Influence of Oxygenated Organic Aerosols (OOAs) on the Oxidative

Jun 13, 2013 - ... burning aerosol: a modeling case study of the 2010 mega-fire event in ... Berezin , H. Petetin , T. Mielonen , I. N. Kuznetsova , M...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/est

Influence of Oxygenated Organic Aerosols (OOAs) on the Oxidative Potential of Diesel and Biodiesel Particulate Matter S. Stevanovic,†,‡ B. Miljevic,† N. C. Surawski,†,§ K. E. Fairfull-Smith,‡ S. E. Bottle,‡ R. Brown,∥ and Z. D. Ristovski*,†,∥ †

International Laboratory for Air Quality and Health, ‡ARC Centre of Excellence for Free Radical Chemistry and Biotechnology, and Biofuels Engine Research Facility, Queensland University of Technology, 2 George Street, Brisbane, Queensland 4001, Australia



ABSTRACT: Generally, the magnitude of pollutant emissions from diesel engines running on biodiesel fuel is ultimately coupled to the structure of the fuel's constituent molecules. Previous studies demonstrated the relationship between the organic fraction of particulate matter (PM) and its oxidative potential. Herein, emissions from a diesel engine running on different biofuels were analyzed in more detail to explore the role that different organic fractions play in the measured oxidative potential. In this work, a more detailed chemical analysis of biofuel PM was undertaken using a compact time of flight aerosol mass spectrometer (c-ToF AMS). This enabled a better identification of the different organic fractions that contribute to the overall measured oxidative potentials. The concentration of reactive oxygen species (ROS) was measured using a profluorescent nitroxide molecular probe 9-(1,1,3,3tetramethylisoindolin-2-yloxyl-5-ethynyl)-10-(phenylethynyl)anthracene (BPEAnit). Therefore, the oxidative potential of the PM, measured through the ROS content, although proportional to the total organic content in certain cases, shows a much higher correlation with the oxygenated organic fraction as measured by the c-ToF AMS. This highlights the importance of knowing the surface chemistry of particles for assessing their health impacts. It also sheds light onto new aspects of particulate emissions that should be taken into account when establishing relevant metrics for assessing health implications of replacing diesel with alternative fuels.

1. INTRODUCTION There is an increasing need for the adoption of biofuels as alternatives to petroleum-based fuels to reduce greenhouse gas emissions and improve the long-term availability of fuel resources. To achieve this, biofuels need to meet a number of criteria, such as high energy content, low production costs, and sustainability of the food supply. In addition, health and environmental aspects of biofuels need to be assessed before their implementation is put in place. Biodiesels currently present on the market are mainly methyl and ethyl esters of fatty acids from plant and animal provenance. Biodiesel can be used as a neat fuel in current automotive engines, or it can be blended with petroleum diesel. Generally, there is a consensus that the combustion of biodiesel results in the reduction of the total mass of particulate matter (PM), reduced soot emissions, and a substantial decrease in CO.1,2 In addition, biodiesels have low sulfur and aromatic contents. However, some studies have shown that biodiesel combustion usually produces increased levels of NOx (a known ozone precursor),3 increased particle-bound organic carbon,4 and decreased particle diameter.5,6 All of these factors can influence health prospects of biodiesel use in compression ignition (CI) engines.7 © XXXX American Chemical Society

It has also been reported that biodiesel increases emissions of some carbonyl species.8−10 Because vehicles are already the major source of carbonyl compounds in urban air, this aspect needs to be explored in more detail. Greater biodiesel use could lead to higher levels of carbonyl compounds, and these compounds can be significant precursors for secondary pollutants.10 Generally, the magnitude of pollutant emissions from diesel engines running on biodiesel is related to the chemical structure of molecules in the fuel. The presence of oxygen atoms in the molecular structure of the substances present in biodiesel could ultimately give rise to significant levels of oxygenated toxic species (including typically formaldehyde and acetaldehyde).11 Taking into account the potential of biodiesel to form aldehydes, the potential for the increase in levels of the soluble organic fraction, and the presence of different additives that may have an impact on the toxicity of biodiesel exhaust, there is a clear need for more detailed research in this area. Received: February 20, 2013 Revised: June 3, 2013 Accepted: June 13, 2013

