-world motor vehicle emissions in the Allegheny and Tuscarora Mountain Tunnels was carried out between 1970 and 1979 by Pierson and coworkers (Pierson & Brachaczek, 1983, and references

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1087

therein). The PM samples were subject to a variety of chemical analyses, including atomic absorption, inductively coupled plasma atomic emission spectroscopy, XRF, ion chromatography, neutron activation and Soxhlet extraction, that allowed attribution of the PM to tailpipe, tire wear, brake wear, and resuspended soil sources, with only 3.5% remaining of unknown origin. Analysis of these emissions relative to strong variations in traffic composition permitted the PM emissions to be further resolved between gasoline and diesel vehicles. About 80% was identified as being carbonaceous, primarily from the 10–15% HDD traffic, 5% as sulfate, 10% as inorganic material from fuel and oil, 1% from tires, and 10% as soil dust. Additional chemical analysis at the Allegheny tunnel by Szkarlat and Japar (1983) showed that 24% of the vehicle PM was extractable material and 75% of the total mass was carbon, with ∼ 75% of this being EC. Alkanes dominated the extractable material, suggesting unburned fuel and lube oil as the main sources. Roadway tunnel studies, such as this, provide a convenient means to sample a large number of vehicles, albeit with the disadvantage of a narrow range of speed and load. Studies by Weingartner, Keller, Stahel, Burtscher, and Baltensperger (1997) at the Gubrist tunnel near Zürich in 1993, Kirchstetter, Harley, Kreisberg, Stolzenburg, and Herring (1999) at the Caldecott tunnel east of San Francisco in 1997, Allen, Mayo, Hughes, Salmon, and Cass (2001) also at the Caldecott tunnel in 1997, and Geller, Sardar, Phuleria, Fine, and Sioutas (2005) at the Caldecott tunnel in 2004 incorporate varying degrees of chemical analysis, including EC/OC, inorganic ions, and elemental analysis, with size resolved composition in the latter two cases. These studies arrive at consistent conclusions that carbon (elemental and organic) comprises the largest component of PM mass, HDD vehicles emit ∼ 20 times more PM mass per mass of fuel burned than gasoline vehicles, the EC fraction is higher for diesel derived PM (50–70%) than from gasoline engines (30–40%), and sulfate represents a small fraction of PM mass (< 5%). And taken together they indicate a decrease in PM emissions over time, consistent with tightening regulations and engine/ aftertreatment improvements. Much of the early work on organic speciation was aimed at PAHs, partly because of their designation as hazardous air pollutants, but also as a potential source tracer for gasoline vehicles after the phase-out of leaded gasoline. Chemical composition studies from the pre-catalyst era, both tunnel studies and individual vehicle testing, were reviewed by Daisey, Cheney, and Lioy (1986) as part of an assessment of using PAHs for source apportionment modeling. The situation changed after introduction of catalytic converters for gasoline fueled vehicles. Absolute concentrations of PAHs measured in the Baltimore Harbor Tunnel by Benner, Gordon, and Wise (1989) in 1985/1986 were factors of 5–10 lower than recorded a decade earlier in the same tunnel by Fox and Staley (1976), presumably from fleet turnover to a high percentage of catalyst equipped gasoline vehicles. Benner et al. (1989) report particulate PAH emission rates, averaged over the vehicle fleet, ranging from 8 g/km each for fluoranthene and pyrene to ∼ 2 g/km for a variety of other PAHs including benzo[a]pyrene and chrysene. The later work of Miguel, Kirchstetter, Harley, and Hering (1998) in the Caldecott tunnel took advantage of bore 2 being restricted to only light-duty vehicles to assign separate emission factors for the diesel and gasoline fleets. On a per kilogram mass of fuel used, diesel PAH emissions were about 60 times higher than from gasoline vehicles for fluoranthene and pyrene, but became comparable for heavier PAHs, such as benzo[k]fluoranthene. Averaged over the vehicle mix in the Baltimore Harbor Tunnel, these emissions are consistent with the results of Benner et al. (1989). Venkataraman, Lyons, and Friedlander (1994) and Miguel et al. (1998) reported size distributions of the particulate PAH and found the majority to reside in the region of 100 nm, near the maximum of the soot, or accumulation, mode of vehicle exhaust PM. Fraser, Cass, and Simoneit (1998, 1999) studied a wider variety of vapor and particle phase organic compound emissions, including petroleum biomarkers such as hopanes and steranes, in the Van Nuys Tunnel in September 1993. Using organic tracer techniques, they attributed from 7.5% to 18.3% of the ambient PM in the Los Angeles air basin to motor vehicle emissions. This site to site variation was comparable to the time of day variation where 19.1% of the PM was of vehicle origin during the morning rush hour, but 6.3% during the afternoon. 2.5. Motor vehicle PM composition—laboratory studies While roadway tunnel studies provide a convenient means to sample PM emissions from a large number of vehicles, the range of vehicle operation is too small for developing emission inventories and models. The alternative approach is dynamometer testing using one of a variety of drive cycles, e.g., US Federal Test Procedure (FTP), New European Drive Cycle (NEDC), California Unified Cycle (also known as the UCDS) (detailed descriptions of drive cycles are available at Dieselnet, 2007). Use of engine or chassis dynamometers permits tight control of engine and aftertreatment

1088

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Fig. 4. Material balance for the organic composition of motor vehicle PM in the Van Nuys Tunnel, Los Angeles (reprinted with permission from Fraser, M. P., Cass, G. R., Simoneit, B. R. T. Atmos. Environ. 33, 2715. Copyright Elsevier 1999).

parameters, but has two drawbacks. First, accurate models and inventories require testing a large and representative set of vehicles. Second, dynamometer testing requires artificial dilution of engine exhaust that might not reflect realworld conditions. For conventional diesel engine exhaust, where PM is dominated by the combustion-formed soot, the distinction is likely minor, but it may be more important for new clean burning vehicles. Much of the PM collection from vehicles in the 1970s was carried out by a combination of filters and condensation of the exhaust. These are reviewed by Daisey et al. (1986) along with a number of other source profiles. The PAH ratios, relative to benzo[e]pyrene, were found to agree to within a factor of 3 amongst the published gasoline vehicle (non-catalyst) studies, but the diesel PAH data were too limited to reach a conclusion about the use of these ratios to distinguish diesel from spark ignition vehicle emissions. In the mid-1980s Cass and coworkers began a comprehensive source sampling campaign to characterize PM emissions in the Los Angeles basin, including two series of motor vehicle measurements. These measurements utilized a dilution tunnel sampler (Hildemann et al., 1989), that incorporated a residence time chamber to allow semivolatile organics to partition amongst the gas and particle phases, and cyclones to give a well defined upper particle size. The first series of tests included six non-catalyst and seven catalyst equipped spark ignition vehicles, and two HDD trucks, with model years ranging from 1965 to 1987 (Hildemann, Markowski et al., 1991; Hildemann, Mazurek, & Cass, 1991; Rogge, Hildemann, Mazurek, & Cass, 1993). The second series examined two 1995 vintage medium duty diesel trucks (Schauer et al., 1999a) and nine catalyst equipped gasoline vehicles from the 1981–1994 time frame (Schauer, Kleeman, Cass, & Simoneit, 2002). Each of these studies quantified over 100 compounds including n-alkanes, nalkanoic acids, benzoic acids, benzaldehydes, PAHs, oxy-PAHs, steranes, pentacyclic triterpanes, and azanapthalenes. These compounds originate from unburned fuel and lube oil components, their partially oxidized products, and metallic ash from fuel and oil additives. A key feature of this work is the consideration of material balance, as illustrated in Fig. 4. Depending on source, diesel versus gasoline vehicle, the extractable and elutable fraction of exhaust PM ranges from ∼ 20% to 90% (∼ 90% in the roadside tunnel example of Fig. 4). Of this extractable and elutable material most, ∼ 85%, is unresolved by GC, termed the “unresolved complex mixture”. About 70% of the resolved mixture cannot be identified; thus, the roughly 100 identified compounds constitute less than 5% of the organic material in the PM. Yet, these species provide useful information relating exhaust PM emissions to engine and fuel type. The PM composition emitted from catalyst versus non-catalyst equipped gasoline vehicles showed some significant differences. Non-catalyst cars exhibited roughly 25-fold higher emissions of PAHs than catalyst equipped vehicles, in rough agreement with the difference in the PAH concentrations in the Baltimore Harbor Tunnel recorded by Brenner et al. (1989) versus earlier measurements by Fox and Staley (1976). Catalyst equipped vehicles, though, had higher emissions of n-alkanoic acids and benzoic acids, likely promoted by the catalyst aftertreatment. The diesel trucks exhibited PAH emissions intermediate between the catalyst and non-catalyst gasoline vehicles. Whereas dimethyl phenanthrenes and anthracenes predominate as the three or higher ringed PAHs in diesel fuel, unsubstituted analogues are more prevalent in the exhaust (Schauer et al., 1999b). Fossil petroleum markers, hopanes and steranes, were found in the emissions of all the test vehicles, but as these are not present in either gasoline or diesel fuel,

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1089

they reflect the relative oil consumption of the test vehicles. Fraser et al. (1999) were able to model the vehicular contribution to the Los Angeles atmosphere based on these tracers, but not separately those from gasoline versus diesel vehicles. Motivated by potential health benefits, an important application for the chemical analysis of PM has been to evaluate “environmentally friendly” fuels. Westerholm and Egebäck (1994) describe fuel and exhaust aftertreatment impact of PM composition recorded as part of the Swedish Urban Air Project. For non-catalyst gasoline vehicles, PAH emissions followed the their content in the gasoline, but they were dramatically decreased by the catalyst. Two diesel vehicles were tested with eight different fuels, including blends with kerosene, use of a cetane enhancer, and low sulfur fuel. Important fuel parameters were found to be density, 90% distillation point, final boiling point, specific energy, total aromatics, di-aromatics, tri-aromatics, and PAH contents. A multivariate statistical analysis supported a relationship between the PAH contents of the diesel fuel and their emissions in the exhaust (Westerholm & Li, 1994). More recently, Westerholm et al. (2001) compared Swedish Environmental and European reference fuels and found particle-bound PAH and 1nitropyrene emissions to be reduced by 88% and 98%, respectively, while semivolatile PAHs and 1-nitropyrene were reduced by 77% and 80%. Characterization of PM from heavy-duty vehicles includes the work of Lowenthal et al. (1994), which examined 13 heavy duty trucks and buses, finding the PM to consist of 80–90% organic and elemental carbon and up to 14% sulfate. Detailed PM chemical analysis was part of the 1999–2003 ECD Technology Validation Study into the effects of DPF aftertreatment and ultra-low sulfur diesel fuel on their emissions from trucks and buses (Lev-On, LeTavec, Uihlein, Alleman et al., 2002; Lev-On, LeTavec, Uihlein, Kimura et al., 2002; Lev-On & Zielinska, 2004). The subset of vehicles selected for chemical analysis included a diesel school bus, two diesel grocery trucks, a diesel transit bus, and two compressed natural gas (CNG) transit buses. Ultra-low sulfur (also Fischer–Tropsch) fuel in combination with the DPF resulted in ∼ 95% reductions in PM mass emissions, and lower levels of 2, 3, and 4+ ring PAHs than emitted by the CNG fueled vehicles in the study. Nitro-PAH emissions were close to detection limits. A lesson from this study was the need for great care in the design and execution of PM sampling and chemical analysis at low emission levels. Testing in-use vehicles is important, as it forms a basis for emission inventories. Sagebiel et al. (1997) examined the exhaust PM from 23 high emitter gasoline vehicles recruited for inspection/maintenance checks, including six visible smokers. OC dominated the emissions, at ∼ 80% for both the smoking and non-smoking vehicles. Total PAHs (gas plus particle phase) were higher in the smoking vehicles, but inorganic ion emissions were low, except in one case where sulfate storage might have occurred. But, the IM240 test cycle used in this work made the data of limited value for inventories. Later, Cadle et al. (2001) compared in-use gasoline vehicle PM emissions for three drive cycles. They observed a small increase in OC fraction from 36% for the FTP to 45% for the REP05 drive cycle, but particulate anions, predominantly sulfate, increased from ∼ 2% to ∼ 10%. Realistic inventories require emissions data as a function of ambient temperature. A large fleet of Denver area gasoline and diesel vehicles was tested in the summer of 1996 and again in the winter of 1997, including EC/OC, sulfate, nitrate, and elemental analysis (Cadle et al., 1999). The OC fraction depended strongly on engine type, but less on temperature. Gasoline vehicle PM averaged 74% OC in summer versus 65% in winter, whereas the corresponding OC fractions for diesel vehicles were 45% in summer versus 35% in winter. Sulfate emissions were 0.45 and 3.5 mg/mi for gasoline and diesel vehicles, less than expected from EPA’s mobile emissions model. Zielinska, Sagebiel, McDonald et al. (2004) included PAH, hopane, and sterane speciation for a small fleet of in-use gasoline and diesel vehicles tested at 22 ◦ C and −1 ◦ C over the Unified Driving Cycle. Temperature had a strong impact on the emissions of these species, especially PAHs. The higher molecular weight PAHs were present at 5–10 times higher levels at −1 ◦ C, likely because of less efficient fuel combustion. Hopanes and steranes increased ∼ 3fold at low temperatures for the gasoline vehicles, but diesel vehicle emissions of these species were insensitive to temperature. Off-line chemical analysis of filter- and PUF-collected material remains an important part of current day PM emissions testing from combustion engines. Examples of recent multi-vehicle programs with a substantial chemical speciation component, some still in the process of writing up results, include: ACEA Programme on the emissions of fine particles from passenger cars, including SOF and sulfate analysis from 3 diesel and 4 gasoline vehicles (ACEA, 2002); Coordinating Research Council (CRC) E-55/E-59 program, with speciation for 13 heavy heavy-duty diesel trucks (HHDDT) in phase 1 and 5 in phase 2 (Gautum & Clark, 2003); Gasoline/Diesel PM Split Study, with 59 light-duty vehicles (2 diesel) and 34 heavy-duty diesel vehicles (NREL, 2007), and the Kansas City Light-Duty Vehicle Emission Study (CRC project E-69). In October 2005 the CRC initiated project E-75, the Compilation of Diesel Emissions Speciation Data to generate a database of results from these and other studies.

1090

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

3. Next generation diesels To assess the impact of “clean diesels” on air quality and health, it is important to understand what if any impact the new technology has on the nature of the PM emissions. The challenges posed to chemical characterization of the PM is best understood after first summarizing the numerous differences of next generation diesels as compared to conventional diesels. Zelenka, Egert, and Cartellieri (2000) discuss the general approach to “clean diesel” technology in terms of three major steps: (a) hardware changes to the engine, (b) incorporation of a DPF, and (c) NOx reduction. To this we can add a fourth component, namely changes to fuel and lubricant composition. The objective is to reduce regulated emissions to levels that meet the upcoming EPA Tier II and EU Stage 5 and 6 levels. With respect to PM, the main impact is the 90 + % reduction in mass emissions provided by the DPF. However, the DPF, as well as the other engine and aftertreatment changes, can in principle also affect the size, density, and composition of the particles. This implies the need for analytical techniques with considerably greater sensitivity than previously used to investigate conventional diesel PM composition. 3.1. Advances in diesel engine technology The major engine modifications include migration from indirect to direct fuel injection, the use of a high pressure, common rail, fuel system (Guerrassi & Dupraz, 1998), and the use of cooled exhaust gas recirculation (EGR) (Ladommatos, Abdelhalim, Zhao, & Hu, 1998). High pressure direct injection systems allow adding pilot and post fuel pulses to the main injection event during each engine cycle. These permit control of the noise and harshness of the engine, improve fuel economy, and reduce emissions (Park et al., 2004). The pilot fuel pulse shortens the ignition delay for the main fuel charge, reducing noise and NOx . Post injection can help reduce PM emissions, and it is used as a source of hydrocarbons for lean NOx reduction. It is also possible that such injection strategies can alter the elemental carbon to organic carbon ratio (EC/OC) in the exhaust stream. EGR is employed to regulate the relative concentrations of NOx and PM in the exhaust. Ladommatos et al. (1998) examine three mechanisms by which EGR can affect diesel combustion: dilution of the available O2 , increased heat capacity of the combustion mixture, and enhanced chemistry from the dissociation of CO2 and H2 O at combustion temperatures. Using controlled replacement of O2 by CO2 , they show that EGR acts mainly by lowering the O2 concentration. This leads to a lower local peak temperature in the combustion zone and, therefore, lower NOx levels. Likewise, by lowering the efficiency of soot oxidation, O2 dilution is a primary factor in the increase in PM levels brought about by EGR. Cooling the exhaust permits the use of high EGR levels without excessively heating the intake fuel–air charge. Zheng, Reader, and Hawley (2004) give a review of advanced EGR concepts, but caution that heavy use of EGR could degrade the energy efficiency and mechanical durability of the engine. The high soot concentrations resulting from high EGR levels can also lead to EGR cooler fouling and valve deposits and, thereby degrade engine performance. These engine changes are likely to have a secondary effect on PM composition. At post combustion temperatures of > 1000 K the PM consists of soot and ash, with organic and sulfate material almost entirely in the gas phase. While particle number concentration and size depend on injection pressure, post injection fuel pulses, and EGR, the composition of the particles exiting the combustion cylinder should not be much affected. The principal effect will be the extent to which the fuel injection and EGR strategies alter hydrocarbon and sulfate emissions that subsequently condense onto the particles, or nucleate, as the exhaust cools. 3.2. DPF/oxidation catalyst Although changes in engine technology may be able to meet certain interim emission regulations, it is widely accepted that diesel vehicles can only meet the circa 2010 tailpipe standards with exhaust aftertreatment (see reviews by Khair, 2003; Twigg, 2007; van Setten, Makkee, & Moulijn, 2001). The overall system, still under development, requires an oxidation catalyst to remove hydrocarbons and CO, a DPF to trap PM, and either an urea SCR catalyst or catalyzed NOx trap to remove oxides of nitrogen. Fig. 5 illustrates a wall-flow DPF commonly used in diesel vehicle applications. These devices, alone and in concert, have potentially larger ramifications on PM composition than the changes to the engine itself.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1091

Fig. 5. Wall-flow diesel particulate filter (Courtesy Corning, Inc.). Alternate channels in a ceramic monolith are blocked, forcing exhaust to flow through the porous walls and trapping the soot particles.