A

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Previous studies have demonstrated that there is a relationship between the organic fraction and the oxidative potential (OP) of PM. The emissions from burning logwood, after removing all of the semi-volatile species by heating the aerosol stream inside a thermodenuder, resulted in the reduction of measured OP by 80−100%12 as measured using the in-house developed profluorescent nitroxide probes.13 A similar reduction in PM OP after thermal conditioning was also reported by Biswas et al.14,15 In this study, they measured the OP of PM produced by heavy duty vehicles using dithiothreitol (DTT) as an assay. The role of organic content in the particle OP was further explored in the studies that included diesel particulate matter (DPM) and various alternative fuels. Fuels investigated included ethanol,16 Fischer−Tropsch diesel (gas to liquid), and various biodiesel feedstocks (soy, canola, and tallow) in various blend percentages.4 All of these studies support the link between the oxygen content in the fuel and OP. All of the previous studies have shown that the semivolatile component of PM is the most important factor influencing the OP of these particles. Although a clear link between the OP and the organic fraction of PM exists, in the case of biodiesel PM,4 no correlation was observed when the reactive oxygen species (ROS) concentration was plotted against the organic volume percentage. The study looking at wood smoke12 showed a good linear correlation between the OP and organic matter but only for the same burning conditions. This indicates that the relationship between these two parameters is complex, and a more detailed investigation needs to be undertaken before it will be possible to rationalize the observed results. The primary objective of the work presented here was to characterize emissions from a diesel engine running on different alternative fuels to gain a better understanding of the role that different organic fractions play in influencing biodiesel particle surface chemistry. In this work, a detailed chemical analysis of biodiesel PM was undertaken using a compact time of flight aerosol mass spectrometer (c-ToF AMS), which enabled identification of the different organic fractions that contribute to the overall OPs measured. Despite the reduction in the total mass of PM generated as well as reduced CO and polycyclic aromatic hydrocarbon (PAH) emissions, the results reported herein indicate that combustion of alternative fuels leads to greater oxidative reactivity for PM and, therefore, potentially higher health risks associated with exposure to those. This finding may require authorities to review the planned usage of alternative fuels that are currently under consideration as replacements for more traditional fuel sources.

Table 1. Specification of the Diesel Engine Used in This Study model

Cummins ISBe220 31

cylinders capacity (L) bore × stroke (mm) maximum power (kW/rpm) maximum torque (N m/rpm) compression ratio aspiration fuel injection emissions certification

6 in-line 5.9 102 × 120 162/2500 820/1500 17.3 turbocharged and after cooled common rail Euro III

rpm and 50% load. Tallow was tested as a B20 blend (20% of biodiesel and 80% of baseline fuel) at the same conditions as soy biodiesel. All of the palm oil samples were tested as B20 blends but at different loads to the previous tests. The sample labeled palm oil 3 was tested at the same engine speed of 1500 rpm but at 25% load, while samples labeled palm oil 1 and 2 were tested at high-speed idle (1200 rpm) at two different dilution conditions. This almost random choice of loads and fuels was performed to test if the linear correlation between DPM organic content and the OP, as measured by the concentration of ROS, could still stand. 2.3. Particulate Emissions Measurement Methodology. The schematic of the experimental setup is shown in Figure 1. The raw exhaust was subjected to a two-stage dilution process, involving a partial flow dilution tunnel, followed by a Dekati ejector diluter. To achieve turbulent mixing conditions in the partial flow dilution tunnel, the flow rates were kept above 400 lpm, which corresponded to a Reynolds number above 4000 in the partial flow dilution tunnel. This has resulted in first-stage dilution ratios of 15 or above. The second-stage dilution, in the ejector dilutor, was kept constant at 9. Particle number size distributions were measured following the method described by Surawski et al.4 A Dusttrak model TSI 8530 was used for PM2.5 measurement. Dusttrak mass measurements were converted to gravimetric measurements according to a procedure described in ref 17. OP was measured using a profluorescent nitroxide molecular probe 9-(1,1,3,3-tetramethylisoindolin-2-yloxyl-5-ethynyl)-10(phenylethynyl)anthracene (BPEAnit). This probe provides an entirely novel, rapid, and non-cell-based assay for measuring OP.13 BPEAnit has been applied in situ to assess the OP of cigarette smoke, DPM, wood smoke, and traffic emission within a tunnel.12,18−20 Samples were collected by bubbling aerosol after the first-stage dilution through an impinger containing 20 mL of 4 μM BPEAnit solution [using analytical reagent (AR)grade dimethyl sulfoxide (DMSO) as a solvent], followed by fluorescence measurements with a spectrofluorometer (Ocean Optics).21 Because the efficiency of particle collection in the impingers is size-dependent, all of the reported data were corrected for the losses in impingers, according to the loss function provided in ref 22. The amount of BPEAnit reacting with the combustion aerosol was calculated from a standard curve obtained by plotting known concentrations of the methanesulfonamide adduct of BPEAnit (fluorescent)20 against the fluorescence intensity at 485 nm. For each setting and particulate source, two samples were taken. The first sample was produced by the exposure of BPEAnit solution to the particle-free gas phase. This was

2. EXPERIMENTAL SECTION 2.1. Engine. Particulate emission measurements were performed on a Euro III Cummins diesel engine coupled to a water brake dynamometer. The engine was equipped with an ethanol fumigation system. Details on the engine specifications can be seen in Table 1. 2.2. Fuels and Testing Conditions. Commercially available diesel was used as the baseline fuel, along with three ethanol fumigation experiments, where 10, 20, and 30% of the total fuel by energy was supplied by ethanol, with the remainder of fuel energy provided by diesel (labeled as E10, E20, and E30, respectively). All of the ethanol tests were conducted at 1500 rpm and 50% load. In addition, three different biodiesel fuels (soy, palm oil, and tallow) were tested. Soy was tested as a B100 blend (pure soy biodiesel with no baseline fuel) at 1500 B

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Figure 1. Schematic representation of the experimental configuration used in this study.

Figure 2. Particle number size distributions (corrected for dilution) for all fuel types tested (includes ethanol-fumigated diesel with different percentages of ethanol used (10, 20, and 30%) and three biodiesels from different feedstocks).

operation is available in the study by Drenwick et al. 23 The aerosol mass spectrometer (AMS) provides size-resolved mass concentrations of the non-refractory components of submicrometer particles. The term “non-refractory” is assigned to those components that are vaporized at 600 °C under vacuum conditions (organics, sulfate, nitrate, ammonium, and chloride). The flow calibration, the servo position check, and the size calibration were performed prior to each measurement campaign; the ionization efficiency (IE) calibration was conducted every 4−5 days, and the m/z, the baseline, and single-ion calibrations were performed every day. The averaging time was 1 min, and the AMS alternated between mass spectrometry (MS) and particle time of flight (PToF) modes

achieved by placing a high-efficiency particulate air (HEPA) filter between an impinger and the aerosol source. The second test sample was collected following exposure of the BPEAnit solution to the combined particle and the gas phase. On the basis of the difference in fluorescence signals at 485 nm between the test and a HEPA-filtered control sample, the amount of particle-associated ROS emitted for each test sample was calculated and normalized to the particle mass to give ROS concentrations (nmol/mg), hereon referred to as OP. OP was expressed in nmoles per milligram of particles. For this purpose, ROS results were normalized to the PM mass. An Aerodyne c-ToF AMS was used to measure particle composition. A detailed description of the c-ToF AMS and its C

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Figure 3. PM2.5 mass emissions and emissions of organic matter from an engine run at intermediate speed (1500 rpm) using various fuels.

every 30 s. For each mode, sampling lasted 10−15 min. Also, for each mode, a 10−15 min background measurement, carried out by placing a HEPA filter in front of the AMS sampling inlet, was performed. The results of background measurements were used to apply gas-phase corrections in the AMS fragmentation table.24 AMS data analysis was performed in Igor Pro 6.2 (Wavemetrics, Lake Oswego, OR) using the ToF-AMS Analysis Toolkit Squirrel, version 1.51H. A collection efficiency (CE) of 1 was assumed to estimate the non-refractory particle mass concentration.