It is difficult to examine in isolation the effects of the engine, each of the aftertreatment components, and the fuel on PM emissions. The impact of the oxidation catalyst depends also on factors such as the fuel composition. On one hand it removes hydrocarbons that could otherwise condense on soot particles or nucleate as the exhaust subsequently cools. But on the other hand, it oxidizes SO2 to SO3 . In the presence of water vapor this yields sulfuric acid, which can promote nucleation or condense on soot. The result is a synergistic effect between the catalyst and fuel sulfur level that alters the chemical makeup of diesel engine PM, and affects the particle size distribution. Light duty diesel vehicles that have an oxidation catalyst or are operated with conventional diesel fuel (350 ppm sulfur), but not both, typically exhibit a single lognormal accumulation mode of particle emissions. But the combination of the two adds a nucleation mode to the emissions (Maricq, Chase, Xu, & Laing, 2002; Vogt, Scheer, Casati, & Benter, 2003), which the particle MS work by Scheer et al. (2005) suggests is dominated by sulfate. In other engines the reverse may be true. Tandem nano-DMA/thermal desorption particle MS measurements of HDD PM by Tobias et al. (2001) indicate that the nucleation particles are predominantly hydrocarbon in origin, likely from lube oil and heavy end fuel components. This distinction between light and HDD emissions is discussed in more detail below when addressing the question of the nucleation mode. The addition of a DPF increases the complexity in the engine/aftertreatment system. The DPF substantially lowers PM mass emissions, but its effect on particle number is ambiguous. This is because the mass is dominated by the soot accumulation mode, which is efficiently trapped, but the number can be dominated by nucleation particles formed downstream of the DPF, a process that is very sensitive to engine operating conditions (Abdul-Khalek, Kittelson, & Brear, 1998; Holmén & Ayala, 2002; Liu, Verdegan, Badeau, & Sonsalla, 2002). Recently, Vaaraslahti, Virtanen, Ristimäki, and Keskinen (2004a, 2004b), have noted with respect to a HDD engine that low load conditions lead to a nucleation mode without a DPF, but that when a catalyzed DPF is added, the nucleation mode occurs at high load. Based on indirect arguments, they suggest that the nucleation mode arises from hydrocarbons when a DPF is not present, but from sulfate with the DPF in place. A further complication is that particle traps need to be regenerated periodically to avoid excessive back pressure on the engine. This introduces a new set of circumstances for particle formation that needs to be examined. Again, there are possible interactions with fuel structure, where high sulfur fuel can yield substantial numbers of nucleation particles during regeneration (Guo, Xu, Laing, Hammerle, & Maricq, 2003). Clearly, chemical analysis of the PM would benefit our understanding of DPF function, help optimization of aftertreatment systems, as well as provide insight into the potential environmental impact of the remaining emissions. 3.3. NOx aftertreatment Because diesel NOx aftertreatment is still under development, its impact on PM composition is not well understood. Two paths are currently pursued, lean NOx traps and urea SCR. Lean NOx traps have the advantage that their use does not require any fuel infrastructure changes. However, there are a number of drawbacks including sulfur poisoning, thermal durability and an energy penalty to regenerate that trap, as discussed by Cheng, Cavataio, Belanger, Hoard, and Hammerle (2004), that remain to be overcome. The relatively low NOx conversion efficiencies necessitate the use of high levels of EGR, while regeneration is accomplished during periods of rich engine opera-

1092

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

tion (West et al., 2004). These are atypical operating conditions for conventional diesels that can potentially influence particle composition. The technical advantages of urea SCR over lean NOx traps are discussed by Lambert, Hammerle, McGill, Khair, and Sharp (2004). They include a wider operating temperature range, a lower fuel economy penalty, higher sulfur tolerance, and higher durability. However, it requires periodic replenishment of urea, held in a reservoir. And there are additional emission issues to contend with, such as ammonia slip and N2 O formation. Besides ammonia, urea yields other thermal decomposition products including cyanate ion, biuret, cyanuric acid, ammelide, ammeline, and melamine (Ball, 2001). Except for ammonia, Sluder, Storey, Lewis, and Lewis (2005) did not detect these downstream of the SCR catalyst. However, they did find small amounts of HCN formed along with the ammonia, and trace amounts of volatile nitro-organic compounds, nitromethane, produced upstream of the SCR catalyst. Both NOx aftertreatment methods are promising, but they are under development and our understanding of their implications on the PM composition is in its infancy. 3.4. Fuel/lubricant modifications There are two forces driving the search for alternative diesel fuels. One is directed at alleviating difficulties in achieving the upcoming emission standards, particularly for NOx . The effect might be indirect; for example, a lower sooting fuel could permit higher EGR levels that enable larger NOx reductions. The other interest is to enable greenhouse gas reductions via the use of biofuels (Knothe, Sharp, & Ryan, 2006; Mayer, Czerwinski, Wyser, Mattrel, & Heitzer, 2005). The major difference between current and future diesel fuels is the dramatically lower sulfur content, already available in parts of Europe, and that will be phased in from 2006 to 2010 in the US (Leister, 2003; US EPA, 2001). The sulfur reduction is aimed both at reducing PM emissions and as an enabler for aftertreatment technology that would otherwise be poisoned by sulfur or produce too much SO3 (Bardasz, Antoon, Schiferl, Wang, & Totten, 2004). Other efforts are aimed at changing the organic content of the fuel. This can be done either by introducing a new fuel, such as the Fischer–Tropsch fuel, which offers a very low aromatic content. Or it can take the form of a fuel additive, such as dimethoxy methane (DMM), ethanol, and other oxygenated fuel additives (Ball et al., 2005; Frank et al., 2004). Changes are also under consideration for engine lubricants (Whitacre, 2003). The main considerations are reductions in sulfur and ash. This primarily come from the “additive package”, which contains detergents and anti-wear compounds. At typical sulfur levels of 5000 ppm, the lube oil contributes only a small fraction relative to the sulfate emissions from present diesel fuels, but the oil contribution may become significant as low sulfur fuels are introduced. The primary concern about ash is that it is trapped along with soot in the DPF. But unlike the soot, it is not regenerated. It therefore accumulates in the DPF, limiting its lifetime. 3.5. Measurement challenges at low emission levels “New diesel” technology affects not only tailpipe emissions, but also the measurement protocols. Although emission regulations are not directed at particulate composition, it plays a role in the measurement method nonetheless. Hydrocarbon adsorption artifacts on filter media limit the performance of the regulatory gravimetric PM mass measurement at low emission levels (Chase et al., 2004; Khalek, 2005). And there is a potential for nitro-PAH artifact formation on the filter during PM collection due to the higher NO2 /NO ratios present in oxidation catalyst/DPF equipped diesel vehicles. Although the very low PM levels from “clean diesels” challenge the analysis of chemical composition, this knowledge would help define better approaches to PM emissions measurement. 4. Emerging tools for PM chemical analysis Diesel engine technology is not alone in undergoing fundamental change. The past decade and a half has seen the emergence of qualitatively new paradigms in diesel exhaust PM measurement (Burtscher, 2005). This includes chemical analysis of aerosol particles. Most notable is the application of on-line single particle and thermal desorption mass spectrometric techniques and the capability for real-time speciation that they bring. New off-line techniques have been introduced as well, for example, the possibility to probe functional group structure using near-edge X-ray absorption fine structure (NEXAFS). These advanced characterization techniques are the subject of this section.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1093

Fig. 6. Schematic view of the aerosol time-of-flight mass spectrometer (ATOFMS) (reprinted with permission of TSI, Inc.).

4.1. Single particle MS A comprehensive review of aerosol MS and its historical development is given by Suess and Prather (1999). There are two main types of aerosol mass spectrometers—those that rely on single particle laser ablation/ionization and those that utilize thermal desorption/electron impact ionization. These have sufficient differences in terms of ion generation and interpretation of spectra to warrant separate discussion. Single particle mass spectrometers are based on the following general design elements: an inlet to sample the aerosol into a high vacuum region, laser scattering to detect a particle, ablation/ionization of the particle, and time-of-flight measurement of the resultant ion masses. Various versions have been introduced, for example: particle analysis by laser mass spectrometry (PALMS) (Murphy & Thomson, 1995), the aerosol time-of-flight mass spectrometer (ATOFMS) (Prather, Nordmeyer, & Salt, 1994), and laser mass analysis of particles in the airborne state (LAMPAS) (Hinz, Kaufmann, & Spengler, 1994, 1996). To describe their operation, let us take the ATOFMS as an example. This instrument combines MS with aerodynamic particle sizing. A schematic is shown in Fig. 6. Earlier versions introduced particles through a converging nozzle (Gard et al., 1997), but current inlet systems incorporate an aerodynamic lens (Liu, Ziemann, Kittelson, & McMurry, 1995a, 1995b) that focuses the particles into a beam and increases sampling efficiency. As a particle exits the lens into the high vacuum of the mass spectrometer, collisions with gas molecules accelerate it to a terminal velocity dependent on its aerodynamic diameter and shape. The particle is sized by its transit time through a pair of continuous wave (cw) laser probe beams set a fixed distance apart. Its velocity is used to time the firing of a pulsed laser that volatizes and ionizes the particle as it enters the source region of a pair of back-to-back time-of-flight mass spectrometers (typically using a 266 nm quadrupled NdYAG laser or a 193 nm excimer laser at a fluence of 2×107 to 4×108 W/cm2 ). An electric field in the source sends the resultant positive and negative ions in opposite directions through the two reflectron time-of-flight tubes. All ions are given a fixed kinetic energy dictated by the electric potential at which they are born. Consequently, an ion’s arrival time at one of the two microchannel plate detectors establishes its mass. Current technology instruments can perform this size and mass analysis at particle rates of up to ∼ 10 Hz. Simultaneous measurement of a particle’s constituent positive and negative ions helps assign the particle’s chemical constituents (Noble & Prather, 1996). EC is typically characterized by the appearance of carbon cluster ion series of both polarities spaced by m/z 12. OC is readily identified by the presence positive ion fragments such as CH+ 3, m/z = 15; C2 H3+ , m/z = 27; CNH+ , m/z = 27; C2 H5+ , m/z = 29; COH+ , m/z = 29; CHNO+ , m/z = 43; C2 H3 O+ , m/z = 43; C3 H7+ , m/z = 43; C7 H7+ , m/z = 91; and so forth. Metals are recognized by peaks at the elemental and metal oxide m/z values. Inorganic aerosol components, such as sulfate and nitrate, are identified by their characteristic anion masses. A detailed discussion of how to interpret single particle mass spectra of organic compounds is given by Silva and Prather (2000).

1094

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Although mass measurement allows one to identify the chemical constituents of aerosol particles, assessing their concentrations using single particle MS proves more difficult (Suess & Prather, 2002). Given an isolated gas phase molecule, one could hope to calibrate its ionization efficiency at a given laser wavelength and fluence. But with the molecule embedded in a particle, the physical and chemical nature of the surrounding matrix has a strong influence on ionization efficiency, and this can vary from one particle to another. Laser targeting and pulse energy represent added sources of variability. This problem is analogous to the issue of recovery efficiency in the traditional off-line approach of collecting PM onto filters. Spiking the filters with deuterated internal standards solves the latter problem, but this is not possible with single particle analysis. Consequently, single particle MS has adopted a different approach. Particles are segregated into composition classes, each represented by a characteristic mass spectral pattern. The need to accomplish this for large numbers of spectra led to off-line statistical analysis of the mass data by methods such as principal components analysis (Hinz et al., 1996; Ro, Musselman, & Linton, 1991) and hierarchical cluster and discriminant analysis (Shattuck, Germani, & Buseck, 1991). More recently, Song, Hopke, Fergenson, and Prather (1999) applied neural net methods to establish particle classes. They iteratively assign spectra using adaptive resonance theory to decide which class, refining the class standard as spectra are added, and defining new classes when necessary. Tan, Malpica, Evans, Owega, and Fila (2002) examined a discriminant analysis method that could be trained to be sensitive or insensitive to specific chemical markers as needed to improve performance under conditions of poor spectral reproducibility. Single particle MS is amenable to a variety of adaptations. To an extent it is possible to overcome the limitations on quantifying particle constituents by using laser pulse energies sufficiently high to completely ablate and ionize the particle into its atomic constituents (Reents & Ge, 2000). Lee, Miller, Kittelson, and Zachariah (2006) used this approach to study nanoparticle formation from the combustion of ferrocene doped diesel fuel. While this approach cannot identify the molecular composition, it was used to study the particle size dependence of the iron to carbon ratio. In the opposite vein, Reilly, Gieray, Whitten, and Ramsey (1998) coupled laser-based particle ablation/ionization with an ion-trap mass spectrometer. This approach enables MS/MS capability to help identify complex chemical structures, in this case PAH components in diesel exhaust particles. It is also possible to use single particle MS to perform on-line EC/OC analysis. Okada et al. (2003) found that the ATOFMS assignment of particles to the EC and OC spectral classes correlated well with thermal EC/OC analysis of filter-collected diesel PM and with the expected behavior on engine operation, e.g., increasing OC with decreasing engine load. Ferge et al. (2006) assigned peaks in the series Cn Hy+ , with 0 y 3 to EC, and those with y > 3 to OC. But negative ion spectra could not effectively discriminate between EC and OC. After correction for interference from metal ion peaks, the ratios of EC-associated to OC-associated positive ion peak areas showed good agreement relative to the NIOSH method for EC/OC analysis of various sources of soot, including industrial, diffusion flame, and diesel engine. 4.2. Thermal desorption aerosol MS An alternative approach to aerosol MS is based on thermal desorption of PM and its electron impact ionization (Jayne et al., 2000; Tobias, Kooiman, Docherty, & Ziemann, 2000). Fig. 7 illustrates the design of Tobias et al. (2000). If desired, the aerosol is first sent through a differential mobility analyzer (DMA) to pre-select particle size. It then enters the mass spectrometer via an aerodynamic lens (Liu et al., 1995a, 1995b) that focuses the particles into a narrow, low divergence, beam. The beam passes through regions of differential pumping and impacts onto a resistively heated V-shaped molybdenum foil where volatile and semivolatile material is vaporized. The liberated molecules diffuse into an electron impact source operated at 70 eV and the resulting ions are extracted into a quadrupole mass spectrometer for mass analysis. Operation at a fixed, or set of fixed, m/z ratios allows real-time monitoring of chemical composition changes in an aerosol. Or given a continuous stream of particles, the quadrupole can be scanned to record the aerosol mass spectrum. This instrument is designed for particles in the range of 20–500 nm diameter at concentrations of 0.1.1 g/m3 . The Jayne et al. (2000) version (Aerodyne Research Inc.) utilizes similar aerosol inlet and mass detection schemes, but adds a time-of-flight region to measure particle size. Particles exit the aerodynamic lens in a supersonic expansion that accelerates the particles to terminal velocities dependent on their vacuum aerodynamic diameter. The particles next pass through a chopper wheel that defines the start of a time-of-flight region. It ends at the closed end of a resistively heated tube where flash vaporization of the non-refractory particle constituents and mass spectral detection take place.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

computer

1095

quadrupole mass spectrometer

power supply

aerodynamic aerosol source lenses skimmers nozzle particle relaxation tube

PID

orifice thermocouple LN2 Dewar

vaporizer (rotated 90°) turbo pumps

Fig. 7. Schematic view of the thermal desorption particle beam mass spectrometer (TDPBMS) (reprinted with permission of P. Ziemann, U. C. Riverside).

The quadrupole setting defines the ion m/z ratio, whereas the time-of-flight establishes the size of the particle from which it originated. Very recently DeCarlo et al. (2006) introduced a new high resolution version of the Aerodyne aerosol mass spectrometer. The major change replaces the quadrupole with a time-of-flight mass analyzer that operates in two modes. As a single reflectron it provides a resolving power of 2100 at m/z 200 and a 1 min average detection limit of 0.04 g/m3 . Switched into the double reflectron mode the resolving power increases to 4300, but at ∼ 10 times lower sensitivity. This resolution is sufficiently high to allow separation of different molecular species having the same nominal m/z value. Aiming at chemical analysis of nucleation mode particles in the range below 10 nm, Voisin, Smith, Sakurai, McMurry, and Eisele (2003) devised a thermal desorption chemical ionization mass spectrometer. Particles are collected onto a metal filament by electrostatic precipitation. After a suitable averaging time the filament is moved into the source region of a triple quadrupole mass spectrometer where the PM is vaporized by resistive heating and ionized by charge transfer reactions. The triple quadrupole arrangement assists chemical identification by MS/MS analysis of ion fragmentation patterns. Single particle versus thermal desorption mass spectrometers offer interesting tradeoffs. The former ideally identifies the chemical constituents of each individual particle. But relative concentrations are poorly determined because the laser ablation/ionization efficiency depends on the particle’s chemical and physical structure. In contrast, electron impact ionization is reproducible and yields quantitative data, but only for those particle species that can be thermally desorbed. The single particle instrument’s strength lies in its ability to classify heterogeneous aerosols according to chemical signatures on a particle by particle basis. An application relevant to diesel PM is the source apportionment of urban aerosols (Gross et al., 2000, 2005; Noble & Prather, 1996; Vogt et al., 2003). When a steady stream of particles is available, the thermal desorption approach may be of more benefit. Its ability to determine relative concentrations allowed Tobias et al. (2001) and Sakurai, Tobias et al. (2003) to conclude that nucleation mode particles selected from heavy-duty diesel exhaust originate primarily (95%) from lube oil. Interestingly, the same approach shows light-duty diesel vehicle exhaust is typically dominated by sulfate (Scheer et al., 2005; Schneider et al., 2005). 4.3. Multiple laser approach In single particle MS, the laser performs two functions: it vaporizes the particle and ionizes the liberated molecules. High laser power is needed to accomplish both tasks, which risks substantial fragmentation of the parent ions. Reproducible ionization is further hampered by matrix effects. With two lasers it is possible to separate these tasks. One laser vaporizes the particle, leaving the second laser to ionize gas phase molecules under more reproducible conditions. Cabalo, Zelenyuk, Baer, and Miller (2000) and Woods, Smith, Dessiaterik, Baer, and Miller (2001) investigated a CO2

1096

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Fig. 8. Apparatus for laser induced breakdown spectroscopy.

laser vaporization/UV laser ionization scheme and studied how laser power and timing affect the mass analysis of liquid organic aerosols. Application to diesel particles to date has been on collected samples, as opposed to on-line analysis. NIST diesel PM (SRM 1650) was examined for PAHs by Hankin and John (1999), in this case using two UV lasers. Two lasers also make possible selective examination of internally mixed particles; thus, Zhan, Voumard, and Zenobi (2005) used one laser to desorb the adsorbed material on particles collected from rural, industrial, and roadway areas, and the second laser to ionize this material for mass spectral analysis. 4.4. DMA/ICP-MS ICP-MS is a highly sensitive and accurate method for quantifying the elemental composition of PM. Often used for analyzing filter-collected samples, e.g., the examination of roadway metal emissions by Lough et al. (2005), efforts to develop on-line versions have also been reported. Okada, Yabumoto, and Takeuchi (2002) combined differential mobility and ICP-MS analysis for in situ size and composition measurement of metal particles relevant to semiconductor processing. By operating the DMA with an argon sheath gas, it acts also as a gas converter (Myojo, Takaya, & Ono-Ogasawara, 2002). Thus, the monodisperse particles at the outlet find themselves in an argon carrier gas and they can be introduced directly into the plasma torch where the argon plasma evaporates and ionizes the particle constituents. Composition analysis of particles in the 5–40 nm range was possible at concentrations of ∼ 105 cm−3 . If extended to ∼ 100 nm this method offers the possibility of on-line investigations into the origin of metal ash in diesel PM, for example from lube oil additives, engine wear, or aftertreatment deterioration. It would be a useful compliment to thermal desorption MS’s ability to speciate semivolatile PM components. 4.5. Laser breakdown spectroscopy Another approach to on-line elemental analysis of aerosol particles is via laser induced breakdown spectroscopy (LIBS). The applicability of this technique to aerosols is reviewed by Hahn and Lunden (2000). Fig. 8 illustrates the basic method. An intense laser pulse (10 ns, 50–400 mJ, NdYAG at 1064 nm) is focused through a pierced mirror into the aerosol chamber, where it generates a microplasma with a characteristic volume on the order of 10−4 cm3 and temperature of 104 K (Hahn & Lunden, 2000). The plasma dissociates matter within this volume into its atomic components, leaving many in highly excited energy states. Atomic emission is collected by one lens and focused by a second lens onto the fiber optic input of a UV/visible spectrometer, where it is dispersed versus wavelength. Emission line positions and intensities indicate the species present and their relative concentrations in the PM. Setting a suitable threshold level makes it possible to distinguish laser shots that hit particles from misses. LIBS has seen success as a continuous environmental monitor for toxic metals, for example, from stack emissions (Zhang, Yueh, & Singh, 1999; see also references in Hahn & Lunden, 2000). Recently Hybl, Lithgow, and Buckley