emission factors decrease with an increased ethanol percentage.27 Contrary, the tested biodiesel stocks PM2.5 reductions are evident but dependent upon the fuel stock. In general, this trend matches literature reports.27 Because the existing standards and guidelines aimed at reducing PM mass do not consider which component is the most influential in negative health effects, this graph also illustrates the organic mass concentrations detected by the AMS (in gray), because this factor can be a more relevant metric for observed effects. The organic content in the PM increases with an increasing ethanol substitution, along with the reduction in PM2.5 mass. For the differing biodiesel feedstocks, the results are conflicting. Soy biodiesel gives an organic content similar to diesel, and tallow produces a reduction in the PM2.5 with similar organic mass concentrations, while combustion of palm-oil-sourced biodiesel produces the highest organic compound percentage with respect to the emitted mass. 3.3. Correlation between OP and Particle Volatility. It has been well-established through a number of studies that the measurement of the OP of airborne particles can be a good estimate of their capacity to induce adverse health effects.28 The presented OPs measured here are normalized to the concentration of the methanesulfonamide adduct of BPEAnit formed during the reaction between the radical quencher (BPEAnit) and biodiesel exhaust in DMSO. In the study by Stevanovic et al.,20 the main product responsible for the observed increase in fluorescence upon exposure to biodiesel exhaust has been isolated and detected. Notably, the same product was generated upon exposure of BPEAnit to high concentrations of hydrogen peroxide and peroxyl radicals. The most challenging task that remains is to identify the specific PM fraction that represents the most important causal pollutant component. This is a very complex issue because of composition continuum as well as the dynamic continuum between carbon particulates and both volatile organic compounds (VOCs) and semi-volatile organic compounds (SVOCs). More than 200 species have been identified, many of

3. RESULTS AND DISCUSSION 3.1. Particle Number Size Distribution. Size distribution data of some of the fuels tested are presented in Figure 2. With an increase in ethanol content, the total particle concentration decreased in a monotonic fashion. The highest reduction in the particle concentration occurred with E30 fuel, as observed previously with an older type engine.16 The biodiesel size distribution also shows feedstock and blend dependency as observed previously.4 For soy, tested as B100, a reduction in the biodiesel particle number occurs along with a shift to a smaller particle diameter [18.7% reduction in count median diameter (CMD)]. The B20 blend of palm oil also exhibits a reduction in the accumulation mode particle concentration and size at 1500 rpm (19.25% reduction in CMD). At idle (1200 rpm), a nucleation mode can also be observed for palm oil. This is the only case where a nucleation mode was observed, but it was not likely due to the fuel tested but the fact that the nucleation mode is commonly observed at idle.25,26 Tallow, which was also tested as a B20 blend, also exhibits a reduction in the number concentration of particles but only for particles smaller than 50 nm. 3.2. Total PM2.5 Mass Emissions and Its Organic Fraction. The results for the PM2.5 mass emissions as well as the organic fraction emissions, as measured by the AMS, for the above fuels are shown in Figure 3. Figure 3 shows that PM2.5 D

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Figure 4. Dependence between OP and organic content of particles. A line of linear fit, with R2 = 0.163884, is presented as well.

method m/z was used.29 The types of OAs investigated were hydrocarbon-like OA (HOA) and oxygenated OA (OOA) species. Two markers were used, f44 and f57, which reflect the contribution of the particular organic fragment at each molecular ion (m/z 44 or 57) to the total organic mass. The HOA mass profile is, as expected, dominated by alkyl fragments. As a representative from the CnH2n + 1 series, we chose m/z 57 (C4H9+) as a HOA tracer peak. OOA is represented by its prominent m/z 44 (CO2+) peak. HOA levels are usually in accordance with primary tracers, such as EC, NOx, and CO.30,31 OOAs are however mainly considered as surrogates for secondary organic aerosols (SOAs) and are in a good correlation with sulfates, O3 levels, and/or nitrates. OOAs can thus represent heterogeneous products of primary organic aerosols (POAs) and/or oxidation products of SVOCs that volatilized from POAs and oxidized in a gas phase.30,32 To illustrate the difference between the composition of the particle-bound organic fractions, the average mass spectra for low OP (diesel), medium OP (30% ethanol), and high OP (soy biodiesel) samples are shown in Figure 5. In addition, the f44 and f57 values for all samples are presented in Table 2. It can be seen from Table 2 the E30 and diesel f57 values of 0.078 and 0.081, respectively, which are typical for OAs dominated by hydrocarbons. Despite having a large organic fraction, the mass spectrum for E30 showed that the most of the organic content was not oxygenated, which can explain the significantly lower production of ROS observed in comparison to the soy biodiesel. Figure 6 demonstrates the importance of the OOAs for the OP. OP was expressed as a concentration of ROS per unit mass of PM, while the value of the f44 marker presented the OOA fraction. A strong linear correlation (R2 = 0.97) between the ROS concentration and OOAs can be clearly observed, which implies that the OP is consistent with the level of oxygenated organics.

which can undergo chemical modifications and alter their reactivity and toxicity. It was previously indicated in the literature4,15 that SVOCs are components of airborne PM that contribute the most to the measured OP of particles in question. These components undergo partitioning between particulates and the gaseous phase. Upon heating in a thermodenuder, however, they are transferred predominantly to the gas phase. Using this approach, it is possible to attain information on the volatility of the particles. In our previous studies,4,12 the relationship between the organic fraction of particles and the OP measured by the BPEAnit assay was clearly established. This correlation highlights the importance of organic species in particle-induced toxicity and reinforces the need for further investigations to explore this relationship. However, the measurements performed in this study and presented in Figure 4 show that the overall organic content of particles did not correlate strongly with their measured OPs. It has been previously observed16 and further confirmed in this study that the organic content increases with an increasing amount of ethanol present in the fuel. Also, the same study showed that increased levels of ethanol led to the higher OP. However, when a different fuel type, such as soy or palm oil, is introduced, the direct correlation disappears. The same noncorrelation was observed for most of the fuels tested. This indicates that the ROS content, although dependent upon the particle organic fraction, does not always exhibit a direct correlation. Furthermore, using the same profluorescent probe technology, Miljevic et al.12 have observed a similar effect with wood combustion PM, where a good linear correlation was observed only for the same combustion conditions. This was influenced by the differing reactivities of the organic species under the different conditions. 3.4. Influence of Oxygenated Organic Aerosols (OOAs) Content on the OP of DPM. To gain further insight into the chemical composition of organic aerosol (OA), a tracer E

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Figure 5. Average mass spectra for neat diesel, E30, and B100 soybean biodiesel at 50% load.

The engine fueled with 100% soy biodiesel contained substantially higher concentrations of ROS per milligram of PM than any other fuel, possibly related to the higher level of oxygenated organic compounds. The f44 was 0.104 for soy, 0.033 for palm oil, and 0.029 for E20. E30 showed higher levels

of OOAs (0.046), which resulted in higher ROS concentrations and decreases in the HOAs. For all of the other fuels, f44 was below 0.05. In addition, the f57 tracer for HOAs was 0.08 for neat diesel, and it gives similar values for fuels that generate similar OP (0.07−0.08). The most reactive particles were F

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

ROS production, although proportional to the total volatile organic volume percentage, shows a much stronger correlation to the oxygenated organic fraction. This highlights the importance of the surface chemistry of particles for assessing the health impacts because these surfaces are likely to be rich with oxygenated species and structures. These results also shed new light on previously unrecognized aspects of the nature of particulate emissions that should be taken into account when establishing relevant metrics for health implications of various emissions. Although all of the measurements were performed on primary aerosol particles, these findings should be even more relevant for secondary aerosol particles, because the OOA component of these particles is expected to be even larger.