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1097

(2003), and Hybl, Shane, Berry, and Jordan (2006) have applied the method for bioaerosol detection, looking at the possibility of using element ratios to discriminate biological agents from naturally occurring aerosols and using LIBS to augment laser induced fluorescence detection. Applications to the combustion area include the work of Ottesen, Wang, and Radziernski (1989) and Zhang, Singh, Yueh, and Cook (1995), which show the technique to have high sensitivity for the mineral content of coal particles and give results consistent with XRF measurements. Lombaert et al. (2004) used LIBS to determine the semiquantitative composition of diesel PM, showing Fe, Mg, Ca, Cu, and Zn to be the major metal species. Fe can be attributed to engine wear, and Ca and Zn are ubiquitous lube oil additives. 4.6. Advanced X-ray methods The wider availability of synchrotron light sources has opened up the opportunity to apply advanced X-ray techniques to help elucidate PM structure and composition. These include small angle X-ray scattering (SAXS), wide angle X-ray scattering (WAXS), and NEXAFS. The scattering methods yield structural parameters. Work by Braun, Shah, et al. (2005) illustrates their application to diesel PM, in this case to examine the influence of oxygenated diesel fuel. SAXS data reveal multiple structural classes in the PM. The most prominent ones, with diameters ∼ 35 and ∼ 270 nm, correspond to primary particles and aggregates, but substructures can also be distinguished. The WAXS data identify the size of crystalline regions within the primary particles, on the order of 1 nm, and their lattice spacings. In contrast, NEXAFS provides chemical information. As described in the review by Myneni (2002), it is a sensitive probe of functional group chemistry that can be applied to characterize organic compounds in the environment. The irradiation of a material by soft X-rays, < 4000 eV, excites atomic core electrons to valence molecular orbitals, as well as into the continuum. The bound state transitions make up the near edge absorption fine structure observed spectroscopically. The core atomic levels are largely independent of molecules structure, but the final valence levels depend on the local chemical bonding. Electron energy loss spectroscopy (EELS) represents a very similar chemical probe, but using an electron beam instead of X-rays as the excitation source. It has the advantage of sub-nanometer spatial resolution, but the potential for the high energy electron beam to alter sample chemistry is a drawback (Braun, Huggins, et al., 2005). Disadvantages for NEXAFS include the need for relatively large samples and its off-line nature. The carbon K-edge absorption at ∼ 280 eV exhibits a fine structure dependent on the functional groups that are present in the sample. C 1s → ∗ transitions are relatively intense and can be resolved, with C = C and C ≡ C transitions typically appearing at ∼ 285 eV, C = O at 286–287 eV, and COO at ∼ 289 eV (Myneni, 2002). C 1s → ∗ transitions near 288 eV can be used to identify C–H groups and those at 291 eV are associated with C–OH in alcohols. Such characteristic spectral positions make it possible to deconvolve individual functional group contributions to complex species (Stöhr, 1992). However, care is needed since specific transition energies vary from one molecule to another and the peak widths depend on the extent of vibrational coupling. For this reason, the study of model compounds becomes important to help assign spectra of complex materials (Francis & Hitchcock, 1992). Including nitrogen and oxygen K-edge absorption spectra can help confirm assignments. Examples of complex matter analyzed by NEXAFS include the work of Cody et al. (1995) and Cody, Ade, Wirick, Mitchell, and Davis (1998) on various forms of coal. Liptinite and huminite could be distinguished by the abundance of aliphatic and carboxylic carbon present in the former, as compared to the greater aromatic content of the latter. Braun et al. (2004) combined NEXAFS with X-ray microscopy to distinguish regions of graphitic soot from hydrocarbon material in diesel PM. Fig. 9 illustrates the NEXAFS spectrum of diesel soot and compares it to wood smoke. The indicated spectral assignments are approximate, since the detailed chemical makeups of these soots are unknown. Nevertheless, the significant difference between the diesel and wood smoke spectra illustrates the potential use for this technique to distinguish soot emission sources. 4.7. Neutron scattering Common features of coal, soot, and related materials are the high C/H ratio and strong UV/visible absorption cross section. This hinders the application of many spectroscopic tools to their study. The relatively low hydrogen content thus exacerbates the difficulties in studying this aspect of these materials. On the other hand, inelastic neutron scattering (INS) cross sections for hydrogen atoms are an order of magnitude larger than for atoms like C, N, and O, making this a useful tool to study CH groups in coals and soot. Fillaux, Papoular, Lautié, and Tomkinson (1995) compared INS spectra from 30 to 4000 cm−1 for four coal samples with rank from subbituminous to anthracite. The main feature is

1098

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

COOH

diesel PM

diesel fuel

0.5

284

286 phenol C-OH

2.5

2.0

290

292

294

wood smoke

C=O

1.5

1.0

C=C

0.5

0.0 282

288

COOH

0.0 282

NEXAFS absorption (a.u.)

phenol C-OH

C=C b-q C=O

1.0

b-q C=C

NEXAFS absorption (a.u.)

1.5

284

286

288 290 Energy (eV)

292

294

Fig. 9. Comparison of near edge X-ray absorption fine structure (NEXAFS) spectra for particulate matter from diesel exhaust (top panel) versus wood smoke (bottom panel) (data courtesy of A. Braun). Gray symbols in the top panel show the spectrum for diesel fuel.

a continuum assigned to recoiling free protons between graphene planes. The superimposed structure is dominated by C–H modes associated with polyaromatic structures in the case of anthracite, but lower rank coals also exhibit features associated with aliphatic C–H modes. Albers et al. (2000) applied INS to compare diesel soot collected with and without passing the vehicle exhaust through a diesel oxidation catalyst. The spectra were all broadly similar to those of carbon black. The main features are a broad band at ∼ 880 cm−1 corresponding to sp2 C–H out of plane bending and another at ∼ 1180 for sp2 in plane C–H bending. Any effect of the catalyst was primarily on the lower crystalline polyaromatic and aliphatic components of the soot, with little effect on the graphitic portion. Although the necessity of relatively large sample sizes and off-line analysis will hamper widespread use of the technique, it has the potential to address important questions regarding diesel PM. 4.8. Raman spectroscopy As with INS, Raman spectroscopy probes molecular vibrations, but in this case primarily of the carbon atoms in soot samples. Escribano, Sloan, Siddique, Sze, and Dudev (2001) and Sadezky, Muckenhuber, Grothe, Niessner, and Pöschl (2005) compare and model spectra of commercial graphites ranging from highly ordered pyrolytic graphite to printer ink, as well as soot from candles, tobacco smoke, and diesel exhaust. These spectra exhibit a characteristic set of peaks (Fig. 10), the principal ones labeled G (graphite) at 1575 cm−1 , D (defect) at 1350 cm−1 and 2D (D overtone) at 2720 cm−1 . The G peak corresponds to a C–C stretching motion along the longitudinal axis of the graphite plane. This mode has E2g symmetry and is the only allowed transition in crystalline graphite. The D peak originates from

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1099

Raman Spectra graphite powder

Intensity

G

2D

D

0

1000

2000

3000

4000

diesel PM Intensity

G D

2D

0

1000

2000 3000 wavenumber (cm-1)

4000

Fig. 10. Raman vibrational spectra of powdered graphite (top panel) and filter-collected diesel PM (bottom panel) (data courtesy of D. Uy).

a breakdown of the selection rules for graphene atoms near regions of disorder, with an intensity that depends on excitation laser wavelength. Rosen and Novakov (1977) recognized that the characteristic positions of these peaks make Raman spectroscopy a useful means to identify graphite-like carbon in ambient and source PM samples. Furthermore, their widths and relative intensities describe the local order in the sample. Owing to sizable microcrystalline domains, the peaks are sharp in graphite, with widths of ∼ 20 cm−1 , and the G peak intensity exceeds the D peak, as evident in Fig. 10. Increasing disorder, for example, in soot, causes the peaks to broaden and the D peak intensity to grow. The relative band areas can be used to ascertain the domain sizes, but with some care since organic molecules and fragments can also contribute to the breadth (Sadezky et al., 2005). Thus far, the application of Raman analysis to diesel PM has primarily addressed the carbon component (Lee, Zhu, Ciatti, Yozgatligil, & Choi, 2003). And the technique faces the drawback of requiring sample collection onto filters or other media. However, there are some attractive possibilities as well. Odziernkowski, Koziel, Irish, and Pawliszyn (2001) showed how sampling can be simplified by the use of solid phase microextraction fibers. That work and an initial study by Nelson et al. (2001) demonstrate the potential for Raman chemical imaging to help elucidate the molecular structure of individual ambient fine particles. They illustrated Raman spectroscopy’s sensitivity to other PM components than carbon, such as ammonium sulfate and nitrate; thus, future Raman studies of diesel PM may well extend beyond the graphitic component. 4.9. Particle volatility Although not strictly a chemical technique, measurement of particle volatility provides compositional information roughly comparable to EC/OC analysis, but with transient capability. In its basic format, particle number, mass, size distribution, or other property of interest is measured in two branches: “wet” to record total PM and “dry” to determine the “solid particle” or “EC” component. The “dry” branch incorporates a thermodenuder (Burtscher et al., 2001) to

1100

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

evaporate and remove semivolatile components. These include nucleation mode particles and adsorbed material, but as operationally defined by the thermodenuder temperature. The European Commission Particulates Project adopted this approach to address nucleation mode versus solid particle measurement for a variety of vehicle and fuel types (Ntziachristos et al., 2004). The basic method can be extended in various ways. One is the volatility tandem differential mobility analyzer (tDMA) (Kwon, Lee, Saito, Shinozaki, & Seto, 2003; Sakurai et al., 2003; Wehner, Philippin, Wiedensohler, Scheer, & Vogt, 2004). The first DMA selects a specific particle size, for example, nucleation or accumulation mode. These particles pass through a heat pipe, typically set to 100.300 ◦ C, after which shifts in size or number density are recorded with a second DMA. Besides the overall “EC/OC” ratio, this approach reveals the extent to which particles are internally mixed, and can be used to study how this depends on engine operating condition or traffic pattern, for example. A second variation, which provides a different kind of chemical information, is to examine how vapors adsorb onto particles. Weingartner, Gysel, and Baltensperger (2002) describe a hygroscopicity tandem DMA where size selected particles pass through a Gore-Tex tube held at specified temperatures and humidities, and their growth is monitored by the second DMA. A similar approach was used by Joutsensaari, Vaatovaara, Hämeri, and Laaksonen (2001) to categorize atmospheric particles by their relative adsorption of water vapor versus ethanol and by Grose et al. (2006) to examine the sulfate content in diesel nanoparticles downstream of a continuously regenerating trap. Finally, thermal desorption can be used with aerosol MS to help identify chemical constituents having the same m/z ratio, as done by Tobias et al. (2001) to distinguish the sulfate and hydrocarbon contributions to diesel nucleation mode particles.

5. Chemical insight into current questions on diesel PM It is evident that a wide array of analytical chemistry techniques have been brought to bear on the problem of diesel exhaust PM composition. What is relevant to this review is how these techniques help us understand diesel PM in the context of engine design, emission controls, and environmental concerns. The present section examines this through a series of topical questions. In the ensuing discussion, it is useful to recall that diesel exhaust PM is characteristically bimodal (Fig. 11), and that the differences in the particle size and volatility of these modes point to distinct origins. Some studies specifically address this distinction, but most use sampling methods that are operationally defined (e.g., regulatory filter-based methods).

nucleation mode particles

1e+9

dN/dlogdm (cm-3)

solid particles (sampled with thermodenuder)

1e+8

1e+7 thermodenuder temp. 20 C 293 C 1e+6

350 ppm sulfur / active catalyst 10

100 dm (nm)

Fig. 11. Characteristic biomodal size distribution (number concentration versus mobility diameter) of diesel exhaust particulate matter. The two modes exhibit different volatility, suggesting distinct origins—nucleation and soot.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1101

5.1. How does engine design and operation impact PM emissions? Diesel combustion depends on a number of parameters, the principal ones being engine speed, load, fuel injection timing, turbocharging, and EGR. Of these, speed and load receive the most attention. The CARB 8 mode and European ESC 13 mode test cycles, for example, to comprise a sequence of steady state operating points, including idle plus two or three engine speeds at loads from 25% to 100% of the rated maximum. The overall PM emissions tend to increase with increasing load, except at idle, which can have the highest emissions (Kweon, Foster, Schauer, & Okada, 2002). But this dependence is not even directionally the same for all PM components. Shi, Mark, and Harrison (2000) observe for a 1995 model year diesel engine that the EC fraction increases with both engine speed and load, from 25% at 1600 rpm and 25% load to 52% at 2600 rpm and 100% load. In contrast, OC shows the opposite trend, decreasing from 58% to 24%. Using a direct injection engine, Kweon et al. (2002) find the same behavior; EC increases with load from ∼ 10% to ∼ 90%, and OC decreases from ∼ 80% to ∼ 10%, in this case at 1200 and 1800 rpm. Zielinska, Sagebiel, Arnott et al., 2004 observe similar trends in a study of in-use diesel vehicles, but with maximum EC levels of about 30%. This behavior holds true for transient drive cycles as well. Shah, Cocker, Miller, and Norbeck (2004) examined the EC/OC ratio of PM emission from a fleet of 11 HHDDT operated on the CARB transient cycle consisting of cold start/idle, creep, transient, and cruise phases. The EC emission rates (per mile) increase by an average factor of 1.9 from cruise mode (high average load) to creep mode (low load), but the OC emissions increase by a much larger factor of 8.1. The latter is consistent with PAH and n-alkane emission rates recorded by Shah, Ogunyoku, Miller, and Cocker (2005), which were an order of magnitude higher during creep as compared to cruise phase. Interestingly, variations in the ratios of individual alkane and PAH species with operating mode remained modest. Diesel combustion includes both premixed and diffusion components. Higher speed and load allow less mixing; thus, these conditions emphasize diffusion burning and consequently higher EC levels. Supporting this view, Kweon et al. (2002) find that OC emissions correlate with the premixed portion’s contribution to the heat release. In a separate study on injection timing, Kweon et al. (2003a) observe a reduction in EC emissions as the injection timing is advanced, and note that earlier injection leads to a longer ignition delay and, hence, more time for air–fuel mixing. Unburned fuel and oil from the cooler and less efficient combustion at idle and low load can also contribute to OC particulate emissions. During combustion sulfur compounds are predominantly oxidized to SO2 , with a small fraction yielding SO3 , which reacts with water to form sulfuric acid (sulfate). At ∼ 400 ppm fuel sulfur, the sulfate contribution to diesel PM increases with load from a ∼ 1% to ∼ 11% (Kweon et al., 2002; Shi et al., 2000). But with ∼ 10 ppm sulfur fuel, much less sulfate is found in the engine-out emissions. Higher molecular weight PAHs are generally associated with PM emissions, whereas lower weight PAHs partition more to the gas phase (Shi et al., 2000). Zielinska, Sagebiel, Arnott et al. (2004) found that the partitioning of intermediate size PAHs depends on load. At low load 3- to 4-ring PAHs were predominantly associated with gas phase emissions, but at high load they shifted more to the particulate phase. Riddle et al. (2007) observed a similar load effect, where larger 5- and 6-ring PAHs contribute to PM in the creep and idle phases of the CARB HDD test, but very little is found under higher load. To achieve EPA 2010 and Euro 5 emission standards, engine operation must be optimized together with the aftertreatment system. Current technology makes use of multiple fuel injection pulses and high levels of EGR. Little is known about how these affect PM chemistry. High EGR levels increase soot concentrations that can foul EGR coolers and valves. Insight into the PM chemistry is needed to overcome these hurdles. It is also needed to optimize DPF regeneration, where soot oxidation rates depend on its composition. 5.2. What is the influence of ambient temperature on diesel PM? Although PM emissions are generally considered to increase at low ambient temperature, there are rather few studies on this subject. Aside from such potential difficulties as ice blockage in the sampling system, concerns about gas–particle partitioning during sample collection are exacerbated at low ambient temperatures. By independently examining the effects of test cell, dilution air, and instrument temperature on vehicle exhaust particle measurement, Ristimäki, Keskinen, Virtanen, Maricq, and Aakko (2005) concluded that the instruments could be operated at room temperature, but representative results required dilution air temperature consistent with that of the test cell. This implies similar considerations for the sampling trains used for traditional chemical analyses of PM composition, but there is the added difficulty of typically long times between sample collection and analysis.

1102

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Ristimäki et al. (2005) and Mathis, Mohr, and Forss (2005) examined semivolatile versus solid particle emissions at cold temperatures, but not their detailed chemical composition. The light duty diesel vehicles tested showed an influence of cold temperature during the vehicle cold-start, but relatively little effect on PM emissions after the engine warmed up. Even during the cold start, the primary effect of a cold environment is an increase in semivolatile nucleation mode, but not solid particle, emissions. The behavior is similar, but appears more pronounced, for DPF equipped diesel vehicles due to the absence of solid particles. Two-real world vehicle tests nearing completion, the Kansas City and the Gasoline/Diesel PM Split studies (NREL, 2007), reveal higher PM emissions during cold weather rather than warm weather that appears not to be limited to the cold start. But these studies predominantly consider (representative, in-use) gasoline vehicles. Chemical speciation, including PAHs, hopanes, and steranes, was carried out by Zielinska et al. (2004) on a small set of gasoline and diesel vehicles comparing emissions at −1 ◦ C versus 25 ◦ C. Sulfate emissions increased nearly 10-fold for the gasoline vehicles, but decreased slightly for the diesel vehicles at −1 ◦ C. The total PM emissions more than doubled at −1 ◦ C, due almost entirely to higher OC emissions. There was an accompanying increase of semivolatile PAH collected on filters, partially due to the higher partitioning of semivolatile compounds to the particle phase at lower temperature. The small change observed for hopane and sterane content at −1 ◦ C suggested a minor influence of temperature on diesel engine oil consumption. Real-time methods can help clarify the role of ambient temperature by providing speciation data during the important engine warm-up period. 5.3. How do fuel properties affect PM emissions? The question of fuel effects can be divided into three principal areas: fuel formulation, sulfur content, and biodiesel. Because of the strong relationship between sulfur and nucleation mode particles, this topic is discussed below with respect to nanoparticle formation. A good fuel must satisfy many requirements: favorable combustion properties (high cetane number for diesel fuel), sufficient lubricity, satisfactory vapor pressure and viscosity over a wide temperature range, safe handling characteristics, and so on. Within these constraints, additives or alternative fuels are being sought that improve the PM – NOx emissions tradeoff and reduce fossil fuel derived CO2 emissions. Chemical characterization of the emissions can help us understand how to improve combustion, and it can ensure that fuel changes do not introduce new potentially toxic compounds into the emissions. Conventional motor vehicle diesel fuel is a petroleum fractional distillate containing about 75% alkanes (paraffins) and 25% aromatics. The principal alternatives include low aromatic fuels, including Fischer–Tropsch, and oxygenates such as dimethyl ether and DMM. A number of traditional chemical characterization studies show a consistent relationship between the fuel aromatic content and PM emissions. Westerholm and Li (1994) and Westerholm et al. (2001) observed that reductions in PAH emissions correlate with the fuel content of these species. Consistent with this, Kweon et al. (2003b) reported significant reductions of PM2.5 and EC for Fischer–Tropsch versus No. 2 diesel fuel (352 ppm S). And Ball et al. (2001) found Fischer–Tropsch and DMM fuels to yield significantly lower PM mass as well as lower levels of 33 specific gas and particle phase compounds, primarily PAHs, as compared to EPA and CARB certification ˚ diesel fuels. EC/OC analysis by Alander, Leskinen, Raunemaa, and Rantanen (2004) revealed a 10–40% PM mass emissions reduction, almost entirely due to lower OC levels, for a reformulated diesel fuel having 45% lower aromatic and 94% lower sulfur content than conventional diesel fuel. With the introduction of DPFs, the emphasis shifts from a direct to an indirect effect of fuel on exhaust emissions. It is clear (e.g., Frank et al., 2004; Lev-On et al., 2002) that engine aftertreatment is considerably more effective in emission reductions than fuel reformulation. Thus, fuel compositions that benefit engine/aftertreatment system performance become more important than simply those reducing engine-out PM. Of interest for future studies are chemical characteristics of the exhaust gaseous and particulate species that reduce fouling (e.g., EGR system and sensors), promote DPF regeneration, reduce ash content (which limits DPF soot capacity), assist NOx reduction technology, and so forth. Desires for energy independence and reductions in greenhouse gas emissions are the primary drivers for biofuels. Their chemistry is distinct from petroleum fuels. Direct use of vegetable oils or animal fat in diesel engines can degrade performance because of their high viscosity and low volatility. Therefore, they are transesterified, usually with methanol, to produce fatty acid methyl esters. The result is an oxygenated fuel, with essentially zero aromatic and sulfur content. Used neat or in blends with petrodiesel fuel, biodiesel is generally found to lower PM, hydrocarbon and CO emissions, but NOx remains unchanged or slightly increases (Knothe et al., 2006; Wang, Lyons, Clark, Gautam, & Norton, 2000).

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1103

However, Durbin and Norbeck (2002) observed mixed results for biodiesel fuels, with some vehicles registering increases in PM emissions, and Mayer et al. (2005) reported little change in PM for rapeseed methyl ester. Even when overall PM emissions decrease, the SOF can increase. This fraction contains PAHs, albeit at lower levels than for conventional diesel fuel. As they are absent in biodiesel fuel, these PAHs must originate as byproducts of combustion (Bagley, Gratz, Johnson, & McDonald, 1998; Lin, Lee, Wu, & Wang, 2006). Why oxygenated fuels reduce PM emissions is not well understood. Tandem DMA measurements (Jung, Kittelson, & Zachariah, 2006) yielded sixfold higher oxidation rates for biodiesel as opposed to conventional diesel fuel derived soot, suggesting higher burnout rates. Buchholz et al. (2004) employed the 14 C naturally present in biofuels, but not petroleum fuels, to track carbon from the fuel to the emissions. By selectively 14 C labeling dibutyl maleate blended with diesel fuel, they observed that carbon atoms in the 2,3 maleate positions participated in PM formation, but those from the 1,4 positions did not. The latter are bonded to oxygen, suggesting that functional group structure is a factor in the mechanism for soot production. This notion is supported by the work of Song, Alam, Boehman, and Kim (2006) showing that diesel soot from biofuel (B100) is much more reactive towards oxidation than Fischer–Tropsch derived soot. Increased surface oxygen functionality of the B100 soot appears to be more important than initial structure or pore size distribution in enhancing the oxidation rate. Aside from insights into soot formation, these mechanistic studies are important for optimizing DPF performance, especially their regeneration. 5.4. How does lube oil affect PM emissions? Long chain alkanes in the base oil and sulfur supply condensable material for the semivolatile PM fraction and play important roles in the nucleation mode. Calcium, zinc, phosphorus, and other elements in the additive package contribute to the PM ash content. The metals are of interest because of their potential health effects, but there is increasing attention to their impact on exhaust aftertreatment systems. Since ash cannot be regenerated, it accumulates in the DPF over time, as detailed by Givens et al. (2003), increasing back pressure, reducing fuel economy, and adversely affecting durability. Only about 60–70% of the ash expected from fuel consumption collects in the DPF, implying that the remainder deposits in the oxidation catalyst and other parts of the exhaust system. The ash may have one positive side, namely that it lowers the activation energy for soot oxidation. This hypothesis is supported by the high temperature oxidation–tandem DMA study by Kim, Fletcher, and Zachariah (2005) that demonstrates how adding iron to a flame reduces the soot oxidation activation energy from ∼ 162 to ∼ 116 kJ/mol. And it is consistent with the use of fuel borne catalysts to promote DPF regeneration. Ash contributed by lube oil figures prominently in single particle mass spectra of diesel exhaust PM. The three main particle classes observed in the study by Toner, Sodeman, and Prather (2006) of six in-use HHDDTs are assigned to “EC, Ca, OC, and phosphate” (78% of particles), “OC, EC, phosphate, and sulfate” (8%), and “Ca, Na, EC, phosphate, and sulfate (7%), each of which exhibits a strong presence of lube oil components. Particles classes containing EC, Ca, and phosphate or sulfate are similarly prevalent in light duty gasoline vehicle PM (Sodeman, Toner, & Prather, 2005). An attractive application of single particle MS explored by Okada et al. (2003) is the real-time monitoring of oil consumption. They determine the overall consumption rate by chemical mass balance based on ICP-MS measurements of Ca in filter-collected PM samples, while recording time dependent variations with an ATOFMS. New EPA regulations for HDD engines require either closed crankcase systems or routing of the crankcase vapors into the dilution tunnel during certification testing (introduced into light duty gasoline vehicles in the 1960s). With DPFs drastically reducing soot emissions, condensed crankcase vapors may dominate the overall PM emissions. What this means in terms of PM composition remains largely unknown. 5.5. What are diesel nanoparticles? Elucidating the chemical nature of the diesel exhaust nucleation mode (Fig. 11), often referred to as diesel nanoparticles, is an area where aerosol MS has excelled. While typically dominant in number concentration, nucleation mode particles contribute little to the PM mass collected onto filter media, making it difficult for traditional off-line chemical analysis to distinguish them from the larger mass of accumulation mode particles. Tobias et al. (2001) overcame this by combining differential mobility analysis with thermal desorption particle beam mass spectrometry (TDPBMS). They applied this technique to the selective chemical analysis of nanoparticles from heavy-duty diesel engine exhaust.

1104

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

m/z = 69

m/z = 98 organic & sulfate

ion count (s-1)

organic 300

150

200

100

100

50 dmmd = 41 nm

0

dmmd = 41 nm 0

0

50

100 150 200 250 300

vaporizer temperature (C)

0

50

100 150 200 250 300

vaporizer temperature (C)

Fig. 12. Mass selected thermograms of size selected particulate matter from the exhaust of a heavy duty diesel vehicle at 40% load (data courtesy of P. Ziemann). Left panel: m/z = 69—one thermal desorption peak corresponding to organic material. Right panel: m/z = 98—peak at 40 ◦ C corresponds to sulfate (reference material desorbs at 30 ◦ C), and second peak at 150 ◦ C is from organic material with the same m/z ratio. dmmd , denotes mass median diameter.

Fig. 12 illustrates how mass-selected thermal desorption profiles can help elucidate chemical composition. The single peak at ∼ 150 ◦ C in the m/z = 69 temperature profile is assigned to organic material, whereas the ∼ 40 ◦ C peak in the bimodal profile at m/z = 98 matches well to thermal desorption of a sulfuric acid reference standard. Quantitative assessment reveals that the 41 nm particles in this example are composed of ∼ 95% organic material and < 5% sulfate. Comparison of the particle mass spectra to reference mass spectra for fuel–oil mixtures suggests that unburned lube oil is the dominant component, > 95%. This appears to be relatively independent of fuel composition, with comparable results obtained for EPA diesel fuel with 360 ppm S, California low sulfur fuel with 96 ppm S, and Fischer–Tropsch fuel with < 1 ppm S (Sakurai, Tobias et al., 2003). The semivolatile composition is similar for small and large particles, indicating that lube oil is the major component of the condensed material on soot particles, as well as to the semivolatile nucleation mode. Real-world chase measurements of New York City transit bus exhaust by Canagaratna et al. (2004) corroborate this picture. Aerosol mass spectra obtained from a fleet of 121 diesel buses, 20 CRT equipped diesel buses, and 31 CNG fueled buses, showed that the non-refractory PM was dominated by lube oil signatures. Newer engines exhibited somewhat lower emission indices, while those of CRT equipped diesels and CNG vehicles were reduced by ∼ 60%. Laser ablation/ionization single particle mass spectra (ATOFMS) recorded by Toner et al. (2006) give a complementary, but consistent view of HDD PM. While not quantitative on a particle by particle basis, ATOFMS is sensitive to the refractory particulate material. Besides EC, it reveals that 91% of the particles sampled exhibit calcium and phosphate peaks, indicative of lube oil additives. This picture changes for light duty diesel vehicles. Exhaust PM measurements in a wind tunnel showed that the combination of high sulfur fuel (∼ 350 ppm) and an oxidation catalyst produces a nucleation mode, but that neither alone does so (Maricq et al., 2002). The composition remained uncertain until aerosol MS by Schneider et al. (2005) and Scheer et al. (2005) directly demonstrated that, unlike for HDD engines, sulfate is a major semivolatile constituent in light duty diesel PM. This is illustrated in Fig. 13. Chemically specific particle size distributions for high sulfur fuel (360 ppm, left panel) show that the sulfur and organic PM components are both bimodally distributed and that sulfate constitutes about 85% of the semivolatile material. A thermodenuder largely eliminates the nucleation mode, here at ∼ 50 nm vacuum aerodynamic diameter. It also eliminates this mode in the overall particle size distribution measured by scanning mobility particle sizer (not illustrated) confirming the nucleation mode to be entirely semivolatile. With low sulfur fuel (2 ppm, right panel), the nucleation mode is absent and the sulfate contribution to the semivolatile fraction of the soot mode reduces to 50%. That sulfur remains at much higher fractions than for HDD PM is likely related to the relative oil consumption rates between light and HDD engines. The catalyst, which removes hydrocarbons, but stores and oxidizes SO2 , also has an obviously important role in determining the sulfate to organic ratio, as discussed next.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1105

Fig. 13. Sulfate and organic composition versus particle size for exhaust particulate matter from a light duty diesel vehicle (data courtesy of J. Schneider & R. Vogt). Left panel: high sulfur diesel fuel (360 ppm). Right panel: low sulfur fuel (2 ppm). Thermodenuder (TD) removes semivolatile material. dva denotes vacuum aerodynamic diameter.

5.6. How does exhaust aftertreatment affect PM? Diesel exhaust aftertreatment is evolving towards a complex system comprised oxidation catalyst, DPF, and NOx reduction strategy. Each affects PM composition. The catalyst does not directly reduce PM emissions, except perhaps for some diffusion losses, but acts indirectly by altering precursors of semivolatile PM. Oxidation of SO2 to SO3 leads to enhanced nucleation mode particles (Maricq et al., 2002; Vaaraslahti et al., 2004a, 2004b; Vogt et al., 2003), whereas removal of hydrocarbons reduces the amount of material available for condensation onto soot or for nucleation. At a more detailed level, the catalyst can alter soot structure. Albers et al. (2000) made INS and secondary ion MS measurements of soot that show reductions in hydrogen content and adhered SOF of PM after passage through an oxidation catalyst, while leaving the graphitic component unaltered. Collura et al. (2005) examined soot collected by SiC DPF either with or without an oxidation catalyst. Comparative surface characterization by X-ray photoelectron and diffuse reflection infrared spectroscopy showed a significant reduction of the SOF (from 20 to 6 wt%), as well as changes to the surface oxygen groups and porosity. The typically low exhaust temperature of diesel engines requires catalytic assistance for DPF regeneration. There are two basic options: incorporate the catalyst into the DPF substrate or into the fuel. While filtering of accumulation mode soot particles may promote a nucleation mode by reducing the available surface area for condensation, catalyzed DPFs are nonetheless observed to reduce semivolatile organic matter mass. Warner, Johnson, Bagley, and Huynh (2003) and Thalagavara, Johnson, Bagley, and Shende (2005) studied the effect of a catalyzed DPF (with 5 and 50 g Pt/f 3 ) on the emissions from a 10.8 L diesel engine (using 375 and 0.6 ppm S fuel) run on modes 8–11 of the EPA 13 mode cycle. Indeed, nucleation mode particle number emissions increased downstream of the DPF; however, Soxhlet extraction and EC/OC analysis of filter-collected PM consistently showed that both solid particle and semivolatile organic mass emissions decreased. This behavior was also observed by Lev-On et al. (2002) in a study of catalyzed DPF performance on six buses and grocery trucks, where gas and particulate phase PAH emissions were equally reduced. In contrast, sulfate in these studies increased post DPF, presumably from catalytic oxidation of SO2 to SO3 . The sulfate emission rate, however, is controlled by the fuel sulfur level, being more than an order of magnitude lower for the ultra-low sulfur fuel (Thalagavara et al., 2005) as compared to the 375 ppm S fuel (Warner et al., 2003). Looking specifically at the nucleation mode, the catalyzed DPF alters its composition. Based on tandem DMA measurements of particle volatility and hydroscopicity, Grose et al. (2006) concluded that nucleation mode particles downstream of the DPF are composed primarily of sulfate, perhaps sulfuric acid partially neutralized by ammonia. This is in contrast to a similar analysis of PM from a non-DPF equipped diesel engine where the nucleation mode volatility resembled that of C24 to C32 alkanes (Sakurai, Tobias et al. 2003). Use of fuel-borne additives involves a number of design considerations ranging from the DPF regeneration performance of the base metal to additive stability in the fuel (Caprotti, Field, Michelin, Schuerholz, & Terres, 2003). Additives currently in use include cerium and iron. These already influence soot formation during combustion. Tandem

1106

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

DMA investigation of diesel soot oxidation by Jung, Kittelson, and Zachariah (2005) revealed that fuel borne cerium does not alter the activation energy for oxidation, but rather increases the pre-exponential factor, possibly by increasing the number of metal sites already present from lube oil ash. Kasper, Sattler, Siegmann, Matter, and Siegmann (1999) showed how ferrocene produces iron oxide condensation nuclei prior to soot formation, onto which carbonaceous matter preferentially condenses and is oxidized. Single particle MS (Lee et al., 2006) and TEM plus energy dispersive spectroscopy (Miller, Ahlstrand, Kittelson, & Zachariah, 2007) were used to investigate the composition and morphology of soot particles from ferrocene doped diesel fuel. At low levels, iron uptake was primarily via iron condensation onto soot particles, but at higher doping levels they observed homogeneous iron nucleation, consistent with Kasper, et al. (2007). Braun et al. (2006) used NEXAFS and WAXS to characterize diesel soot generated with 1000 ppm ferrocene doped fuel and collected onto quartz filters. Whereas soot from the base fuel was dominated by a graphite-like microstructure, the soot from the doped fuel showed a pronounced aliphatic character, graphene sheets with low stack height, and iron in the form of Fe2 O3 maghemite. While these studies examine the influence of additives on soot composition, other work is aimed at understanding how these changes impact aftertreatment performance, for example the effect of adsorbed hydrocarbons on DPF regeneration (Stratakis, Konstantas, & Stamatelos, 2003; Stratakis & Stamatelos, 2003). Exhaust aftertreatment raises the possibility that these devices themselves become sources of new emissions. The first such concerns arose in connection with precious metals when widespread use of catalytic converters began in gasoline vehicles. Roadside dust and soil samples revealed enhanced Pt, Pd, and Rh contamination, in relative concentrations roughly correlating with those of automotive catalysts (e.g., Ely et al., 2001; Hodge & Stallard, 1986; Zereini, Wiseman, & Püttmann, 2007). Typical soil abundances are on the order of 70, 25, and 5 ng/g, respectively (Ely et al., 2001), decreasing rapidly with distance from the roadside, and increasing over time. Trace measurements in urban air have also revealed elevated concentrations of precious and heavy metals relative to background (Lough et al., 2005; Zereini et al., 2005). Many of these are attributable to brake wear, tire wear, and resuspended road dust, but here too the precious metals are in proportions similar to automotive catalysts, suggesting this as a principal source (Rauch, Hemond, Peucker-Ehrenbrink, Ek, & Morrison, 2005). Specific levels vary widely with location, ∼ 7 pg/m3 Pt in Boston (Rauch et al., 2005) to ∼ 150 pg/m3 in Frankfurt (Zereini et al., 2001). Cascade impactor data indicate that 75% of the Pt and 95% of the Rh is associated with particles > 2 m diameter, with the peak of the distribution at ∼ 5 m (Zereini et al., 2001). This size range is consistent with wear process, a combination of vibration and temperature cycling, as the release mechanism. Direct tailpipe measurements of Pt, Pd, and Rh corroborate roadside and airborne observations. Their emissions decreased dramatically with the introduction of catalyst monoliths to replace the earlier used pellets (König, Hertel, Koch, & Rosner, 1992). Artelt, Kock, König, Levsen, and Rosner (1999) found that Pt emissions decrease with catalyst age, with typical emission rates dropping from 12–90 ng/km for new, to 9–26 ng/km for aged catalytic converters. Consistent with this, Palacios et al. (2000) observed Pt, Pd, and Rh emissions from new catalysts of 100, 250, and 50 ng/km fall to ∼ 7, 14, and 3–12 ng/km, respectively, after aging for 30,000 km. But Pt emissions from diesel oxidation catalysts were higher, 400–800 ng/km for new versus 108–150 ng/km for aged catalysts. This study also reports energy dispersive X-ray spectroscopic and LIBS measurements that characterize the spatial distributions of the precious metals in the monolith and how these change with age. Pt emissions were found almost exclusively bound to aluminum oxide particles, with only minute quantities, less than ∼ 1%, water soluble (Artelt et al., 1999). Approximately 43–74% had aerodynamic diameters larger than 10 m and 10–36% were < 3 m. König et al. (1992) also observed mass median diameters to be 5– > 10 m. The vast majority of these metals analyses are performed off-line using XRF, ICP-MS and other methods. But, Pt and Ce emissions in automobile exhaust have also been observed by single particle MS (Silva & Prather, 1997). The precious metal was again associated with micron sized particles. Pt was found in only a few percent of the particles, in contrast to 12–87% for Ce. By utilizing its high time resolution, future ATOFMS studies may provide a better understanding of just how precious metals are released from the catalyst or DPF. Catalytic converters, DPFs, NOx traps, and urea SCRs are in essence chemical reactors that can potentially produce secondary emissions chemically distinct from engine out species. Polychlorinated dibenzodioxins and dibenzofurans are two classes of compounds receiving attention in this regard. They have been observed in a 1989 roadway tunnel study (VAAlerenga tunnel in Oslo, Norway) with emission factors of 0.04–0.5 ng/km for light duty and 0.8–9.5 ng/km for heavy-duty vehicles, depending on driving direction through the tunnel (Oehme, Larssen, & Brevik, 1991). Dioxin

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1107

formation is a three step process requiring chlorine and aromatics, but can be promoted by the presence of copper (van Setten et al., 2001). Clunies-Ross, Stanmore, and Millar (1996) demonstrated this effect by comparing dioxin levels in diesel emissions produced running a single cylinder, indirect injection, engine on a fuel doped with 34.4 mg/L of copper and with a soot trap versus stock fuel and no trap. This led Heeb et al. (2005, 2007) to develop a secondary emissions risk assessment for DPFs. The tests are conducted using engine dynamometers, collecting gaseous and PM samples from the exhaust, and analyzing them for a variety of toxic compounds, including PAHs, nitro-PAHs, polychlorinated dibenzodioxins, and dibenzofurans. Amongst 20 HDD DPFs tested, two using copper additives to aid regeneration were rejected.