Table 2. f44 and f57 Tracer Values for All Tested Fuels fuel stock

f44

f57

siesel E20 E30 soybean tallow palm oil 3 palm oil 1 palm oil 2

0.015 0.029 0.046 0.104 0.048 0.033 0.018 0.026

0.081 0.075 0.078 0.050 0.055 0.073 0.075 0.074

emitted upon combustion of soybean biodiesel, which gave a HOA fraction of 0.049, which is much lower than that compared to the other fuels tested. Because the measured OP value for soy is very high compared to all of the other values, this correlation was investigated in the absence of this point. Excluding soy from the correlation would give a coefficient of multiple determination (R2) with the value of 0.732. This is still a significantly higher value of R2 compared to the value obtained from the correlation between OP and organic content (R2 = 0.16). However, the data that originated from the combustion of tallow (animal fat) biodiesel was not included in this correlation, because the result does not follow the trend shown by the other fuels. The observed values for f44 were relatively high, and values for f57 were relatively low. To the contrary of what would be expected for this case, a low OP was observed. This result needs to be further investigated. It may arise from some bias in the measurements and require repetitive and more detailed analysis, or it may reflect the range of components present in tallow and the differences in their chemical structures compared to the plant-derived materials. From these results, it can be concluded however that, in general, the OP of the PM, as measured through the levels of



AUTHOR INFORMATION

Corresponding Author

*Telephone: +617-3138-1129. Fax: +617-3138-9079. E-mail: z. [email protected]. Present Address §

N. C. Surawski: CSIRO Ecosystem Sciences, Clunies Ross Street, Acton, Australian Capital Territory 2601, Australia. Notes

The authors declare no competing financial interest.



REFERENCES

(1) Xue, J.; Grift, T. E.; Hansen, A. C. Effect of biodiesel on engine performances and emissions. Renewable Sustainable Energy Rev. 2011, 15 (2), 1098−1116. (2) Lackey, L. G.; Paulson, S. E. Influence of feedstock: Air pollution and climate-related emissions from a diesel generator operating on soybean, canola, and yellow grease biodiesel. Energy Fuels 2011, 26 (1), 686−700. (3) Hedges, J. I.; et al. The biochemical and elemental compositions of marine plankton: A NMR perspective. Mar. Chem. 2002, 78 (1), 47−63.

Figure 6. Correlation between OP, measured as the ROS concentration, and f44 used as a marker for the content of the oxygenated organic fraction. A line of linear fit is also shown. A label for the fuel type is next to each data point. G