5.7. Can PM composition help source apportionment? Carbonaceous fine particles are generally associated with combustion sources. Yet tracers permitting more detailed source apportionment have proven elusive. When the use of leaded gasoline ended, attention turned to PAHs as potential motor vehicle emission markers (see also Section 2). Subsequently more detailed examinations of the organic compounds in motor vehicle PM were undertaken, including n-alkanes, n-alkonoic acids, benzoic acids, benzaldehydes, PAH, oxy-PAH, steranes, pentacyclic triterpanes, and azanaphthalenes. A comparison of source measurements, including catalyst and non-catalyst equipped gasoline vehicles, diesel trucks, wood smoke, meat broiling, and others suggested that hopanes and steranes (which originate from non-synthetic lube oils) could serve as markers of vehicular emissions (Fraser et al., 1999; Rogge et al., 1993; Schauer et al., 1999a, 1999b; Schauer et al., 1996). These, however, vary with engine operating condition and particle size; for example Riddle et al. (2007) found the relative abundances of hopanes and steranes to total HC emissions to shift to smaller particles at higher engine load. It is generally thought that the EC/OC ratio is higher in diesel as compared to gasoline vehicle exhaust (Geller et al., 2005; Kleeman et al., 2000). In an examination of gasoline and diesel vehicle emissions acquired in connection with the Arizona State vehicle inspection program Watson et al. (1994) found carbon composition differences they suggested could be sufficient to aid apportionment through source-receptor models. However, Schauer (2003) cautions that EC originates from many sources, and that this combined with differences between measurement techniques can lead to inaccuracies. Miguel et al. (1998) found from Caldecott tunnel measurements that diesel trucks emitted relatively more lighter, and gasoline vehicles heavier, molecular weight PAHs. Thus, they used the ratios of PAHs to black carbon as a source profile to help distinguish gasoline from diesel vehicle PM in their tunnel study. Other compounds that could help distinguish gasoline vehicle PM from diesel exhaust and wood smoke include isoprenoids and tricyclic terpanes (Schauer et al., 2002). The advent of single particle MS has brought a new tool for source apportionment capable of much higher time resolution than conventional off-line chemical analysis. While not quantitative in the traditional sense of mass emissions, mass spectral patterns can be used to classify particles on an individual basis as belonging to combustion, mineral, sea salt, etc. Vogt, Kirchner et al. (2003) found eight such classes at rural, urban, and highway sampling sites, including those characteristic of diesel exhaust particles. In the Caldecott tunnel, Gross et al. (2000) identified three main classes of particles associated with motor vehicle emissions: (1) those with significant PAH features, (2) EC, and (3) inorganic ions such as Ca and PO− 4 related to lube oil additives. These represented 61.4%, 10.3% and 11.0%, respectively, of the total particles sampled during 3 h of heavy traffic in the light-duty vehicle bore, and 57.4%, 11.8%, and 18% of the particles in the mixed heavy duty–light duty tunnel bore. Comparison of vehicle exhaust PM data to single particle mass spectra of aerosolized fuel and lube oil samples provide another avenue to help identify ambient sources (Spencer, Schields, Sodeman, Toner, & Prather, 2006). Single particle MS also points out underlying difficulties in source apportionment. EC, which in filter-based measurements is associated with diesel PM, can dominate non-smoking gasoline vehicle particles when looked at in terms of ultrafine particle number. This suggests that an EC signature alone is not sufficient as a tracer for diesel PM (Sodeman et al., 2005). Ambient temperature and sampling conditions can further affect the suitability of marker compounds. Gross et al. (2005) identified alkylated naphthalenes as potential heavy-duty diesel tracers, but observed them in significantly higher concentrations during November, with temperatures of 10–15 ◦ C, as compared to July, with temperatures of 26–32 ◦ C. Such factors may help account for the conflicting estimates of diesel versus gasoline vehicle contributions to PM inventories discussed by Gertler (2005), where methods, such as receptor modeling, show diesel PM to dominate, while others, for example, roadway emission factors, suggest the opposite.

1108

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Other new techniques for PM composition can contribute as well. For example, the ability of NEXAFS to probe functional group signatures (Fig. 9) may prove useful for source apportionment of diesel PM relative to wood smoke (Braun et al., 2007). Whether by MS or by spectroscopy, chemical structure is a key factor for source apportionment and emission inventories. An important current activity that these techniques can support includes development of EPA’s motor vehicle emission simulator (MOVES) (US EPA, 2007a). 5.8. How do atmospheric process alter diesel exhaust? Atmospheric chemistry of diesel exhaust impacts PM in two basic ways: secondary aerosol formation and soot aging. In the former case, chemical compounds emitted into the atmosphere are oxidized by reactions with OH, O3 , and other reactive trace species. This process introduces carbonyl and carboxylic acid functionality into the originally hydrophobic hydrocarbon emissions, decreases their saturation vapor pressures, and makes them more prone to partition into the particle phase whether by condensation or nucleation. Environmental simulation chambers provide a powerful method to examine these atmospheric transformations (e.g., Zielinska et al., 2006). Secondary aerosol formation can proceed directly from gas phase chemistry. In the environmentally relevant case of gasoline vapors, Odum, Jungkamp, Griffin, Flagan, and Seinfeld (1997) showed the secondary aerosol to be almost entirely attributable to the aromatic content of the fuel. But soot particles can also participate in this secondary chemistry. Lee, Jang, and Kamens (2004) examined the photooxidation of -pinene in the presence of fresh diesel soot and observed both changes to the soot composition and increases in the secondary aerosol production. The relationship between what is emitted at the tailpipe and the eventual ambient PM burden is likely more complex. Recent work by Robinson et al. (2007) suggests that semivolatile material initially emitted in the particulate phase can partially evaporate, undergo atmospheric photooxidation, and subsequently nucleate or condense into secondary organic aerosol, something not taken into account in current atmospheric models. The analytical techniques to substantiate such mechanisms have relied primarily on gas chromatographic and mass spectrometric methods similar to those described in Section 2 for diesel PM analysis. But, additional procedures, such as conversion to silyl derivatives (Nolte et al., 2002), are needed to interrogate the polar species produced by photooxidation. Increasingly, though, aerosol MS is employed, as in the study by Robinson et al. (2007). Diesel soot itself undergoes atmospheric transformations subsequent to its emission. These largely ultrafine particles rapidly disperse and soon coagulate with other atmospheric particles, processes that contribute to their rapid decrease in number concentration away from freeways (Zhu, Hinds, Kim, Shen, & Sioutas, 2002). Exposure to ozone and UV light increases soot particle hygroscopicity (Weingartner, Burtscher, & Baltensperger, 1997). The impact of these oxidative changes to the soot surface has been examined by Decesari et al. (2002) using a combination of ion exchange chromatography, total carbon analysis and proton nuclear magnetic resonance. They found ozone exposure to increase the water soluble organic content of soot, consisting mainly of aromatic polyacids. Laboratory studies by Johnson et al. (2005) and Zuberi et al. (2005) using electron microscopy, energy dispersive X-ray spectroscopy, and secondary ion MS demonstrate that this aging causes the soot to become more hydrophilic, giving it the ability to retain water and be removed from the atmosphere either by entrapment into existing water droplets or by acting as cloud condensation nuclei. Diesel soot also reacts with NO2 . In the exhaust system this is used to help regenerate DPFs, but in the atmosphere it leads to HONO production. In principle this could affect the oxidative capacity of the atmosphere, but atmospheric processing reduces the soot reactivity towards NO2 , leaving this a minor pathway (Arens, Gutzwiller, Baltensperger, Gäggeler, & Ammann, 2001; Lelièvre, Bedjanian, Laverdet, & Le Bras, 2004). The preceding discussion addresses merely a fraction of the atmospheric chemical transformations of diesel exhaust, the details of which are mostly unknown. A better understanding of the compositional changes is needed to help clarify questions as to human exposure, potential health effects, and climate effects of these emissions. It could also help point to improved emissions measurement methods. 5.9. What can PM composition contribute to health effects studies? Much of the early interest in diesel PM composition was motivated by health effects concerns and aimed at the identification of PAH and nitro-PAH exhaust components (Bechtold et al., 1984; Henderson et al., 1983, 1984). Besides examining the potential for mutagenic effects from diesel PM extracts, the possibility was noted of nitro- and oxy-PAH formation by oxidation of material collected by filters (Lies, Hartung, Postulka, Gring, & Schulze, 1986; Risby & Lestz,

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1109

1983). This artifact was deemed minor for FTP testing, but could become more significant at high NO2 concentrations or long sampling times. As more complex aftertreatment systems are introduced and overall emission rates drastically decline, this issue needs to be kept in mind. Present attention has shifted to the potential toxicological and cardiovascular effects of motor vehicle exhaust. As these studies examine the comparative effects of emissions from various engine technologies and operating conditions, chemical analysis can play an important supporting role. The changing engine, aftertreatment, and fuel landscape has led to the desire for rapid screening methodologies to ascertain the potential biological effects (McDonald, Harrod, Seagrave, Seilkop, & Mauderly, 2004; Seagrave et al., 2002; Seagrave, Seilkop, & Mauderly, 2003). As our ability to examine exhaust chemical composition improves it will provide a better context within which to view any changes to the relative potencies of exhaust emissions (Geller et al., 2006). Just as PM emissions undergo chemical and physical changes in the atmosphere, they can do so as well in animal exposure studies. The importance of this is evident from the chemical characterization of inhalation exposure atmospheres by McDonald, Barr et al. (2004). These were found to be generally consistent with on-road and laboratory emission measurements, but secondary PM was also observed from the interaction of the diluted vehicle exhaust with ammonia from the animals. Additionally, some exhaust components did not scale as expected with dilution ratio owing to background material from the rodents (e.g., their respiration) and dilution air. These are important considerations for the upcoming ACES program on the health effects of “clean diesels”. 5.10. What about PM from non-road diesel engines? The vast majority of diesel PM characterization work to date has been on motor vehicle exhaust emissions. Yet diesel engines are employed in numerous other applications, including ships, locomotives, construction equipment, agricultural machinery, and back-up generators (BUG). With the recent regulatory changes dramatically reducing motor vehicle tailpipe emission rates, increased attention has turned to the environmental impact of PM from non-road sources. These emissions are additionally subject to occupational exposure regulations. A review by Cantrell and Watts (1997) found exposures to be highest in locations such as mines that use heavy diesel equipment in confined spaces. The airborne PM standards established for mines by the U.S. Department of Labor’s Mine Safety and Health Administration are based on total carbon concentrations. The use of this metric as a tracer for diesel exhaust risks interference from other carbons sources such as oil mists and cigarette smoke. To examine this issue, McDonald, Zielinska, Sagebiel, and McDaniel (2002) and McDonald, Zielinska, McDaniel, and Mousset-Jones (2003) combined source sampling and chemical mass balance modeling to apportion the PM at various locations within a gold mine. Their analysis included size segregated measurements of PAH, oxy-PAH, hopanes and steranes, which although < 1% of the total OC, were found consistent with diesel emissions. While distinctions between mines do not permit generalizations, their analysis demonstrated diesel exhaust to contribute 78–98% of the fine particulate mass and > 90% of the fine particle carbon in the gold mine studied. Locomotive and marine diesel engines are in principle amenable to the same exhaust aftertreatment strategies, oxidation catalyst, DPF, SCR, as those of on-road vehicles. Indeed, the EPA has recently introduced regulations for locomotive and marine emissions (US EPA, 2007b), but efforts to evaluate various emissions control measures began already before this (e.g., the case studies described in MECA, 2006). There are, however, few scientific investigations into how diesel emissions in these applications differ from those of motor vehicles. Kasper, Aufdenblatten, Forss, Mohr and Burtscher (2007) investigated PM emissions from a 2-stroke marine diesel engine and found them considerably different from their automotive counterparts. The particles, predominantly 20–40 nm in diameter, were substantially smaller than typical of 4-stroke diesel engines. Thermodenuder analysis of particle volatility showed about 80% of the mass to be comprised semivolatile organic matter, higher than for heavy duty, and much higher than for automobile, diesel engines. Only about 1.4% of the fuel sulfur was converted to sulfate; however, the introduction of an oxidation catalyst could well increase this fraction. The operating modes of locomotives and ships are distinct from motor vehicles, relying much more on low speed, high power, operation. In this way, they are similar to BUG, which typically operate near steady state at mid-load. Shah et al. (2006) recorded regulated emissions from 18 diesel BUGs ranging from 60 to 2000 kW in size. The PM emission factors were 83% lower for small and 50% lower for large BUGs than in EPA’s AP-42 inventory. Durbin et al. (2007) investigated the influence of biofuels on these emissions, but for 20% blends did not find consistent trends. The influence of aftertreatment on non-road diesel applications still needs study. Strategies such as urea SCR

1110

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

were originally designed for stationary sources, so it might be anticipated that NOx as well as PM reductions can be successfully achieved with stationary diesel engines. 6. Conclusion The field of diesel engine emissions characterization is presently in a time of considerable change. Due to tightening regulations and new engine and aftertreatment technologies, PM and other exhaust emissions are rapidly declining with new model vehicles. This brings with it significant implications for emission inventories and modeling. Sources once minor in comparison to on-road emissions are becoming comparatively more important and receiving increased regulatory attention. In turn, this places increased demands on measurement methods to sample and analyze PM emissions.At conventional diesel engine PM emission levels, the regulatory procedure of dilution tunnel sampling and filter collection proved satisfactory. But at the ∼ 1 mg/km PM emission rate of DPF equipped vehicles, gaseous sampling artifacts begin to compete with the collected PM mass. Nor is this traditional method amenable to recent regulations aimed at in-use emissions monitoring or to the vast variety of off-road applications. Thus, new PM measurement techniques are needed. And as they are developed and evaluated, it no longer makes sense to rely solely on an operational definition of PM, but rather one should take advantage of advances in the chemical analysis of PM to help develop more reliable and robust methods. Chemical composition can provide insight into a variety of problems related to PM emissions. Whether particles are solid or semivolatile has implications for sampling, atmospheric fate, and health effects. Whether particles contain metal, inorganic ion, PAH, or other molecular species presumably at some level affects their health implications. And the pattern of composition can provide a means for source apportionment, which in turn supports emissions inventory development. These issues are becoming increasingly complex as new aftertreatment technologies and fuel structures enter the marketplace and alter previously used signatures. Particle chemical composition is needed for new technology to ascertain future emission benefits and to avoid unintended consequences. But the overall lower emissions from new vehicles means that more data are required from non-road sources as their relative burden increases. Successfully dealing with these demands will rely on continued development of new methods for investigating PM chemical composition. Acknowledgments The author sincerely thanks D. Gross (Carleton), G. Huffman (U. Kentucky), and D. Lawson (NREL) for their help supplying reprints and references. T. Perez (Ford), M. Galei (TSI Inc.), T. Johnson (Corning, Inc.), and P. Ziemann (U.C. Riverside) kindly helped with some of the illustrations. A. Braun (EMPA), D. Uy (Ford), J. Schneider (Max Planck Institute, Mainz), R. Vogt (Ford Forschungszentrum Aachen), and P. Ziemann (U.C. Riverside) generously provided data for graphs displayed in this review. Finally, the author is grateful to R. Chase (Ford) and J. Szente (Ford) for their critical reading of the manuscript and their helpful suggestions. References Abdul-Khalek, I. S., Kittelson, D. B., & Brear, F. (1998). Diesel trap performance: Particle size measurements and trends. SAE Technical Paper 982599. ACEA (2002). Report on small particle emissions from passenger cars [2]. European Automobile Manufacturers Association http://www.acea.be. ˚ Alander, T. J. A., Leskinen, A. P., Raunemaa, T. M., & Rantanen, L. (2004). Characterization of diesel particles: Effects of fuel reformulation, exhaust aftertreatment, and engine operation on particle carbon composition and volatility. Environmental Science & Technology, 38, 2707–2714. Albers, P. W., Klein, H., Lox, E. S., Seibold, K., Prescher, G., & Parker, S. F. (2000). INS-, SIMS- and XPS-investigations of diesel engine exhaust particles. Physical Chemistry Chemical Physics, 2, 1051–1058. Allen, J. O., Mayo, P. R., Hughes, L. S., Salmon, L. G., & Cass, G. R. (2001). Emissions of size-segregated aerosols from on-road vehicles in the Caldecott tunnel. Environmental Science & Technology, 35, 4189–4197. Arens, F., Gutzwiller, L., Baltensperger, U., Gäggeler, H. W., & Ammann, M. (2001). Heterogeneous reaction of NOx on diesel soot particles. Environmental Science & Technology, 35, 2191–2199. Arnott, W. P., Zielinska, B., Rogers, C. F., Sagebiel, J., Park, K., Chow, J. et al. (2005). Evaluation of 1047-nm photoacoustic instruments and photoelectric aerosol sensors in source sampling of black carbon aerosol and particle-bound PAHs from gasoline and diesel powered vehicles. Environmental Science & Technology, 39, 5398–5406. Artelt, S., Kock, H., König, H. P., Levsen, K., & Rosner, G. (1999). Engine dynamometer experiments: Platinum emissions from differently aged three-way catalytic converters. Atmospheric Environment, 33, 3559–3567.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1111

Bagley, S. T., Gratz, L. D., Johnson, J. H., & McDonald, J. F. (1998). Effects of an oxidation catalytic converter and a biodiesel fuel on the chemical, mutagenic, and particle size characteristics of emissions from a diesel engine. Environmental Science & Technology, 32, 1183–1191. Ball, J. (2001). A toxicological evaluation of potential thermal degradation products of urea. SAE Technical Paper 2001-01-3621. Ball, J. C., Hilden, D. L., Wolf, L. R., Natarajan, M., Yost, D. M., Schulman, M. E., et al. (2005). Emissions of toxicologically relevant compounds using dibutyl maleate and tripropylene glycol monomethyl ether diesel fuel additives to lower NOx emissions. SAE Technical Paper 2005-01-0475. Ball, J. C., Lapin, C., Buckingham, J., Frame, E., Yost, D., Garbak, J., et al. (2001). Dimethoxy methane in diesel fuel: Part 1. The effect of fuels and engine operating modes on emissions of toxic air pollutants and gas/solid phase PAH. SAE Technical Paper 2001-01-3627. Bardasz, E. A., Antoon, F. A., Schiferl, E. A., Wang, J. C., & Totten, W. (2004). The impact of lubricant and fuel derived sulfur species on efficiency and durability of diesel NOx adsorbers. SAE Technical Paper 2004-01-3011. Bechtold, W. E., Dutcher, J. S., Mokler, B. V., Lopez, J. A., Wolf, I., Li, A. P. et al. (1984). Chemical and biological properties of diesel exhaust particles collected during selected segments of a simulated driving cycle. Toxicological Sciences, 4, 370–377. Benner, B. A., Jr., Gordon, G. E., & Wise, S. A. (1989). Mobile sources of atmospheric polycyclic aromatic hydrocarbons: A roadway tunnel study. Environmental Science & Technology, 23, 1269–1278. Bezabeh, D. Z. (1999). Screening of aerosol filter samples for PAHs and Nitro-PAHs by laser desorption ionization TOF mass spectrometry. Aerosol Science and Technology, 30, 288–299. Bezabeh, D. Z., Bamford, H. A., Schantz, M. M., & Wise, S. A. (2003). Determination of nitrated polycyclic aromatic hydrocarbons in diesel particulate-related standard reference materials by using gas chromatography/mass spectrometry with negative ion chemical ionization. Analytical & Bioanalytical Chemistry, 375, 381–388. Braun, A., Huggins, F. E., Kelly, K. E., Mun, B. S., Ehrlich, S. N., & Huffman, G. P. (2006). Impact of ferrocene on the structure of diesel exhaust soot as probed with wide-angle X-ray scattering and C(1s) NEXAFS spectroscopy. Carbon, 44, 2904–2911. Braun, A., Huggins, F. E., Kubátová, A., Wirick, S., Maricq, M. M., Mun, B. S., et al. (2007). Towards distinguishing wood-smoke and diesel exhaust particulates in ambient particulate matter. Environmental Science & Technology, submitted for publication. Braun, A., Huggins, F. E., Shah, N., Chen, Y., Wirick, S., Mun, S. B. et al. (2005). Advantages of soft X-ray absorption over TEM-EELS for solid carbon studies—a comparative study on diesel soot with EELS and NEXAFS. Carbon, 43, 117–124. Braun, A., Shah, N., Huggins, F. E., Huffman, G. P., Wirick, S., Jacobsen, C. et al. (2004). A study of diesel PM with X-ray microspectroscopy. Fuel, 83, 997–1000. Braun, A., Shah, N., Huggins, F. E., Kelly, K. E., Sarofim, A., Jacobsen, C. et al. (2005). X-ray scattering and spectroscopy studies on diesel soot from oxygenated fuel under various engine load conditions. Carbon, 43, 2588–2599. Buchholz, B. A., Mueller, C. J., Martin, G. C., Cheng, A. S., Dibble, R. W., & Frantz, B. R. (2004). Tracing fuel component carbon in the emissions from diesel engines. Nuclear Instruments & Methods in Physics Research B, 223–224, 837–841. Burtscher, H. (2005). Physical characterization of particulate emissions from diesel engines: A review. Journal of Aerosol Science, 36, 896–932. Burtscher, H., Baltensperger, U., Bukowiecki, N., Cohn, P., Hüglin, C., Mohr, M. et al. (2001). Separation of volatile and non-volatile aerosol fractions by thermodesorption: Instrumental development and applications. Journal of Aerosol Science, 32, 427–442. Cabalo, J., Zelenyuk, A., Baer, T., & Miller, R. E. (2000). Two-color laser induced evaporation dynamics of liquid aerosols probed by time-of-flight mass spectrometry. Aerosol Science & Technology, 33, 3–19. Cadle, S. H., Groblicki, P. J., & Stroup, D. P. (1980). An automated carbon analyzer for particulate samples. Analytical Chemistry, 52, 2201–2206. Cadle, S. H., Mulawa, P., Groblicki, P., Laroo, C., Ragazzi, R. A., Nelson, K. et al. (2001). In-use light-duty gasoline vehicle particulate matter emissions on three driving cycles. Environmental Science & Technology, 35, 26–32. Cadle, S. H., Mulawa, P. A., Hunsanger, E. C., Nelson, K., Ragazzi, R. A., Barrett, R. et al. (1999). Composition of light-duty motor vehicle exhaust particulate matter in the Denver, Colorado area. Environmental Science & Technology, 33, 2328–2339. Canagaratna, M. R., Jayne, J. T., Ghertner, D. A., Herndon, S., Shi, Q., Jimenez, J. L. et al. (2004). Chase studies of particulate emissions from in-use New York City vehicles. Aerosol Science and Technology, 38, 555–573. Cantrell, B. K., & Watts, W. F., Jr. (1997). Diesel exhaust aerosol: Review of occupational exposure. Applied Occupational & Environmental Hygiene, 12, 1019–1027. Caprotti, R., Field, I., Michelin, J., Schuerholz, S., & Terres, F. (2003). Development of a novel DPF additive. SAE Technical Paper 2003-01-3165. CFR (2007). Code of Federal Regulations Title 40, Protection of Environment, Part 1065—Engine Testing http://ecfr.gpoaccess.gov/cgi/t/text/ text-idx?c = ecfr&sid = 94cf84cc9b9f6927a1ff4daf8d7f8642&tpl = /ecfrbrowse/Title40/40cfr1065_main_02.tpl. Chase, R. E., Duszkiewicz, G. J., Lewis, D., & Podsiadlik, D. H. (2005). Reducing PM measurement variability by controlling static charge. SAE Technical Paper 2005-01-0193. Chase, R. E., Duszkiewicz, G. J., Richert, J. F. O., Lewis, D., Maricq, M. M., & Xu, N. (2004). PM measurement artifact: Organic vapor deposition on different filter media. SAE Technical Paper 2004-01-0967. Cheng, Y., Cavataio, J. V., Belanger, W. D., Hoard, J. W., & Hammerle, R. H. (2004). Factors affecting diesel LNT durability in lab reactor studies. SAE Technical Paper 2004-01-0156. Chow, J. C., Watson, J. G., Crow, D., Lowenthal, D. H., & Merrifield, T. (2001). Comparison of IMPROVE and NIOSH carbon measurements. Aerosol Science and Technology, 34, 23–34. Chow, J. C., Watson, J. G., Pritchett, L. C., Pierson, W. R., Frazier, C. A., & Purcell, R. G. (1993). The DRI thermal/optical reflectance carbon analysis system: Description, evaluation and applications in U.S. air quality studies. Atmospheric Environment, 27A, 1185–1201. Clunies-Ross, C., Stanmore, B. R., & Millar, G. J. (1996). Dioxins in diesel exhaust. Nature, 381, 379. Cody, G. D., Ade, H., Wirick, S., Mitchell, G. D., & Davis, A. (1998). Determination of chemical-structural changes in vitrinite accompanying luminescence alteration using C-NEXAFS analysis. Organic Geochemistry, 28, 441–455. Cody, G. D., Botto, R. E., Ade, H., Behal, S., Disko, M., & Wirick, S. (1995). Inner-shell spectroscopy and imaging of a subbituminous coal: In-situ analysis of organic and inorganic microstructures using C(1s)-, Ca(2p)-, and Cl(2s)-NEXAFS. Energy & Fuels, 9, 525–533.

1112

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Collura, S., Chaoui, N., Azambre, B., Finqueneisel, G., Heintz, O., Krzton, A. et al. (2005). Influence of the soluble organic fraction on the thermal behaviour, texture and surface chemistry of diesel exhaust soot. Carbon, 43, 605–613. Cutler, W. A. (2004). Overview of ceramic materials for diesel particulate filter applications. Ceramic Engineering and Science Proceedings, 25, 421–430. Daisey, J. M., Cheney, J. L., & Lioy, P. J. (1986). Profiles of organic particulate matter emissions from air pollution sources: Status and needs for receptor source apportionment modeling. Journal of the Air Pollution Control Association, 36, 17–33. DeCarlo, P. F., Kimmel, J. R., Trimborn, A., Northway, M. J., Jayne, J. T., Aiken, A. C. et al. (2006). Field-deployable, high-resolution, time-of-flight aerosol mass spectrometer. Analytical Chemistry, 78, 8281–8289. Decesari, S., Facchini, M. C., Matta, E., Mircea, M., Fuzzi, S., Chughtai, A. R. et al. (2002). Water soluble organic compounds formed by oxidation of soot. Atmospheric Environment, 36, 1827–1832. Demokritou, P., Kavouras, I. G., Ferguson, S. T., & Koutrakis, P. (2002). Development of a high volume cascade impactor for toxicological and chemical characterization studies. Aerosol Science and Technology, 36, 925–933. Dieselnet (2007). Emission test cycles http://www.dieselnet.com/standards/cycles/. Durbin, T. D., Cocker III, D. R., Sawant, A. A., Johnson, K., Miller, J. W., Holden, B. B., et al. (2007). Regulated emissions from biodiesel fuels from on/off-road applications. Atmospheric Environment, 41, 5647–5658. Durbin, T. D., & Norbeck, J. M. (2002). Effects of biodiesel blends and Arco EC-diesel on emissions from light heavy-duty diesel vehicles. Environmental Science & Technology, 36, 1686–1691. Ely, J. C., Neal, C. R., Kulpa, C. F., Schneegurt, M. A., Seidler, J. A., & Jain, J. C. (2001). Implications of platinum-group element accumulation along U.S. roads from catalytic-converter attrition. Environmental Science & Technology, 35, 3816–3822. Escribano, R., Sloan, J. J., Siddique, N., Sze, N., & Dudev, T. (2001). Raman spectroscopy of carbon-containing particles. Vibrational Spectroscopy, 26, 179–186. Ferge, T., Karg, E., Schröppel, A., Coffee, K. R., Tobias, H. J., Frank, M. et al. (2006). Fast determination of the relative elemental and organic carbon content of aerosol samples by on-line single particle aerosol time-of-flight mass spectrometry. Environmental Science & Technology, 40, 3327–3335. Fillaux, F., Papoular, R., Lautié, A., & Tomkinson, J. (1995). Inelastic neutron-scattering study of the proton dynamics in coals. Fuels, 74, 865–873. Fox, M. A., & Staley, S. W. (1976). Determination of polycyclic aromatic hydrocarbons in atmospheric particulate matter by high pressure liquid chromatography coupled with fluorescence techniques. Analytical Chemistry, 48, 992–998. Francis, J. T., & Hitchcock, A. P. (1992). Inner-shell spectroscopy of p-benzoquinone, hydroquinone, and phenol: Distinguishing quinoid and benzenoid structures. Journal of Physical Chemistry, 96, 6598–6610. Frank, B. P., Tang, S., Lanni, T., Rideout, G., Beregszanszy, C., Meyer, N., et al. (2004). A study of the effects of fuel type and emission control systems on regulated gaseous emissions from heavy-duty diesel engines. SAE Technical Paper 2004-01-1085. Fraser, M. P., Cass, G. R., & Simoneit, B. R. T. (1998). Gas-phase and particle-phase organic compounds emitted from motor vehicle traffic in a Los Angeles roadway tunnel. Environmental Science & Technology, 32, 2051–2060. Fraser, M. P., Cass, G. R., & Simoneit, B. R. T. (1999). Particulate organic compounds emitted from motor vehicle exhaust and in the urban atmosphere. Atmospheric Environment, 33, 2715–2724. Gard, E., Mayer, J. E., Morrical, B. D., Dienes, T., Fergenson, D. P., & Prather, K. A. (1997). Real-time analysis of individual atmospheric aerosol particles: Design, and performance of a portable ATOFMS. Analytical Chemistry, 69, 4083–4091. Gautum, M., & Clark, N. N. (2003). E-55/E-59 reports on heavy-duty vehicle emissions. Coordinating research council, recent reports and study results http://www.crcao.com/. Geller, M. D., Kim, S., Misra, C., Sioutas, C., Olson, B. A., & Marple, V. A. (2002). A methodology for measuring size-dependent chemical composition of ultrafine particles. Aerosol Science and Technology, 36, 748–762. Geller, M. D., Ntziachristos, L., Mamakos, A., Samaras, Z., Schmitz, D. A., Froines, J. R. et al. (2006). Physicochemical and redox characteristics of particulate matter (PM) emitted from gasoline and diesel passenger cars. Atmospheric Environment, 40, 6988–7004. Geller, M. D., Sardar, S. B., Phuleria, H., Fine, P. M., & Sioutas, C. (2005). Measurements of particle number and mass concentrations and size distributions in a tunnel environment. Environmental Science & Technology, 39, 8653–8663. Gertler, A. W. (2005). Diesel vs. gasoline emissions: Does PM from diesel or gasoline vehicles dominate in the US?. Atmospheric Environment, 39, 2349–2355. Givens, W. A., Buck, W. H., Jackson, A., Kaldor, A., Hertzberg, A., Moehrmann, W., et al. (2003). Lube formulation effects on transfer of elements to exhaust after-treatment system components. SAE Technical Paper 2003-01-3109. Grose, M., Sakurai, H., Savstrom, J., Stolzenburg, M. R., Watts, W. F., Morgan, C. G. et al. (2006). Chemical and physical properties of ultrafine diesel exhaust particles sampled downstream of a catalytic trap. Environmental Science & Technology, 40, 5502–5507. Gross, D. S., Barron, A. R., Sukovich, E. M., Warren, B. S., Jarvis, J. C., Suess, D. T. et al. (2005). Stability of single particle tracers for differentiating between heavy- and light-duty vehicle emissions. Atmospheric Environment, 39, 2889–2901. Gross, D. S., Gälli, M. E., Silva, P. J., Wood, S. H., Liu, D.-Y., & Prather, K. A. (2000). Single particle characterization of automobile and diesel truck emissions in the Caldecott tunnel. Aerosol Science and Technology, 32, 152–163. Guerrassi, N., & Dupraz, P. (1998). A common rail injection system for high speed direct injection diesel engines. SAE Technical Paper 980803. Guo, G., Xu, N., Laing, P., Hammerle, R. H., & Maricq, M. M. (2003). Performance of a catalyzed diesel particulate filter system during soot accumulation and regeneration. SAE Technical Paper 2003-01-0047. Hahn, D. W., & Lunden, M. M. (2000). Detection and analysis of aerosol particles by laser-induced breakdown spectroscopy. Aerosol Science and Technology, 33, 30–48. Hankin, S. M., & John, P. (1999). Laser time-of-flight mass analysis of PAHs on single diesel particulates. Analytical Chemistry, 71, 1100–1104. Heeb, N. V., Ulrich, A., Emmenegger, L., Czerwinski, J., Mayer, A., & Wyser, M. (2005). Secondary emissions risk assessment of diesel particulate traps for heavy duty applications. SAE Technical Paper SIAT 2005-ABS-165.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1113

Heeb, N. V., Zennegg, M., Gujer, E., Honegger, P., Zeyer, K., Gfeller, U., et al. (2007). Secondary effects of catalytic diesel particulate filters: Copper-Induced Formation of PCDD/Fs. Environmental Science & Technology, 41, 5789–5794. HEI (2002). Understanding the health effects of components of the PM mix: Progress and next steps. HEI Perspectives, Boston, MA: Health Effects Institute. HEI (2003). Revised analyses of time-series studies of air pollution and health. Special Report, Boston, MA: Health Effects Institute. Henderson, T. R., Sun, J. D., Li, A. P., Hanson, R. L., Bechtold, W. E., Harvey, T. M. et al. (1984). GC/MS and MS/MS studies of diesel exhaust mutagenicity and emissions from chemically defined fuels. Environmental Science & Technology, 18, 428–434. Henderson, T. R., Sun, J. D., Royer, R. E., Clark, C. R., Li, A. P., Harvey, T. M. et al. (1983). Triple-quadrupole mass spectrometry studies of nitroaromatic emissions from different diesel engines. Environmental Science & Technology, 17, 443–449. Hildemann, L. M., Cass, G. R., & Markowski, G. R. (1989). A dilution stack sampler for collection of organic aerosol emissions: Design, characterization and field tests. Aerosol Science and Technology, 10, 193–204. Hildemann, L. M., Markowski, G. R., & Cass, G. R. (1991). Chemical composition of emissions from urban sources of fine organic aerosol. Environmental Science & Technology, 25, 744–759. Hildemann, L. M., Mazurek, M. A., & Cass, G. R. (1991). Quantitative characterization of urban sources of organic aerosol by high-resolution gas chromatography. Environmental Science & Technology, 25, 1311–1325. Hinz, K.-P., Kaufmann, R., & Spengler, B. (1994). Laser-induced mass analysis of single particles in the airborne state. Analytical Chemistry, 66, 2071–2076. Hinz, K.-P., Kaufmann, R., & Spengler, B. (1996). Simultaneous detection of positive and negative ions from single airborne particles by real-time laser mass spectrometry. Aerosol Science and Technology, 24, 233–242. Hodge, V. F., & Stallard, M. O. (1986). Platinum and palladium in roadside dust. Environmental Science & Technology, 20, 1058–1060. Holmén, G. A., & Ayala, A. (2002). Ultrafine PM emissions from natural gas, oxidation-catalyst diesel and particle-trap diesel heavy-duty transit buses. Environmental Science & Technology, 36, 5041–5050. Huntzicker, J. J., Johnson, R. L., Shah, J. J., & Cary, R. A. (1982). Analysis of organic and elemental carbon in ambient aerosol by a thermal-optical method. In: G. T. Wolff, & R. L. Klimisch (Eds.), Particulate carbon: Atmospheric life cycle New York: Plenum. Hybl, J. D., Lithgow, G. A., & Buckley, S. G. (2003). Laser-induced breakdown spectroscopy detection and classification of biological aerosols. Applied Spectroscopy, 57, 1207–1215. Hybl, J. D., Shane, M. T., Berry, S. R., & Jordan, M. P. (2006). Laser-induced fluorescence-cued, laser-induced breakdown spectroscopy biologicalagent detection. Applied Optics, 45, 8806–8814. Jacobson, M. Z. (2001). Strong radiative heating due to the mixing state of black carbon in atmospheric aerosols. Nature, 409, 695–697. Japar, S. M., Szkarlat, A. C., Gorse, R. A., Heyerdahl, E. K., Johnson, R. L., Rau, J. A. et al. (1984). Comparison of solvent extraction and thermal-optical carbon analysis methods: Application to diesel vehicle exhaust aerosol. Environmental Science & Technology, 18, 231–234. Jayne, J. T., Leard, D. C., Zhang, X., Davidovits, P., Smith, K. A., Kolb, C. E. et al. (2000). Development of an aerosol mass spectrometer for size and composition analysis of submicron particles. Aerosol Science and Technology, 33, 49–70. Johnson, K. S., Zuberi, B., Molina, L. T., Molina, M. M., Iedema, M. J., Cowin, J. P. et al. (2005). Processing of soot in an urban environment: Case study from the Mexico City metropolitan area. Atmospheric Chemistry and Physics, 5, 3033–3043. Johnson, T. V. (2003). Diesel emission control technology—2003 in review. SAE Technical Paper 2004-01-0070. Joutsensaari, J., Vaatovaara, P., Hämeri, K., & Laaksonen, A. (2001). A novel tandem differential mobility analyzer with organic vapor treatment of aerosol particles. Atmospheric Chemistry and Physics Discussion, 1, 1–22. Jung, H., Kittelson, D. B., & Zachariah, M. B. (2006). Characteristics of SME biodiesel-fueled diesel particle emissions and the kinetics of oxidation. Environmental Science & Technology, 40, 4949–4955. Jung, H., Kittelson, D. B., & Zachariah, M. R. (2005). The influence of a cerium additive on ultrafine diesel particle emissions and kinetics of oxidation. Combustion and Flame, 142, 276–288. Kasper, A., Aufdenblatten, S., Forss, A., Mohr, M., & Burtscher, H. (2007). Particulate emissions from a low-speed marine diesel engine. Aerosol Science and Technology, 41, 571–580. Kasper, M., Sattler, K., Siegmann, K., Matter, U., & Siegmann, H. C. (1999). The influence of fuel additives on the formation of carbon during combustion. Journal of Aerosol Science, 30, 217–225. Kawamura, K., Ng, L.-L., & Kaplan, I. R. (1985). Determination of organic acids (C1 .C10 ) in the atmosphere, motor exhausts, and engine oils. Environmental Science & Technology, 19, 1082–1086. Kellog, R., & Winberry Jr., W. (1999). EPA compendium method IO-3.3: Determination of metals in ambient particulate matter using X-ray fluorescence (XRF) spectroscopy. U.S. Environmental Protection Agency. Khair, M. K. (2003). A review of diesel particulate filter technologies. SAE Technical Paper 2003-01-2303. Khalek, I. A. (2005). 2007 Particulate measurement research. Project E-66-Phase 1 Final Report. Coordinating Research Council http://www.crcao.com/. Kim, S. H., Fletcher, R. A., & Zachariah, M. R. (2005). Understanding the difference in oxidative properties between flame and diesel soot nanoparticles: The role of metals. Environmental Science & Technology, 39, 4021–4026. Kirchstetter, T. W., Harley, R. A., Kreisberg, N. M., Stolzenburg, M. R., & Herring, S. V. (1999). On-road measurement of fine particle and nitrogen oxide emissions from light- and heavy-duty motor vehicles. Atmospheric Environment, 33, 2955–2968. Kittelson, D. B. (1998). Engines and nanoparticles: A review. Journal of Aerosol Science, 29, 575–588. Kleeman, M. J., Schauer, J. J., & Cass, G. R. (2000). Size and composition distribution of fine particulate matter emitted from motor vehicles. Environmental Science & Technology, 34, 1132–1142. Knothe, G., Sharp, C. A., & Ryan, T. W., III (2006). Exhaust emissions of biodiesel, petrodiesel, neat methyl esters, and alkanes in a new technology engine. Energy & Fuels, 20, 403–408.

1114

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

König, H. P., Hertel, R. F., Koch, W., & Rosner, G. (1992). Determination of platinum emissions from a three-way catalyst-equipped gasoline engine. Atmospheric Environment, 26A, 741–745. Kweon, C.-B., Foster, D. E., Schauer, J. J., & Okada, S. (2002). Detailed chemical composition and particle size assessment of diesel engine exhaust. SAE Technical Paper 2002-01-2670. Kweon, C.-B., Okada, S., Stetter, J. C., Christenson, C. G., Shafer, M. M., Schauer, J. J., et al. (2003a). Effect of injection timing on detailed chemical composition and particulate size distributions of diesel exhaust. SAE Technical Paper 2003-01-1794. Kweon, C.-B., Okada, S., Stetter, J. C., Christenson, C. G., Shafer, M. M., Schauer, J. J., et al. (2003b). Effect of fuel composition on combustion and detailed chemical/physical characteristics of diesel exhaust. SAE Technical Paper 2003-01-1899. Kwon, S.-B., Lee, K. W., Saito, K., Shinozaki, O., & Seto, T. (2003). Size-dependent volatility of diesel nanoparticles: Chassis dynamometer experiments. Environmental Science & Technology, 37, 1794–1802. Ladommatos, N., Abdelhalim, S. M., Zhao, H., & Hu, Z. (1998). The effects of carbon dioxide in exhaust gas recirculation on diesel engine emissions. Proceedings of the I MECH E Part D Journal of Automobile Engineering, 212, 25–42. Lambert, C., Hammerle, R., McGill, R., Khair, M., & Sharp, C. (2004). Technical advantages of urea SCR for light-duty and heavy-duty diesel vehicle applications. SAE Technical Paper 2004-01-1292. Lee, D., Miller, A., Kittelson, D., & Zachariah, M. R. (2006). Characterization of metal-bearing diesel nanoparticles using single-particle mass spectrometry. Journal of Aerosol Science, 37, 88–110. Lee, K. O., Zhu, J., Ciatti, S., Yozgatligil, A., & Choi, M. Y. (2003). Sizes, graphitic structures and fractal geometry of light-duty diesel engine particulates. SAE Technical Paper 2003-01-3169. Lee, S., Jang, M., & Kamens, R. M. (2004). SOA formation from the photooxidation of -pinene in the presence of freshly emitted diesel soot exhaust. Atmospheric Environment, 38, 2597–2605. Leister, M. (2003). USLD Composition—2007. ACES Workshop, Denver. Lelièvre, S., Bedjanian, Y., Laverdet, G., & Le Bras, G. (2004). Heterogeneous reaction of NO2 with hydrocarbon flame soot. Journal of Physical Chemistry A, 108, 10807–10817. Lev-On, M., LeTavec, C., Uihlein, J., Alleman, T. L., Lawson, D. R., Vertin, K., et al. (2002). Chemical speciation of exhaust emissions from trucks and busses fueled on ultra-low sulfur diesel and CNG. SAE Technical Paper 2002-01-0432. Lev-On, M., LeTavec, C., Uihlein, J., Kimura, K., Alleman, T. L., Lawson, D. R., et al. (2002a). Speciation of organic compounds from the exhaust of trucks and buses: Effect of fuel and after-treatment on vehicle emission profiles. SAE Technical Paper 2002-01-2873. Lev-On, M., & Zielinska, B. (2004). Compilation of PM characterization data for diesel fueled heavy-duty trucks and buses: Impact of tailpipe controls and fuel sulfur on the EC/OC ratio and the particles’ chemical species content. SAE Technical Paper 2004-01-1991. Lies, K. H., Hartung, A., Postulka, A., Gring, H., & Schulze, J. (1986). Composition of diesel exhaust with particular reference to particle bound organics including formation of artifacts. Developments in Toxicology and Environmental Science, 13, 65–82. Lin, Y.-C., Lee, W.-J., Wu, T.-S., & Wang, C.-T. (2006). Comparison of PAH and regulated harmful matter emissions from biodiesel blends and paraffinic fuel blends on engine accumulated mileage test. Fuel, 85, 2516–2523. Liu, P., Ziemann, P. J., Kittelson, D. B., & McMurry, P. H. (1995a). Generating particle beams of controlled dimensions and divergence: I. Theory of particle motion in aerodynamic lenses and nozzle expansions. Aerosol Science and Technology, 22, 293–313. Liu, P., Ziemann, P. J., Kittelson, D. B., & McMurry, P. H. (1995b). Generating particle beams of controlled dimensions and divergence: II. Experimental evaluation of particle motion in aerodynamic lenses and nozzle expansions. Aerosol Science and Technology, 22, 314–324. Liu, Z. G., Verdegan, B. M., Badeau, K. M. A., & Sonsalla, T. G. (2002). Measuring the fractional efficiency of diesel particulate filters. SAE Technical Paper 2002-01-1007. Lloyd, A. C., & Cackette, T. A. (2001). Diesel engines: Environmental impact and control. Journal of the Air Waste & Management Association, 51, 809–847. Lombaert, K. et al. (2004). Nondestructive analysis of metallic elements in diesel soot collected on filter: Benefits of laser induced breakdown spectroscopy. Chemistry of Material Science, 24, 41–56. Lough, G. C., Schauer, J. J., Park, J.-S., Shafer, M. M., Deminter, J. T., & Weinstein, J. P. (2005). Emissions of metals associated with motor vehicle roadways. Environmental Science & Technology, 39, 826–836. Lowenthal, D. H., Zielinska, B., Chow, J. C., Watson, J. G., Gautam, M., & Stevens, K. D. (1994). Characterization of heavy-duty diesel vehicle emissions. Atmospheric Environment, 28, 731–743. Maricq, M. M., Chase, R. E., Xu, N., & Laing, P. M. (2002). The effects of the catalytic converter and fuel sulfur level on motor vehicle particulate matter emissions: Light duty diesel vehicles. Environmental Science & Technology, 36, 283–289. Mathis, U., Mohr, M., & Forss, A.-M. (2005). Comprehensive particle characterization of modern gasoline and diesel passenger cars at low ambient temperatures. Atmospheric Environment, 39, 107–117. Mayer, A., Czerwinski, J., Wyser, M., Mattrel, P., & Heitzer, A. (2005). Impact of RME/diesel blends on particle formation, particle filtration, and PAH emissions. SAE Technical Paper 2005-01-1728. McDonald, J. D., Barr, E. B., White, R. K., Chow, J. C., Schauer, J. J., Zielinska, B. et al. (2004). Generation and characterization of four dilutions of diesel engine exhaust for a subchronic inhalation study. Environmental Science & Technology, 38, 2513–2522. McDonald, J. D., Harrod, K. S., Seagrave, J.-C., Seilkop, S. K., & Mauderly, J. L. (2004). Effects of low sulfur fuel and a catalyzed particle trap on the composition and toxicity of diesel emissions. Environmental Health Perspectives, 112, 1307–1312. McDonald, J. D., Zielinska, B., McDaniel, M. R., & Mousset-Jones, P. (2003). Source apportionment of airborne fine particulate matter in an underground mine. Journal of the Air & Waste Management Association, 53, 386–395. McDonald, J. D., Zielinska, B., Sagebiel, J. C., & McDaniel, M. R. (2002). Characterization of fine particle material in ambient air and personal samples from an underground mine. Aerosol Science and Technology, 36, 1033–1044. MECA (2006). Case studies of the use of exhaust emission controls on locomotives and large marine diesel engines http://www.meca.org/ galleries/default-file/MECA%20locomotive%20and%20marine%20case%20study%20report%201006.pdf.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1115

Miguel, A. H., Kirchstetter, T. W., Harley, R. A., & Hering, S. V. (1998). On-road emissions of particulate polycyclic aromatic hydrocarbons and black carbon from gasoline and diesel vehicles. Environmental Science & Technology, 32, 450–455. Miller, A., Ahlstrand, G., Kittelson, D., & Zachariah, M. (2007). The fate of metal (Fe) during diesel combustion: Morphology, chemistry, and formation pathways of nanoparticles. Combustion and Flame, 149, 129–143. Mohr, M., Lehmann, U., & Margaria, G. (2003a). ACEA programme on the emissions of fine particulates from passenger cars (2). Part 1. Particle characterization of a wide range of engine technologies. SAE Technical Paper 2003-01-1889. Mohr, M., Lehmann, U., & Margaria, G. (2003b). ACEA programme on the emissions of the fine particulates from passenger cars (2). Part 2. Effect of sampling conditions and fuel sulphur content on the particle emission. SAE Technical Paper 2003-01-1890. Moosmüller, H., Arnott, W. P., Rogers, C. F., Bowen, J. L., Gillies, J. A., Pierson, W. R. et al. (2001). Time resolved characterization of diesel particulate emissions. 2. Instruments for elemental and organic carbon measurements. Environmental Science Technology, 35, 1935–1942. Murphy, D. M., & Thomson, D. S. (1995). Laser ionization mass spectroscopy of single aerosol particles. Aerosol Science and Technology, 22, 237–249. Myneni, S. C. B. (2002). Soft X-ray spectroscopy and spectromicroscopy studies of organic molecules in the environment. In P. Fenter, M. Rivers, N. Sturchio, & S. Sutton (Eds.), Reviews in mineralogy and geochemistry applications of synchrotron radiation in low-temperature geochemistry and environmental science (Vol. 49, pp. 485–579). Myojo, T., Takaya, M., & Ono-Ogasawara, M. (2002). DMA as a gas converter from aerosol to “argonsol” for real-time chemical analysis using ICP-MS. Aerosol Science and Technology, 36, 76–83. Nelson, M. P., Zugates, C. T., Treado, P. J., Casuccio, G. S., Exline, D. L., & Schlaegle, S. F. (2001). Combining Raman chemical imaging and scanning electron microscopy to characterize ambient fine particulate matter. Aerosol Science and Technology, 34, 108–117. Newton, D. L., Erickson, M. D., Tomer, K. B., Pellizzari, E. D., Gentry, P., & Zweidinger, R. B. (1982). Identification of nitroaromatics in diesel exhaust particulate using gas ghromatorgraphy/negative ion chemical ionization mass spectrometry and other techniques. Environmental Science & Technology, 16, 206–213. NIOSH (1996). Elemental carbon (diesel exhaust). In NIOSH manual of analytical methods. Cincinnati, OH: National Institute of Occupational Safety and Health. Noble, C.A., & Prather, K.A. (1996). Real-time measurement of correlated size and composition profiles of individual aerosol particles. Environmental Science & Technology, 30, 2667–2680. Nolte, C. G., Schauer, J. J., Cass, G. R., & Simoneit, B. R. T. (2002). Trimethylsilyl derivatives of organic compounds in source samples and in atmospheric fine particulate matter. Environmental Science & Technology, 36, 4273–4281. NREL (2007). Gasoline/diesel PM split study. National Renewable Energy Laboratory http://www.nrel.gov/vehiclesandfuels/nfti/feat_split_ study.html. Ntziachristos, L., Mamakos, A., Samaras, Z., Mathis, U., Mohr, M., Thompson, N., et al. (2004). Overview of the European “Particulates” Project on the characterisation of exhaust particulate emissions from road vehicles: Results for light-duty vehicles. SAE Technical Paper 2004-01-1985. Odum, J. R., Jungkamp, T. P. W., Griffin, R. J., Flagan, R. C., & Seinfeld, J. H. (1997). The atmospheric aerosol-forming potential of whole gasoline vapor. Science, 276, 96–99. Odziernkowski, M., Koziel, J. A., Irish, D. E., & Pawliszyn, J. (2001). Sampling and Raman confocal microspectroscopic analysis of airborne particulate matter using poly(dimethylsiloxane) solid-phase microextraction fibers. Analytical Chemistry, 73, 3131–3139. Oehme, M., Larssen, S., & Brevik, E. M. (1991). Emission factors of PCDD and PCDF for road vehicles obtained by tunnel experiment. Chemosphere, 23, 1699–1708. Okada, S., Kweon, C.-B., Stetter, J. C., Foster, D. E., Shafer, M. M., Christensen, C. G., et al. (2003). Measurement of trace metal composition in diesel engine particulate and its potential for determining oil consumption. SAE Technical Paper 2003-01-0076. Okada, Y., Yabumoto, J., & Takeuchi, K. (2002). Aerosol spectrometer for size and composition analysis of nanoparticles. Journal of Aerosol Science, 33, 961–965. Ottesen, D. K., Wang, J. C. F., & Radziernski, L. J. (1989). Real-time laser spark spectroscopy of particulates in combustion environments. Applied Spectroscopy, 43, 967–976. Palacios, M. A., Gómez, M. M., Moldovan, M., Morrison, G., Rauch, S., McLeod, C. et al. (2000). Platinum-group elements: Quantification in collected exhaust fumes and studies of catalyst surfaces. Science of the Total Environment, 257, 1–15. Park, C., Kook, S., & Bae, C. (2004). Effects of multiple injections in a HSDI diesel engine equipped with common rail injection system. SAE Technical Paper 2004-01-0127. Penner, J. E., Chuang, C. C., & Grant, K. (1998). Climate forcing by carbonaceous and sulfate aerosols. Climate Dynamics, 14, 839–851. Pierson, W. R., & Brachaczek, W. W. (1983). Particulate matter associated with vehicles on the road. II. Aerosol Science and Technology, 2, 1–40. Piñeiro-Iglesias, M., López-Mahía, P., Vázquez-Blanco, E., Muniategui-Lorenzo, S., & Prada-Rodríquez, D. (2002). Problems in the extraction of polycyclic aromatic hydrocarbons from diesel particulate matter. Polycyclic Aromatic Compounds, 22, 129–146. Pope III, C. A., Burnett, R. T., Thurston, G. D., Thun, M. J., Calle, E. E., Krewski, D. et al. (2004). Cardiovascular mortality and long-term exposure to particulate air pollution. Circulation, 109, 71–77. Prather, K. A., Nordmeyer, T., & Salt, K. (1994). Real-time characterization of individual aerosol particles using time-of-flight mass spectrometry. Analytical Chemistry, 66, 1403–1407. Rauch, S., Hemond, H. F., Peucker-Ehrenbrink, B., Ek, K. H., & Morrison, G. M. (2005). Platinum group element concentrations and osmium isotopic composition in urban airborne particles from Boston Massachusetts. Environmental Science & Technology, 39, 9464–9470. Reents, W. D., & Ge, Z. (2000). Simultaneous elemental composition and size distributions of submicron particles in real time using laser atomization/ionization mass spectrometry. Aerosol Science and Technology, 33, 122–134. Reilly, P. T. A., Gieray, R. A., Whitten, W. B., & Ramsey, J. M. (1998). Real-time characterization of the organic composition and size of individual diesel engine smoke particles. Environmental Science & Technology, 32, 2672–2679.

1116

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Riddle, S. G., Robert, M. A., Jakober, C. A., Hannigan, M. P., & Kleeman, M. J. (2007). Size distribution of trace organic species emitted from heavy-duty diesel vehicles. Environmental Science & Technology, 41, 1962–1969. Risby, T. H., & Lestz, S. S. (1983). Is the direct mutagenic activity of diesel particulate matter a sampling artifact?. Environmental Science & Technology, 17, 621–624. Ristimäki, J., Keskinen, J., Virtanen, A., Maricq, M., & Aakko, P. (2005). Cold temperature PM emissions measurement: Method evaluation and application to light duty vehicles. Environmental Science & Technology, 39, 9424–9430. Ro, C.-U., Musselman, I. H., & Linton, R. W. (1991). Molecular speciation of microparticles: Application of pattern recognition techniques to laser microprobe mass spectrometric data. Analytica Chimica Acta, 243, 139–147. Robinson, A. L., Donahue, N. M., Shrivastava, M. K., Weitkamp, E. A., Sage, A. M., Grieshop, A. P. et al. (2007). Rethinking organic aerosols: Semivolatile emissions and photochemical aging. Science, 315, 1259–1262. Rogge, W. F., Hildemann, L. M., Mazurek, M. A., & Cass, G. R. (1993). Sources of fine organic aerosol. 2. Noncatalyst and catalyst-equipped automobiles and heavy-duty diesel trucks. Environmental Science & Technology, 27, 636–651. Rosen, H., & Novakov, T. (1977). Raman scattering and the characterization of atmospheric aerosol particles. Nature, 266, 708–710. Sadezky, A., Muckenhuber, H., Grothe, H., Niessner, R., & Pöschl, U. (2005). Raman microspectroscopy of soot and related carbonaceous materials: Spectral analysis and structural information. Carbon, 43, 1731–1742. Sagebiel, J. C., Zielinska, B., Walsh, P. A., Chow, J. C., Cadle, S. H., Mulawa, P. A. et al. (1997). PM-10 exhaust samples collected during IM-240 dynamometer tests of in-service vehicles in Nevada. Environmental Science & Technology, 31, 75–83. Sakurai, H., Park, K., McMurry, P. H., Zarling, D. D., Kittelson, D. B., & Ziemann, P. J. (2003). Size-dependent mixing characteristics of volatile and nonvolatile components in diesel exhaust aerosols. Environmental Science & Technology, 37, 5487–5495. Sakurai, H., Tobias, H. J., Park, K., Zarling, D., Docherty, K. S., Kittelson, D. B. et al. (2003). On-line measurements of diesel nanoparticle composition and volatility. Atmospheric Environment, 37, 1199–1210. Schauer, J. J. (2003). Evaluation of elemental carbon as a marker for diesel particulate matter. Journal of Exposure Analysis and Environmental Epidemiology, 13, 443–453. Schauer, J. J., Kleeman, M. J., Cass, G. R., & Simoneit, B. R. T. (1999a). Measurement of emissions from air pollution sources. 1. C1 through C29 organic compounds from meat charbroiling. Environmental Science & Technology, 33, 1566–1577. Schauer, J. J., Kleeman, M. J., Cass, G. R., & Simoneit, B. R. T. (1999b). Measurement of emissions from air pollution sources. 2. C1 through C30 organic compounds from medium duty diesel trucks. Environmental Science & Technology, 33, 1578–1587. Schauer, J. J., Kleeman, M. J., Cass, G. R., & Simoneit, B. R. T. (2002). Measurement of emissions from air pollution sources. 5. C1 –C32 organic compounds from gasoline-powered motor vehicles. Environmental Science & Technology, 36, 1169–1180. Schauer, J. J., Rogge, W. F., Hildemann, L. M., Mazurek, M. A. R., Christensen, C. G., & Simoneit, B. R. T. (1996). Source apportionment of airborne particulate matter using organic compounds as tracers. Atmospheric Environment, 30, 3837–3855. Scheer, V., Kirchner, U., Casati, R., Vogt, R., Wehner, B., Philippin, S., et al. (2005). Composition of semi-volatile particles from diesel exhaust. SAE Technical Paper, 2005-01-0197. Schneider, J., Hock, N., Weimer, S., Borrmann, S., Kirchner, U., Vogt, R. et al. (2005). Nucleation particles in diesel exhaust: Composition inferred from in situ mass spectrometric analysis. Environmental Science & Technology, 39, 6153–6161. Seagrave, J.-C., McDonald, J. D., Gigliotti, A. P., Nikula, K. J., Seilkop, S. K., Gurevich, M. et al. (2002). Mutagenicity and in vivo toxicity of combined particulate and semivolatile organic fractions of gasoline and diesel engine emissions. Toxicological Sciences, 70, 212–226. Seagrave, J. C., Seilkop, S. K., & Mauderly, J. L. (2003). In vitro relative toxicity screening of combined particulate and semivolatile organic fractions of gasoline and diesel engine emissions. Journal of Toxicology and Environmental Health, 66, 1113–1132. Shah, S., Cocker III, D. R., Johnson, K. C., Lee, J. M., Soriano, B. L., & Miller, J. W. (2006). Emissions of regulated pollutants from in-use diesel back-up generators. Atmospheric Environment, 40, 4199–4209. Shah, S. D., Cocker, D. R., Miller, J. W., & Norbeck, J. M. (2004). Emission rates of particulate matter and elemental and organic carbon from in-use diesel engines. Environmental Science & Technology, 38, 2544–2550. Shah, S. D., Ogunyoku, T. A., Miller, J. W., & Cocker, D. R., III (2005). On-road emission rates of PAH and n-alkane compounds from heavy-duty diesel vehicles. Environmental Science & Technology, 39, 5276–5284. Shattuck, T. W., Germani, M. S., & Buseck, P. R. (1991). Multivariate statistics for large data sets: Applications to individual aerosol particles. Analytical Chemistry, 63, 2646–2656. Shi, J. P., Mark, D., & Harrison, R. M. (2000). Characterization of particles from a current technology heavy-duty diesel engine. Environmental Science & Technology, 34, 748–755. Silva, P. J., & Prather, K. A. (1997). On-line characterization of individual particles from automobile emissions. Environmental Science & Technology, 31, 3074–3080. Silva, P. J., & Prather, K. A. (2000). Interpretation of mass spectra from organic compounds in aerosol time-of-flight mass spectrometry. Analytical Chemistry, 72, 3553–3562. Simão, J., Ruiz-Agudo, E., & Rodriguez-Navarro, C. (2006). Effects of particulate matter from gasoline and diesel vehicle exhaust emissions on silicate stones sulfation. Atmospheric Environment, 40, 6905–6917. Sioutas, C., Kim, S., & Chang, M. (1999). Development and evaluation of a prototype ultrafine particle concentrator. Journal of Aerosol Science, 30, 1001–1012. Sioutas, C., Koutrakis, P., Godleski, J. J., Ferguson, S. T., Kim, C. S., & Burton, R. M. (1997). Fine particle concentrators for inhalation exposures—effect of particle size and composition. Journal of Aerosol Science, 28, 1057–1071. Sluder, C. S., Storey, J. M. E., Lewis, S. A., & Lewis, L. A. (2005). Low temperature urea decomposition and SCR performance. SAE Technical Paper, 2005-01-1858. Small, H. (1989). Ion chromatography. New York: Plenum.

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

1117

Sodeman, D. A., Toner, S. M., & Prather, K. A. (2005). Determination of single particle mass spectral signatures from light duty vehicle emissions. Environmental Science & Technology, 39, 4569–4580. Song, J., Alam, M., Boehman, A. L., & Kim, U. (2006). Examination of the oxidation behavior of biodiesel soot. Combustion and Flame, 146, 589–604. Song, X. H., Hopke, P. K., Fergenson, D. P., & Prather, K. A. (1999). Classification of single particles analyzed by ATOFMS using an artificial neural network ART-2A. Analytical Chemistry, 71, 860–865. Spencer, M. T., Schields, L. G., Sodeman, D. A., Toner, S. M., & Prather, K. A. (2006). Comparison of oil and fuel particle chemical signatures with particle emissions from heavy and light duty vehicles. Atmospheric Environment, 40, 5224–5235. Stöhr, J. (1992). NEXAFS spectroscopy. Springer series in surface science (Vol. 25). Berlin: Springer. Stratakis, G. A., Konstantas, G. S., & Stamatelos, A. M. (2003). Experimental investigation of the role of soot volatile organic fraction in the regeneration of diesel filters. Proceedings of the Institution of Mechanical Engineers D, 217, 307–317. Stratakis, G. A., & Stamatelos, A. M. (2003). Thermogravimetric analysis of soot emitted by a modern diesel engine run on catalyst-doped fuel. Combustion and Flame, 132, 157–169. Stumpp, G., & Ricco, M. (1996). Common rail—an attractive fuel injection system for passenger car DI diesel engines. SAE Technical Paper, 960870. Suess, D. T., & Prather, K. A. (1999). Mass spectrometry of aerosols. Chemical Reviews, 99, 3007–3035. Suess, D. T., & Prather, K. A. (2002). Reproducibility of single particle chemical composition during a heavy duty diesel truck dynamometer study. Aerosol Science and Technology, 36, 1139–1141. Szkarlat, A. C., & Japar, S. M. (1983). Optical and chemical properties of particle emissions from on-road vehicles. Journal of the Air Pollution Control Association, 33, 592–597. Tan, P. V., Malpica, O., Evans, G. J., Owega, S., & Fila, M. S. (2002). Chemically-assigned classification of aerosol mass spectra. Journal of the American Society for Mass Spectrometry, 13, 826–838. Thalagavara, A. M., Johnson, J. H., Bagley, S. T., & Shende, A. S. (2005). The effects of a catalyzed particulate filter and ultra low sulfur fuel on heavy duty diesel engine emissions. SAE Technical Paper 2005-01-0473. Tobias, H. J., Beving, D. E., Ziemann, P. J., Sakurai, H., Zuk, M., McMurry, P. H. et al. (2001). Chemical analysis of diesel engine nanoparticles using a nano-DMA/thermal desorption particle beam mass spectrometer. Environmental Science & Technology, 35, 2233–2243. Tobias, H. J., Kooiman, P. M., Docherty, K. S., & Ziemann, P. J. (2000). Real-time chemical analysis of organic aerosols using a thermal desorption particle beam mass spectrometer. Aerosol Science Technology, 33, 170–190. Toner, S. M., Sodeman, D. A., & Prather, K. A. (2006). Single particle characterization of ultrafine and accumulation mode particles from heavy duty diesel vehicles using aerosol time-of-flight mass spectrometry. Environmental Science & Technology, 40, 3912–3921. Twigg, M. V. (2007). Progress and future challenges in controlling automotive exhaust gas emissions. Applied Catalysis B, 70, 2–15. US EPA (2001). Control of air pollution from new motor vehicles: Heavyduty engine and vehicle standards and highway diesel fuel sulfur control requirements. Final Rule, Federal Register, 66, 5002–5173. US EPA (2004). Clean air nonroad diesel—tier 4 Final Rule http://www.epa.gov/nonroad-diesel/2004fr.htm. US EPA (2007a). MOVES (Motor Vehicle Emission Simulator) http://www.epa.gov/otaq/ngm.htm. US EPA (2007b). Control of emissions of air pollution from locomotive engines and marine compression-ignition engines less than 30 liters per cylinder. Federal Register, 72, 15937–15986 http://www.epa.gov/fedrgstr/EPA-AIR/2007/April/Day-03/a1107a.htm. Vaaraslahti, K., Virtanen, A., Ristimäki, J., & Keskinen, J. (2004a). Nucleation mode formation in heavy-duty diesel exhaust with and without a particulate filter. Environmental Science & Technology, 38, 4884–4890. Vaaraslahti, K., Virtanen, A., Ristimäki, J., & Keskinen, J. (2004b). Effect of after-treatment systems on size distribution of heavy duty diesel exhaust aerosol. SAE Technical Paper 2004-01-1980. van Setten, B. A. A. L., Makkee, M., & Moulijn, J. A. (2001). Science and technology of catalytic diesel particulate filters. Catalysis Reviews, 43, 489–564. Venkataraman, C., Lyons, J. M., & Friedlander, S. K. (1994). Size distributions of polycyclic aromatic hydrocarbons and elemental carbon. 1. Sampling, measurement methods, and source characterization. Environmental Science & Technology, 28, 555–562. Vogt, R., Kirchner, U., Scheer, V., Hinz, K. P., Trimborn, A., & Spengler, B. (2003). Identification of diesel exhaust particles at an autobahn, urban and rural location using single-particle mass spectrometry. Journal of Aerosol Science, 34, 319–337. Vogt, R., Scheer, V., Casati, R., & Benter, T. (2003). On-road measurement of particle emission in the exhaust plume of a diesel passenger car. Environmental Science & Technology, 37, 4070–4076. Voisin, D., Smith, J. N., Sakurai, H., McMurry, P. H., & Eisele, F. L. (2003). Thermal desorption chemical ionization mass spectrometer for ultrafine particle chemical composition. Aerosol Science and Technology, 37, 471–475. Vouitsis, E., Ntziachristos, L., & Samaras, Z. (2003). Particulate matter mass measurements for low emitting diesel powered vehicles: What’s next?. Progress in Energy Combustion Science, 29, 635–672. Wang, W. G., Lyons, D. W., Clark, N. N., Gautam, M., & Norton, P. M. (2000). Emissions from nine heavy trucks fueled by diesel and biodiesel blend without engine modification. Environmental Science & Technology, 34, 933–939. Warner, J. R., Johnson, J. H., Bagley, S. T., & Huynh, C. T. (2003). Effects of a catalyzed particulate filter on emissions from a diesel engine: Chemical characterization data and particulate emissions measured with thermal optical and gravimetric methods. SAE Technical Paper 2003-01-0049. Watson, J. G., Chow, J. C., Lowenthal, D. H., Pritchett, L. C., Frazier, C. A., Neuroth, G. R. et al. (1994). Differences in the carbon composition of source profiles for diesel- and gasoline-powered vehicles. Atmospheric Environment, 28, 2493–2505. Wehner, B., Philippin, S., Wiedensohler, A., Scheer, V., & Vogt, R. (2004). Variability of non-volatile fractions of atmospheric aerosol particles with traffic influence. Atmospheric Environment, 38, 6081–6090. Weingartner, E., Burtscher, H., & Baltensperger, U. (1997). Hygroscopic properties of carbon and diesel soot particles. Atmospheric Environment, 31, 2311–2327.

1118

M. Matti Maricq / Aerosol Science 38 (2007) 1079 – 1118

Weingartner, E., Gysel, M., & Baltensperger, U. (2002). Hygroscopicity of aerosol particles at low temperatures. 1. New low-temperature H-TDMA instrument: Setup and first applications. Environmental Science & Technology, 36, 55–62. Weingartner, E., Keller, C., Stahel, W. A., Burtscher, H., & Baltensperger, U. (1997). Aerosol emission in a road tunnel. Atmospheric Environment, 31, 451–462. Weiss, M. A., Heywood, J. B., Drake, E., Schafer, A., & AuYeung, F. (2000). On the road in 2020: A life-cycle analysis of new automobile technologies. Energy Laboratory Report #MIT EL-00-003, Massachusetts Institute of Technology. West, B., Huff, S., Parks, J., Lewis, S., Choi, J.-S., Partridge, W., et al. (2004). Assessing reductant chemistry during in-cylinder regeneration of diesel lean NOx traps. SAE Technical Paper 2004-01-3023. Westerholm, R., Christensen, A., Törnqvist, M., Ehrenberg, L., Rannug, U., Sjögren, M. et al. (2001). Comparison of exhaust emissions from Swedish environmental classified diesel fuel (MK1) and European program on emissions, fuels and engine technologies (EPEFE) reference fuel: A chemical and biological characterization, with viewpoints on cancer risk. Environmental Science & Technology, 35, 1748–1754. Westerholm, R., & Egebäck, K.-E. (1994). Exhaust emissions from light- and heavy-duty vehicles: Chemical composition, impact of exhaust after treatment, and fuel parameters. Environmental Health Perspectives, 102(Suppl. 4), 13–23. Westerholm, R., & Li, H. (1994). A multivariate statistical analysis of fuel-related polycyclic aromatic hydrocarbon emissions from heavy-duty vehicles. Environmental Science & Technology, 28, 965–972. Whitacre, S. D. (2003). 2007 Lubricant impact on emissions. ACES Workshop, Denver. Woods, E., Smith, G. D., Dessiaterik, Y., Baer, T., & Miller, R. E. (2001). Quantitative detection of aromatic compounds in single aerosol particle mass spectrometry. Analytical Chemistry, 73, 2317–2322. Zelenka, P., Egert, M., & Cartellieri, W. (2000). Ways to meet future emissions standards with diesel engine powered sport utility vehicles SUV. SAE Technical Paper 2000-01-0181. Zereini, F., Alt, F., Messerchmidt, J., Wiseman, C. Feldmann, I., von Bohlen, A., et al. (2005). Concentration and distribution of heavy metals in urban airborne particulate matter in Frankfurt am Main, Germany. Environmental Science & Technology, 39, 2983–2989. Zereini, F., Wiseman, C., Alt, F., Messerschmidt, J., Müller, J., & Urban, H. (2001). Platinum and rhodium concentrations in airborne particulate matter in Germany from 1988 to 1998. Environmental Science & Technology, 35, 1996–2000. Zereini, F., Wiseman, C., & Püttmann, W. (2007). Changes in palladium, platinum, and rhodium concentrations, and their spatial distribution in soils along a major highway in Germany from 1994 to 2004. Environmental Science & Technology, 41, 451–456. Zhan, O., Voumard, P., & Zenobi, R. (2005). Application of two-step laser mass spectrometry to the chemical analysis of aerosol particle surfaces. Rapid Communications in Mass Spectrometry, 9, 119–127. Zhang, H., Singh, J. P., Yueh, F. Y., & Cook, R. L. (1995). Laser-induced breakdown spectra in a coal-fired MHD facility. Applied Spectroscopy, 49, 1617–1623. Zhang, H., Yueh, F.-Y., & Singh, J. P. (1999). Laser-induced breakdown spectrometry as a multimetal continuous-emission monitor. Applied Optics, 38, 1459–1466. Zheng, M., Reader, G. T., & Hawley, J. G. (2004). Diesel engine exhaust gas recirculation—a review on advanced and novel concepts. Energy Conversion & Management, 45, 883–900. Zhu, Y., Hinds, W. C., Kim, S., Shen, S., & Sioutas, C. (2002). Study of ultrafine particles near a major highway with heavy-duty diesel traffic. Atmospheric Environment, 36, 4323–4335. Zielinska, B., Sagebiel, J., Arnott, W. P., Rogers, C. F., Kelly, K. E., Wagner, D. A. et al. (2004). Phase and size distribution of polycyclic aromatic hydrocarbons in diesel and gasoline vehicle emissions. Environmental Science & Technology, 38, 2557–2567. Zielinska, B., Sagebiel, J., McDonald, J. D., Whitney, K., & Lawson, D. R. (2004). Emission rates and comparative chemical composition from selected in-use diesel and gasoline fueled vehicles. Journal of the Air Waste Management Association, 54, 1138–1150. Zielinska, B., Sagebiel, J., Stockwell, W., McDonald, J., Seagrave, J.-C., Wiesen, P., et al. (2006). Investigation of atmospheric transformations of diesel emissions in the European photoreactor (EUPHORE). In I. Barnes, & K. J. Rudzinski (Eds.), NATO science series (Vol. 62, pp. 279–284). Netherlands: Springer. Zuberi, B., Johnson, K. S., Aleks, G. K., Molina, L. T., Molina, M. J., & Laskin, A. (2005). Hydrophilic properties of aged soot. Geophysical Research Letters, 32, L01807.