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

(4) Surawski, N. C.; et al. Physicochemical characterization of particulate emissions from a compression ignition engine: The influence of biodiesel feedstock. Environ. Sci. Technol. 2011, 45 (24), 10337−10343. (5) Tzamkiozis, T.; et al. Aerodynamic and mobility size distribution measurements to reveal biodiesel effects on diesel exhaust aerosol. Aerosol Sci. Technol. 2011, 45 (5), 587−595. (6) Chuepeng, S.; et al. Particulate matter size distribution in the exhaust gas of a modern diesel engine fuelled with a biodiesel blend. Biomass Bioenergy 2011, 35 (10), 4280−4289. (7) Ristovski, Z. D.; et al. Respiratory health effects of diesel particulate matter. Respirology 2012, 17 (2), 201−212. (8) Gaffney, J. S.; et al. Potential air quality effects of using ethanol− gasoline fuel blends: A field study in Albuquerque, New Mexico. Environ. Sci. Technol. 1997, 31 (11), 3053−3061. (9) Grosjean, D.; Grosjean, E.; Gertler, A. W. On-road emissions of carbonyls from light-duty and heavy-duty vehicles. Environ. Sci. Technol. 2000, 35 (1), 45−53. (10) He, C.; et al. Comparison of carbonyl compounds emissions from diesel engine fueled with biodiesel and diesel. Atmos. Environ. 2009, 43 (24), 3657−3661. (11) Lapuerta, M.; Armas, O.; Rodriguez-Fernandez, J. Effect of biodiesel fuels on diesel engine emissions. Prog. Energy Combust. Sci. 2008, 34 (2), 198−223. (12) Miljevic, B.; et al. Oxidative potential of logwood and pellet burning particles assessed by a novel profluorescent nitroxide probe. Environ. Sci. Technol. 2010, 44 (17), 6601−6607. (13) Fairfull, K. E.; et al. A novel profluorescent dinitroxide for imaging polypropilene degradation. Macromolecules 2008, 41 (5), 3. (14) Biswas, S.; et al. Particle volatility in the vicinity of a freeway with heavy-duty diesel traffic. Atmos. Environ. 2007, 41 (16), 3479− 3493. (15) Biswas, S.; et al. Oxidative potential of semi-volatile and non volatile particulate matter (PM) from heavy-duty vehicles retrofitted with emission control technologies. Environ. Sci. Technol. 2009, 43 (10), 3905−3912. (16) Surawski, N. C.; et al. Particle emissions, volatility, and toxicity from an ethanol fumigated compression ignition engine. Environ. Sci. Technol. 2009, 44 (1), 229−235. (17) Jamriska, M.; et al. Diesel bus emissions measured in a tunnel study. Environ. Sci. Technol. 2004, 38 (24), 6701−6709. (18) Crilley, L. R.; et al. Concentration and oxidative potential of onroad particle emissions and their relationship with traffic composition: Relevance to exposure assessment. Atmos. Environ. 2012, 59, 533−539. (19) Miljevic, B.; et al. The application of profluorescent nitroxides to detect reactive oxygen species derived from combustion-generated particulate matter: Cigarette smokeA case study. Atmos. Environ. 2010, 44 (18), 2224−2230. (20) Stevanovic, S.; et al. The use of a nitroxide probe in DMSO to capture free radicals in particulate pollution. Eur. J. Org. Chem. 2012, 30, 5908−5912. (21) Stevanovic, S.; et al. Application of profluorescent nitroxides for measurements of oxidative capacity of combustion generated particles. Chem. Ind. Chem. Eng. Q. 2012, 18 (4), 653−659. (22) Miljevic, B.; et al. On the efficiency of impingers with fritted nozzle tip for collection of ultrafine particles. Atmos. Environ. 2009, 43 (6), 1372−1376. (23) Drewnick, F.; et al. A new time-of-flight aerosol mass spectrometer (TOF-AMS)Instrument description and first field deployment. Aerosol Sci. Technol. 2005, 39 (7), 637−658. (24) Allan, J. D.; et al. A generalised method for the extraction of chemically resolved mass spectra from Aerodyne aerosol mass spectrometer data. J. Aerosol Sci. 2004, 35 (7), 909−922. (25) Kittelson, D. B.; Watts, W. F.; Johnson, J. P. On-road and laboratory evaluation of combustion aerosolsPart 1: Summary of diesel engine results. J. Aerosol Sci. 2006, 37 (8), 913−930. (26) Ristovski, Z. D.; et al. Influence of diesel fuel sulfur on nanoparticle emissions from city buses. Environ. Sci. Technol. 2006, 40 (4), 1314−1320.

(27) Abu-Qudais, M.; Haddad, O.; Qudaisat, M. The effect of alcohol fumigation on diesel engine performance and emissions. Energy Convers. Manage. 2000, 41 (4), 389−399. (28) Ayres, J. G.; et al. Evaluating the toxicity of airborne particulate matter and nanoparticles by measuring oxidative stress potentialA workshop report and consensus statement. Inhalation Toxicol. 2008, 20 (1), 75−99. (29) Ng, N. L.; et al. Real-time methods for estimating organic component mass concentrations from aerosol mass spectrometer data. Environ. Sci. Technol. 2010, 45 (3), 910−916. (30) Zhang, Q.; et al. Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-influenced Northern Hemisphere midlatitudes. Geophys. Res. Lett. 2007, 34 (13), L13801. (31) Ulbrich, I. M.; Canagaratna, M. R.; Zhang, Q.; Worsnop, D. R.; Jimenez, J. L. Interpretation of organic components from positive matrix factorization of aerosol mass spectrometric data. Atmos. Chem. Phys. Discuss. 2009, 9, 2891−2918. (32) Herndon, S. C.; et al. Correlation of secondary organic aerosol with odd oxygen in Mexico City. Geophys. Res. Lett. 2008, 35 (15), L15804.

H

dx.doi.org/10.1021/es4007433 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX