Gas-Particle Partitioning of Vehicle Emitted ... - ACS Publications

Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

Gas-particle partitioning of vehicle emitted primary organic aerosol measured in a traffic tunnel Xiang Li, Timothy R Dallmann, Andrew A. May, Daniel S Tkacik, Andrew T. Lambe, John T. Jayne, Philip L Croteau, and Albert A Presto Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b01666 • Publication Date (Web): 20 Oct 2016 Downloaded from http://pubs.acs.org on October 20, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37

Environmental Science & Technology

2

Gas-particle partitioning of vehicle emitted primary organic aerosol measured in a traffic tunnel

3 4

Xiang Li1,2, Timothy R. Dallmann1,2,5 , Andrew A. May1,2,6, Daniel S. Tkacik1,3, Andrew T. Lambe4, John T. Jayne4, Philip L. Croteau4, and Albert A. Presto*,1,2

1

5 6

1

7 8

2

9 10

Center for Atmospheric Particle Studies, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, United States Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, United States 3

Department of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, United States 4

Aerodyne Research, Inc., Billerica, Massachusetts 01821, United States

11 12 13 14 15

5

now at: The International Council on Clean Transportation, Washington, DC 20005, United States 6

now at: Department of Civil, Environmental and Geodetic Engineering, The Ohio State University, Columbus, OH 43210, United States

*Correspondence to: [email protected]

16 17 18

Abstract

19

We measured the gas-particle partitioning of vehicle emitted primary organic aerosol

20

(POA) in a traffic tunnel with three independent methods: artifact corrected bare-quartz

21

filters, thermodenuder (TD) measurements, and thermal-desorption gas-

22

chromatography mass-spectrometry (TD-GC-MS). Results from all methods

23

consistently show that vehicle emitted POA measured in the traffic tunnel is semivolatile

24

under a wide range of fleet compositions and ambient conditions. We compared the

25

gas-particle partitioning of POA measured in both tunnel and dynamometer studies and

26

found that volatility distributions measured in the traffic tunnel are similar to volatility

27

distributions measured in the dynamometer studies, and predict similar gas-particle

28

partitioning in the TD. These results suggest that the POA volatility distribution

29

measured in the dynamometer studies can be applied to describe gas-particle

ACS Paragon Plus Environment

1


Page 2 of 37

30

partitioning of ambient POA emissions. The POA volatility distribution measured in the

31

tunnel does not have significant diurnal or seasonal variations, which indicate that a

32

single volatility distribution is adequate to describe the gas-particle partitioning of vehicle

33

emitted POA in the urban environment.

34

35

Introduction

36

Gasoline and diesel vehicles are a significant source of ambient fine particulate matter

37

(PM).1–3 Vehicle emitted PM mainly consists of elemental carbon (EC) and primary

38

organic aerosol (POA).4–7 Many studies have been conducted to measure POA

39

emission factors from on-road gasoline and diesel vehicles.8–11 Traditionally, chemical

40

transport models treated POA as non-volatile, which contributed to a discrepancy

41

between modeled and measured PM mass.12,13 We know now that a substantial fraction

42

of POA from combustion exhaust is semivolatile14–23 and actively partitions between the

43

vapor and condensed phases. Accurate accounting of POA mass therefore requires

44

knowledge of both the emission rate of condensable material and the volatility of the

45

emissions.

46

Gas-particle partitioning of POA is a sorption process. Organics in the gas phase

47

partition into the particle phase either by absorbing into the organic condensed phase or

48

adsorbing onto nonvolatile, often EC, cores. Whether absorption or adsorption is the

49

dominant mechanism depends on the relative abundance of organic carbon (OC) and

50

EC.24 Absorption dominates under most atmospheric conditions.25,26 The fraction of

51

organic mass in the particle phase can be calculated as:27,28


2

Page 3 of 37

52


Xp =∑ i

 C i* (T )    f i 1 +  C OA  

−1

(1)

53

Where X p is the fraction of organic mass in the particle phase; f i is the fraction of

54

species i among total organics (particle + gas phase); C i* (T ) is the effective saturation

55

concentration of species i at temperature T; and COA is the mass concentration of

56

organic aerosol (OA). The emission factor of the POA can be determined by:29 EFOA = X p ⋅ EFtot

57

( 2)

58

Where EFtot is the emission factor of total organics (particle + gas phase); and EFOA is

59

the emission factor of the POA.

60

Determining X p requires knowing f i and C i* (T ) of all species in the POA. However, no

61

measurement techniques can fully speciate POA.6,30 Therefore equation (1) is often

62

applied semi empirically using a set of surrogate compounds.14,21–23,28,30,31 This set of

63

surrogate compounds can be presented with the one-dimensional volatility basis set,28

64

which spreads the semivolatile organics over a logarithmically spaced set of C* bins.

65

The set of fi is the volatility distribution. Several studies have used the volatility basis set

66

to simulate semivolatile POA in chemical transport models (CTMs), 32–34 and improve

67

model-measurement agreement.14

68

Numerous studies have investigated gas-particle partitioning of vehicle emitted POA

69

through both field measurements and laboratory studies.7,10,22,23,30,35,36 Some earlier

70

studies7,10,36 indicate the presence of POA partitioning, but do not provide a volatility

71

distribution. Recently May et. al22,23 tested 51 gasoline vehicles and 5 diesel vehicles on


3


Page 4 of 37

72

chassis dynamometers and derived POA volatility distributions. Worton et al.30

73

conducted tunnel measurements and derived POA volatility distributions, and concluded

74

that the vehicle emitted POA are dominated by lubricating oil. Both May et. al22,23 and

75

Worton et al.30 concluded that volatility distributions of gasoline and diesel vehicles are

76

similar, suggesting that a single volatility distribution can be used to describe the POA

77

gas-particle partitioning of all vehicles.

78

Kuwayama et al.35 also observed POA gas-particle partitioning. They used a

79

combination of two volatility distributions, lower volatility fuel-combustion POA and

80

motor oil POA, to describe POA gas-particle partitioning measured in their

81

dynamometer studies. Kuwayama et al35 used a large range in the fraction of motor oil

82

POA to total POA (24-86%) to describe the emissions from individual vehicles in their

83

test fleet. Their results suggest that POA gas-particle partitioning cannot be predicted by

84

a single fleetwide volatility distribution. However, since Kuwayama et al.35 tested only

85

eight cleaner gasoline vehicles, data from a larger fleet may be required to validate this

86

claim.

87

The measurements presented by both May et al22,23 and Kuwayama et al35 represent

88

relatively small fleets tested under controlled conditions. Nonetheless, POA emission

89

factors and volatility distributions from May et al22,23 and other small test fleets have

90

been incorporated into chemical transport models.34,37 To our knowledge, no studies

91

have been published to demonstrate whether these laboratory-derived POA volatility

92

distributions can be applied to describe POA gas-particle partitioning of a larger mixed

93

vehicle fleet operating under real-world driving conditions.


4

Page 5 of 37


94

In this work we measured the gas-particle partitioning of POA in a traffic tunnel with

95

three independent methods: 1) a thermodenuder (TD), 2) quartz filter sets analyzed by

96

thermal-optical OCEC and 3) analysis of quartz filters with thermal desorption gas

97

chromatography mass spectrometry (TD-GC-MS) to determine volatility distributions.

98

This manuscript is intended to show that POA emitted from a large and real-world

99

driving fleet is semivolatile under a wide range of ambient conditions, the POA volatility

100

distribution derived from the dynamometer studies can be applied to explain gas-particle

101

partitioning of ambient POA, and that gas-particle partitioning of the POA measured in

102

the traffic tunnel does not have large diurnal or seasonal variations.

103

Methods

104

Tunnel measurement station

105

We conducted measurements in the Fort Pitt Tunnel, which carries Interstate 376

106

southwest of downtown Pittsburgh, PA. The tunnel is 1.1-km long with a 2.5% upward

107

grade to the west. All measurements were conducted in the westbound side. A sketch

108

of the tunnel measurement station is shown in Figure S1.

109

Average vehicle speed, vehicle volume and number fractions of heavy-duty diesel

110

vehicles (HDDV) are presented in Figure S2. Fleet age is also discussed in the

111

supplemental information. The traffic volume in the tunnel exhibits two peaks on

112

weekdays corresponding to the morning (6:00-8:00) and afternoon (15:00-18:00) rush

113

hours. On weekends only the afternoon rush hour peak was evident. The number

114

fraction of HDDV peaks overnight at about 10% of the total fleet. During daytime

115

HDDVs make up less than 6% of the total fleet.


5


Page 6 of 37

116

We quantify the presence of diesel vehicles with the diesel fuel fraction ( % fuelD ),the

117

percentage of fuel consumed by HDDV during a certain measurement time period. The

118

description of the calculation of % fuelD is in the Supplemental Information (Equation S1).

119

Measurements were conducted continuously for 9 months. The results presented here

120

focus on two time periods. The spring measurement was conducted in May 2013

121

together with the work presented in Tkacik et al.38 During this period an Aerosol

122

Chemical Speciation Monitor (ACSM, Aerodyne Research Inc.)39 measured non-

123

refractory particle phase organic aerosol mass. The time resolution of ACSM is typically

124

15-30 min but was operated here at 1-min resolution due to high organic aerosol

125

concentrations in the tunnel. A thermodenuder (TD)40 was also deployed to study the

126

POA gas-particle partitioning under different temperatures. We continuously measured

127

CO2 (Model LI820, Li-Cor Biosciences), CO (Model T300, Teledyne-API), and NOx

128

(Model 200EU, Teledyne-API) concentrations. Particulate OC and EC were measured

129

with a semi-continuous OC/EC analyzer (Model 4F, Sunset Laboratory Inc.).

130

Winter measurements were conducted during January and February 2014. In addition

131

to measuring CO2, CO, NOx, OC, and EC, we collected quartz filter sets and analyzed

132

them offline with a Sunset OCEC analyzer (Model 3) and the TD-GC-MS, described

133

below. All the NOx and CO data used in this manuscript are subtracted for ambient

134

background to show only the enhancement caused by the vehicles inside the tunnel.

135

Thermodenuder (TD)

136

The TD used in this study is the same as the one used in May et al.22,23 and a detailed

137

description of the instrument can be found there. The TD uses high temperatures to


6

Page 7 of 37


138

drive the gas-particle partitioning of OA. It consists of a stainless steel heating section

139

(2.7 cm ID x 65 cm L) followed by an activated carbon filled stripper/denuder. The

140

flowrate inside the TD was 4.2 SLPM and the centerline residence time in the heating

141

section was 5.3 seconds at 298 K. Tunnel air was alternately sampled through the TD

142

and an ambient temperature bypass line.

143

We present the TD data using the mass fraction remaining (MFR):

MFR =

144

CTD Cbypass

(3)

145

CTD is the OA concentration measured downstream of the TD by the ACSM. C bypass is

146

the OA concentration measured through the unheated bypass line. The MFR represents

147

the fraction of OA that survived the high temperature inside the TD. In this work the TD

148

was operated at a series of fixed temperatures (25, 40, 60, 100, 150 °C; Figure S3). We

149

define one TD scan as the time period when the TD temperature increased from 25 ⁰C

150

to 150 ⁰C.

151

As discussed in Ng et al.39 the collection efficiency (CE) of the ACSM is similar with the

152

CE of the Aerosol Mass Spectrometer (AMS). Following previous AMS

153

measurements10,22 we assume unity CE for OA measured by the ACSM. Particle

154

number loss inside the TD is discussed in the Supplemental Information (Figure S4).

155

OA partitioning does not necessarily reach thermodynamic equilibrium in

156

thermodenuders with short residence time, such as the one used here.41,42 We used the

157

mass transfer model of Riipinen et al.41 to compare our TD measurements to predictions

158

of OA evaporation using volatility distributions derived in previous dynamometer and


7


Page 8 of 37

159

tunnel studies. The TD model does not explicitly include other processes such as

160

nucleation, chemical reaction, or condensation. Inputs to the model include the TD

161

dimensions and temperature, the volatility distribution of POA, the OA concentrations

162

(COA) and particle mass-median diameter (dp), and the mass accommodation coefficient

163

(α). We used an accommodation coefficient of 1.43 A full description of parameters used

164

in the model is summarized in Table S1 and Table S2.

165

Quartz filter sets

166

During the winter measurement we simultaneously sampled tunnel air onto a bare-

167

quartz filter (bare-Q) and a quartz-behind-Teflon filter (QBT). Quartz filters were 47-mm

168

Tissuquartz (2500QAT-UP, Pall Corp.). Teflon filters were 47-mm Teflon membranes

169

(Pall Corp). The majority (15 out of 18) of the quartz filter sets were collected during

170

weekdays; the remainder were collected on weekends. Filters were collected either

171

during midday (12:00 – 14:00) or in the afternoon rush hour (15:00 – 18:00). During

172

these time periods conditions in the tunnel crossed a wide range: the ambient

173

temperature in the tunnel (same as the filter sampling temperature) ranged from -2 to

174

6 °C; and the % fuelD ranged from 2% to 9%.

175

Sample flow rates were 46 SLPM. Two sharp cut PM2.5 cyclones were placed upstream

176

of filters. Fourteen out of 18 filter sets were sampled for 45 min and another 4 were

177

sampled for 90 min. Before sampling all quartz filters were baked at 550 °C overnight to

178

remove all residual organics. All quartz filter samples were kept in a freezer at -18 °C

179

prior to analysis.


8

Page 9 of 37


180

We also collected 2 sets of handling blanks in the tunnel. For both bare-Q and QBT, the

181

handling blank represents about 4% of the OC collected onto the filter (Figure S5).

182

Therefore, we did not correct quartz filter OC concentrations for handling blanks.

183

Quartz filters were analyzed using a Sunset Laboratories OCEC Aerosol Analyzer

184

(Model 3) following the IMPROVE-A protocol.44 OC and EC are defined as the carbon

185

that is thermally desorbed up to 550 ⁰C in a helium atmosphere and 550 ⁰C – 800 ⁰C

186

in a helium-oxygen atmosphere, respectively; both are corrected for pyrolyzed OC.44

187

We assume the bare-Q filter captures both particle phase organics and absorbed

188

organic vapors, which is considered the positive artifact. This positive artifact is

189

quantified by the QBT filter, which is assumed to only absorb gas phase organics.45,46

190

Therefore, the particle phase OC concentration can be derived from the artifact-

191

corrected bare-Q (bare-Q minus QBT). The particle phase fraction of organics (Xp) can

192

be determined by:

Xp =

193

bare - Q − QBT bare - Q

(4)

194

In this work we assume an organic-mass-to-organic-carbon ratio of 1.2 to convert OC

195

into OA. 47

196

Quartz filters were also analyzed by the TD-GC-MS. It has a thermal extraction and

197

injection system (Gerstel, Inc) followed with a gas chromatograph / mass spectrometer

198

(GC/MS, Agilent 6891 GC / 5975 MS). OC collected onto the quartz filter was thermally

199

desorbed inside the Gerstel desorption unit (TDS3); then it is concentrated at -120 °C

200

by the Gerstel cooled injection system (CIS4); and finally the concentrated sample was

201

thermally injected into GC/MS for anaylsis.


9


Page 10 of 37

202

Volatility distributions were determined from TD-GC-MS analysis of 18 bare-Q filters

203

following the method of Presto et al.21 The mass fragment (m/z) 57 is used to derive the

204

POA volatility distribution. The relative abundance of m/z 57 signal in the total organics

205

was not strongly affected by the % fuelD and the OC concentrations in the tunnel (Figure

206

S6). Therefore, the POA volatility distribution determined from m/z 57 is not strongly

207

influenced by the dilution of OA in the tunnel. The method uses a set of surrogate

208

compounds (C15-C40 n-alkanes) to develop a relationship between C* (10-2 – 106 µg/m3)

209

and the GC retention time. The set of fi are calculated using the fraction of the total m/z

210

57 signal in each of a series of logarithmically spaced C* bins. The m/z 57 signal in

211

each volatility bin was corrected with the recovery of a set of deuterated standards (C16-

212

C36 n-alkanes).

213 214

Results and discussion

215

Quartz filter measurements

216

OC/EC analysis of quartz filter samples shows that there is substantial organic mass

217

present in both the condensed and vapor phases. Organic concentrations from all

218

quartz filter samples (CQ) are shown as the total bar height in Figure 1(a). CQ measured

219

in the tunnel ranges from 8 to 38 µg/m3.

220

CQ shown in Figure 1(a) are not corrected for ambient background concentrations. Thus

221

while we expect the filter samples to be dominated by fresh emissions, as shown by the

222

ACSM data in Figure 2, there are likely contributions from background OA as well. On

223

weekdays the CQ were roughly two times higher than the CQ on the weekends. CQ


10

Page 11 of 37


224

measured at midday (12:00 -14:00) and during the afternoon rush hour (15:00 – 18:00)

225

do not exhibit a large difference on weekdays.

226

The positive artifact (organic vapors) measured by QBT filters are shown as open areas

227

of the bars in Figure 1(a). Artifact corrected particle phase OA concentration (bare-Q

228

minus QBT) are shown as filled areas of bars. The particle phase fractions of organics

229

(Xp) are shown in the lower panel of Figure 1(a). Xp in the tunnel ranged from 40 – 60%

230

and on average about half of the organics collected onto the bare-Q filter are organic

231

vapors.

232

Figure 1(b) presents concentrations of OC (µg/m3) evolved in different temperature

233

stages of the OCEC analysis. Using the IMPROVE-A OC analysis protocol, the

234

desorption temperature of the sample in a Helium environment gradually increases from

235

140 °C to 560 °C. OC1 – OC4 represent the OC evolved from the lowest to the highest

236

temperature stages. Ma et al48 recently showed that OC1-OC4 can be linked to volatility,

237

with higher volatility organics evaporating at lower temperature.

238

Results for both bare-Q (vapor plus particle) and QBT (vapors only) filters in Figure 1(b)

239

show organic mass spread across all OC temperature stages. Organic material

240

captured by the QBT filters (mostly vapor) is clearly more volatile than organics

241

captured on the bare-Q filters (vapor plus particle). The majority (80%) of the vapor

242

phase organics collected onto QBT filters are more volatile OC1 and OC2, whereas 40%

243

of the artifact-corrected particle phase organics (bare-Q minus QBT) consists of less

244

volatile OC3 and OC4.


11


Page 12 of 37

245

About 10% and 50% of the OC1 and OC2 mass on the bare-Q filters, respectively,

246

exists in the particle phase. Even for the least volatile OC4 there is still a small fraction

247

(10-20%) present in the vapor phase. The OC/EC analysis indicates that organic mass

248

evolving across the range of OC/EC temperature stages readily partitions between the

249

vapor and condensed phases. Therefore the quartz filters suggest that POA sampled in

250

the tunnel is semivolatile.

251 252

Direct measurements of gas-particle partitioning: Thermodenuder data

253

The analysis of quartz filters presented in the previous section lacks chemical specificity,

254

and cannot directly separate fresh emissions from background organic aerosol that may

255

enter the tunnel. Our analysis assumes that OA in the tunnel is dominated by fresh POA,

256

but is ultimately indirect. In this section, we use ACSM data to more rigorously separate

257

fresh POA from aged background aerosol. The time series of OA concentration

258

measured in the bypass line by the ACSM is shown in Figure 2; the TD data are

259

excluded.

260

We separated the measured OA into two components, Hydrocarbon-like Organic

261

Aerosol (HOA) and Oxygenated Organic Aerosol (OOA), using the principle component

262

analysis of Zhang et al.49 This method uses the fraction of OA mass at m/z 57 (f57) to

263

apportion HOA and the mass fraction at m/z 44 (f44) to apportion OOA. HOA is

264

indicative of fresh emissions, such as POA,50 and OOA is representative of more aged,

265

regional OA.51 Since vehicles are the only emission source in the traffic tunnel, the HOA

266

measured in the tunnel is a good surrogate for vehicle-emitted POA. The POA mass


12

Page 13 of 37


267

spectrum measured in the tunnel is highly similar with the HOA mass spectrum

268

measured by the AMS 49 (Figure S8). HOA has a stronger correlation with background

269

corrected CO and NOx concentrations (Figure S9), which are indicators of gasoline and

270

diesel vehicle volume in the tunnel, than OOA.11

271

The trend of OA concentration generally follows the background corrected NOx

272

concentration.11 The average OA concentration was higher on weekdays compared with

273

weekends, and the weekend OA time series shows less variation than weekdays. The

274

OA time series showed high peaks around midday and during the afternoon rush hour

275

during weekdays. On weekends a single, smaller OA peak was observed during the

276

early morning hours. These weekday/weekend patterns were also observed over longer

277

time periods in the semi-continuous OC/EC measurements (Figure S7). During

278

weekday daylight hours HOA accounted for 70-90% of total OA mass. In both absolute

279

concentration and as a fraction of total OA mass, HOA peaked during the weekday

280

midday and afternoon rush hour periods (>80% OA mass). This high HOA fraction

281

indicates that the majority of the sampled OA on weekdays was POA from vehicle fresh

282

emissions rather than the OA from the background air.

283

The HOA fraction on weekends was lower than on weekdays. During the weekend HOA

284

made up 30-80% of the total OA, with higher contributions (60-80%) when OA was

285

elevated during morning hours on Saturday. During Sunday and Monday when the OA

286

concentration was low (~3-6 µg m-3) the HOA fraction was approximately 30-40%. In

287

order to assure that gas-particle partitioning results focus on vehicle emitted POA rather

288

than the OA from the background air, our analysis of TD data focuses only on the HOA

289

fraction.


13


Page 14 of 37

290

The effect of temperature on the gas-particle partitioning of HOA is shown in Figure 3.

291

All TD measurements are presented using box-whisker plots. The median HOA MFR is

292

close to unity when the TD is held near ambient temperature (25 °C). MFR values

293

greater than unity may exist due to temporal variability between samples collected

294

through the TD and the bypass line. When the TD was heated up to higher temperature

295

(40-60 °C) the median HOA MFR dropped to 70-80%, indicating that about 20-30% of

296

the POA evaporated with mild heating. About half of the HOA evaporated when the TD

297

temperature reached 80-90 °C. When the TD temperature reached highest stages (100-

298

150 °C) about 80-85% of the POA evaporated. The trend of median HOA MFRs clearly

299

show that the HOA MFR decreases as the TD temperature increases, indicating that the

300

POA is semivolatile and evaporates upon heating.

301

The change of the MFR as a function of TD temperature is not linear. The MFR drops

302

quickly for TD temperatures between 25 °C and 100 °C. However when the TD

303

temperature is higher than 100 °C the MFR levels off. From 100 °C to 150 °C the HOA

304

MFR changed from 0.25 to about 0.15. This reduction in the slope under conditions of

305

higher TD temperature and low MFR is commonly observed.17,18,22,23,52,53 Incomplete

306

evaporation at high TD temperatures (>100 °C) is often explained by a transition from a

307

purely absorptive environment to one where organics are adsorbed to soot cores.21–23

308

Adsorption incurs an additional enthalpy,27 the enthalpy of adsorption, in addition to

309

enthalpy of vaporization, and this additional enthalpy is often invoked to explain the

310

incomplete evaporation at high TD temperatures.

311

Adsorption is expected to be important for OC:EC ratio 110 °C. The POA in the 24%

429

oil case is predicted to evaporate completely at T = 130 °C, and therefore does not

430

describe the observed MFR at T = 150 °C. Agreement at the highest TD temperatures

431

could be improved if the combustion POA consisted of even lower volatility material

432

(e.g., C* < 10-3 µg/m3), but would still over predict MFR at nearly all lower temperatures.

433

This suggests that the vehicle emitted POA in the traffic tunnel is more likely dominated

434

by the motor oil POA rather than the low-volatile fuel-combustion POA.

435

We also tested the sensitivity of predicted MFR to the presence of additional high

436

volatility material by comparing the TD MFR predictions using the volatility distributions

437

from Zhao et al.56,57 Zhao et al expanded the volatility distributions of May et al.22,23 to

438

include IVOCs. The addition of IVOCs to the volatility distribution is expected to have

439

little impact on predicted partitioning, as IVOCs exist as vapors under all but the most

440

concentrated conditions. As shown in Figure S12, this is the case. The MFR predictions

441

based on Zhao et al.56,57 are either within the range or slightly lower than the predictions

442

of May et al. 22,23

443

Temporal variations in POA volatilty

444

The data presented here allow us to investigate variations in POA volatility over daily

445

and seasonal time scales. The data and modeling presented in Figure 3 suggest that

446

POA composition and volatility do not show a large seasonal variation. The quartz filters


20

Page 21 of 37


447

used to determine the volatility distributions were collected in winter at ambient

448

temperatures below 6 °C. The TD data were collected during the spring, when ambient

449

temperatures were approximately 20 °C. As shown in Figure S11 and Figure 3, there is

450

reasonable agreement between the TD data, the predicted MFR based on our filters

451

collected in the winter, and predicted MFR from filters collected by May et al22,23 in a

452

temperature-controlled dynamometer.

453

The composition of the vehicle fleet in the tunnel changes over the course of each day.

454

As shown in Figure 4, variation of the volatility distribution derived from the TD-GC-MS

455

is much less than the diurnal variation of % fuelD .This suggests that POA emissions

456

from gasoline and diesel vehicles have similar volatility, and may have similar

457

composition.

458

We can also use the TD to investigate sub-daily variations in POA partitioning and

459

volatility. In Figure 5 we separated all measured TD MFR in the tunnel into 4 different

460

time periods: high-traffic daytime periods and overnight on either weekdays or

461

weekends. The % fuelD ranged from 5% -20% and the % fuelD in the high-traffic time is

462

roughly 60% lower than overnight. Results show the same trend in MFR for each of the

463

four time periods. Quantitatively, the mean MFRs at each temperature agree within 20%,

464

and there is significant overlap (error bars indicate the standard deviation of MFR). We

465

compared time-temperature MFR pairs with the Mann-Whitney U-test (Table S5) and

466

found that the majority (25 out of 30) of sample pairs are not statistically different.

467

Exceptions occurred at 25 and 150 °C. The results in Figure 5 therefore indicate that

468

variations in MFR between time periods are smaller than variations within a given time

469

period.


21


Page 22 of 37

470

The variations of MFR can be influenced by COA, dp and the POA volatility distribution in

471

the tunnel. We modeled the TD MFR using the median volatility distribution measured in

472

the tunnel and the average COA and dp of each different time period, and we compared

473

the variation of modeled TD MFR with the variation of average measured TD MFR in

474

different time periods (Table S7-1 and Table S7-2). The result suggests that compared

475

with the POA volatility distribution, the variations of COA and dp have a minor impact on

476

the TD MFR. Therefore, the limited temporal variation of MFR further suggests that

477

there are not substantial diurnal changes in POA volatility.

478

Quartz filters cannot capture all gas phase organics in the atmosphere. Thus our

479

discussion on the temporal variation of POA volatility regards only the material collected

480

onto the quartz filter under tunnel ambient conditions. For those gas phase organics that

481

have not been collected onto the quartz filter, we quantify neither the temporal variability

482

nor the contribution to total organic emissions. We therefore do not know how those gas

483

phase organics that broke through the bare-Q filter would influence the POA volatility

484

distribution. However, as these vapors are expected to be IVOCs or more volatile, their

485

impact on partitioning should be minimal.

486

Atmospheric Implications

487

Data from three independent methods all indicate that a significant portion of the vehicle

488

emitted POA is semivolatile and actively partitions between the vapor and condensed

489

phases. The results presented here also indicate that the volatility distributions for

490

gasoline and diesel vehicles determined in a dynamometer setting adequately describe

491

observed partitioning in the tunnel, and that POA partitioning in the ambient atmosphere

492

can be described by a single volatility distribution.


22

Page 23 of 37


493

Our results further suggest that lubricating oil is the major component of vehicular POA,

494

echoing the results of Worton et al.30 If lubricating oil, rather than combustion products,

495

truly dominates vehicle POA emissions, it may explain why seasonal variations in POA

496

volatility seem to be minor. Formulations of lubricating oil composition do not vary

497

seasonally, unlike fuel composition variations between summer and winter in many

498

locations.58,59

499

While the lubricating oil dominated volatility distributions provide reasonable predictions

500

of observed TD data, none capture the ~15% of apparently very low volatility OA mass

501

that does not evaporate at 150 °C. The implication is that the volatility distributions may

502

be too volatile, and therefore overpredict evaporation.

503

As noted above, this disagreement between TD observations and predicted evaporation

504

at high temperatures is commonly observed.22,23 Simply adding additional nonvolatile or

505

extremely low volatility material to the volatility distribution does not reconcile the

506

discrepancy, as the low volatility material persists at lower TD temperatures, leading to

507

over predictions in MFR. Adsorption of organics to soot cores may play a role in

508

creating higher than expected MFR at high temperatures. However the data collected

509

here, all of which had OC/EC < 2, should have been impacted by adsorption at all TD

510

temperatures. This suggests that other factors may also be at play, and deserve future

511

consideration.

512 513

Acknowledgement


23


Page 24 of 37

514

We thank the Pennsylvania Department of Transportation (PennDOT) for granting us

515

access to the Fort Pitt Tunnel and providing traffic data. We thank three anonymous

516

reviewers for providing valuable comments on the manuscript. X. L. acknowledges the

517

Steinbrenner Institute at Carnegie Mellon University for the financial support of graduate

518

study. X. L. also recognizes A. L. Robinson, N. M. Donahue, P. J. Adams, E. M. Lipsky,

519

Y. Zhao, R. Saleh and Y. Tan (CMU) for helpful discussion.

520 521

Supporting Information

522

Supporting Information including a description of the traffic in the tunnel, a schematic of

523

measurement station setup, parameters used in the simulations, relevant calculations

524

and measurement data. This material is available free of charge via the internet at

525

http://pubs.acs.org.

526 527

References

528 529 530 531 532 533

(1)

Zhang, R.; Wang, G.; Guo, S.; Zamora, M. L.; Ying, Q.; Lin, Y.; Wang, W.; Hu, M.; Wang, Y. Formation of Urban Fine Particulate Matter. Chem. Rev. 2015, 115 (10), 3803–3855.

534 535 536

(2)

Dallmann, T. R.; Harley, R. A. Evaluation of Mobile Source Emission Trends in the United States. J. Geophys. Res. Atmos. 2010, 115 (14), 1–12.

537 538 539

(3)

U.S. EPA. The 2011 National Emissions Inventory. http://www3.epa.gov/ttnchie1/net/2011inventory.html.

540

(4)

Maricq, M. M. Chemical Characterization of Particulate Emissions from Diesel


24

Page 25 of 37


Engines: A review. J. Aerosol Sci. 2007, 38 (11), 1079–1118.

541 542 543 544 545 546

(5)

Kleeman, M. J.; Schauer, J. J.; Cass, G. R. Size and Composition Distribution of Fine Particulate Matter Emitted from Motor Vehicles. Environ. Sci. Technol. 2000, 34 (7), 1132–1142.

547 548 549 550 551

(6)

Schauer, J. J.; Kleeman, M. J.; Cass, G. R.; Simoneit, B. R. T. Measurement of Emissions from Air Pollution Sources. 5. C1-C32 Organic Compounds from Gasoline-powered Motor Vehicles. Environ. Sci. Technol. 2002, 36 (6), 1169– 1180.

552 553 554 555 556

(7)

Fujitani, Y.; Saitoh, K.; Fushimi, A.; Takahashi, K.; Hasegawa, S.; Tanabe, K.; Kobayashi, S.; Furuyama, A.; Hirano, S.; Takami, A. Effect of Isothermal Dilution on Emission Factors of Organic Carbon and n-alkanes in the Particle and Gas Phases of Diesel Exhaust. Atmos. Environ. 2012, 59, 389–397.

557 558 559 560

(8)

Grieshop, A. P.; Lipsky, E. M.; Pekney, N. J.; Takahama, S.; Robinson, A. L. Fine Particle Emission Factors from Vehicles in a Highway Tunnel: Effects of Fleet Composition and Season. Atmos. Environ. 2006, 40, 287–298.

561 562 563 564 565

(9)

Dallmann, T. R.; Demartini, S. J.; Kirchstetter, T. W.; Herndon, S. C.; Onasch, T. B.; Wood, E. C.; Harley, R. a. On-road Measurement of Gas and Particle Phase Pollutant Emission Factors for Individual Heavy-duty Diesel Trucks. Environ. Sci. Technol. 2012, 46 (15), 8511–8518.

566 567 568 569 570

(10) Chirico, R.; Prevot, A. S. H.; DeCarlo, P. F.; Heringa, M. F.; Richter, R.; Weingartner, E.; Baltensperger, U. Aerosol and Trace Gas Vehicle Emission Factors Measured in a Tunnel using an Aerosol Mass Spectrometer and Other On-line Instrumentation. Atmos. Environ. 2011, 45 (13), 2182–2192.

571 572 573 574 575

(11) May, A. A.; Nguyen, N. T.; Presto, A. a.; Gordon, T. D.; Lipsky, E. M.; Karve, M.; Gutierrez, A.; Robertson, W. H.; Zhang, M.; Brandow, C.; et al. Gas- and ParticlePhase Primary Emissions from In-use, On-road Gasoline and Diesel Vehicles. Atmos. Environ. 2014, 88, 247–260.

576 577 578 579 580 581

(12) Hodzic, A.; Jimenez, J. L.; Madronich, S.; Canagaratna, M. R.; Decarlo, P. F.; Kleinman, L.; Fast, J. Modeling Organic Aerosols in a Megacity: Potential Contribution of Semi-volatile and Intermediate Volatility Primary Organic Compounds to Secondary Organic Aerosol Formation. Atmos. Chem. Phys. 2010, 10 (12), 5491–5514.


25


Page 26 of 37

582 583 584 585 586

(13) Shrivastava, M. K.; Lane, T. E.; Donahue, N. M.; Pandis, S. N.; Robinson, A. L. Effects of Gas Particle Partitioning and Aging of Primary Emissions on Urban and Regional Organic Aerosol Concentrations. J. Geophys. Res. Atmos. 2008, 113 (18), 1–16.

587 588 589 590 591

(14) Robinson, A. L.; Donahue, N. M.; Shrivastava, M. K.; Weitkamp, E. a; Sage, A. M.; Grieshop, A. P.; Lane, T. E.; Pierce, J. R.; Pandis, S. N. Rethinking Organic Aerosols: Semivolatile Emissions and Photochemical Aging. Science 2007, 315 (5816), 1259–1262.

592 593 594 595

(15) Kuhn, T.; Biswas, S.; Fine, P. M.; Geller, M.; Sioutas, C. Physical and Chemical Characteristics and Volatility of PM in the Proximity of a Light-Duty Vehicle Freeway. Aerosol Sci. Technol. 2005, 39 (4), 347–357.

596 597 598 599

(16) Lipsky, E. M.; Robinson, A. L. Effects of Dilution on Fine Particle Mass and Partitioning of Semivolatile Organics in Diesel Exhaust and Wood Smoke. Environ. Sci. Technol. 2005, 40 (1), 155–162.

600 601 602 603 604

(17) Grieshop, A. P.; Miracolo, M. a.; Donahue, N. M.; Robinson, A. L. Constraining the Volatility Distribution and Gas-particle Partitioning of Combustion Aerosols Using Isothermal Dilution and Thermodenuder Measurements. Environ. Sci. Technol. 2009, 43 (13), 4750–4756.

605 606 607 608 609

(18) Huffman, J. A.; Docherty, K. S.; Mohr, C.; Cubison, M. J.; Ulbrich, I. M.; Ziemann, P. J.; Onasch, T. B.; Jimenez, J. L. Chemically-Resolved Volatility Measurements of Organic Aerosol from Different Sources. Environ. Sci. Technol. 2009, 43 (14), 5351–5357.

610 611 612 613 614

(19) Verma, V.; Pakbin, P.; Cheung, K. L.; Cho, A. K.; Schauer, J. J.; Shafer, M. M.; Kleinman, M. T.; Sioutas, C. Physicochemical and Oxidative Characteristics of Semi-volatile Components of Quasi-ultrafine Particles in an Urban Atmosphere. Atmos. Environ. 2011, 45 (4), 1025–1033.

615 616 617 618

(20) Ranjan, M.; Presto, A. A.; May, A. A.; Robinson, A. L. Temperature Dependence of Gas–Particle Partitioning of Primary Organic Aerosol Emissions from a Small Diesel Engine. Aerosol Sci. Technol. 2012, 46 (1), 13–21.

619 620 621 622

(21) Presto, A. A.; Hennigan, C. J.; Nguyen, N. T.; Robinson, A. L. Determination of Volatility Distributions of Primary Organic Aerosol Emissions from Internal Combustion Engines Using Thermal Desorption Gas Chromatography Mass Spectrometry. Aerosol Sci. Technol. 2012, 46 (10), 1129–1139.


26

Page 27 of 37


623 624 625 626 627

(22) May, A. A.; Presto, A. A.; Hennigan, C. J.; Nguyen, N. T.; Gordon, T. D.; Robinson, A. L. Gas-particle Partitioning of Primary Organic Aerosol Emissions: (1) Gasoline Vehicle Exhaust. Atmos. Environ. 2013, 77, 128–139.

628 629 630 631 632

(23) May, A. A.; Levin, E. J. T.; Hennigan, C. J.; Riipinen, I.; Lee, T.; Collett, J. L.; Jimenez, J. L.; Kreidenweis, S. M.; Robinson, A. L. Gas-Particle Partitioning of Primary Organic Aerosol Emissions: (2) Diesel Vehicles. Environ. Sci. Technol. 2013, 47, 8288–8296.

633 634 635 636

(24) Roth, C. M.; Goss, K. U.; Schwarzenbach, R. P. Sorption of a Diverse Set of Organic Vapors To Diesel Soot and Road Tunnel Aerosols. Environ. Sci. Technol. 2005, 39 (17), 6632–6637.

637 638 639 640

(25) Roth, C. M.; Goss, K. U.; Schwarzenbach, R. P. Sorption of a Diverse Set of Organic Vapors to Urban Aerosols. Environ. Sci. Technol. 2005, 39 (17), 6638– 6643.

641 642 643 644

(26) Liang, C.; Pankow, J. F.; Odum, J. R.; Seinfeld, J. H. Gas/particle Partitioning of Semivolatile Organic Compounds to Model Inorganic, Organic, and Ambient Smog Aerosols. Environ. Sci. Technol. 1998, 31 (11), 3086–3092.

645 646 647

(27) Pankow, J. F. An Absorption Model of Gas/particle Partitioning of Organic Compounds in the Atmosphere. Atmos. Environ. 1994, 28 (2), 185–188.

648 649 650 651

(28) Donahue, N. M.; Robinson, A. L.; Stanier, C. O.; Pandis, S. N. Coupled Partitioning, Dilution, and Chemical Aging of Semivolatile Organics. Environ. Sci. Technol. 2006, 40 (8), 2635–2643.

652 653 654 655

(29) Robinson, A. L.; Grieshop, A. P.; Donahue, N. M.; Hunt, S. W. Updating the Conceptual Model for Fine Particle Mass Emissions from Combustion Systems. J. Air Waste Manage. Assoc. 2010, 60 (10), 1204–1222.

656 657 658 659 660

(30) Worton, D. R.; Isaacman, G.; Gentner, D. R.; Dallmann, T. R.; Chan, A. W. H.; Ruehl, C.; Kirchstetter, T. W.; Wilson, K. R.; Harley, R. A.; Goldstein, A. H. Lubricating Oil Dominates Primary Organic Aerosol Emissions from Motor Vehicles. Environ. Sci. Technol. 2014, 48 (7), 3698–3706.

661

(31) Shrivastava, M. K.; Lipsky, E. M.; Stanier, C. O.; Robinson, A. L. Modeling


27


Page 28 of 37

662 663 664

Semivolatile Organic Aerosol Mass Emissions from Combustion Systems. Environ. Sci. Technol. 2006, 40 (8), 2671–2677.

665 666 667 668

(32) Murphy, B. N.; Pandis, S. N. Simulating the Formation of Semivolatile Primary and Secondary Organic Aerosol in a Regional Chemical Transport Model. Environ. Sci. Technol. 2009, 43 (13), 4722–4728.

669 670 671 672 673 674

(33) Fountoukis, C.; Racherla, P. N.; Denier Van Der Gon, H. a C.; Polymeneas, P.; Charalampidis, P. E.; Pilinis, C.; Wiedensohler, A.; Dall’Osto, M.; O’Dowd, C.; Pandis, S. N. Evaluation of a Three-Dimensional Chemical Transport Model (PMCAMx) in the European Domain During the EUCAARI May 2008 Campaign. Atmos. Chem. Phys. 2011, 11 (20), 10331–10347.

675 676 677 678 679 680

(34) Fountoukis, C.; Megaritis, A. G.; Skyllakou, K.; Charalampidis, P. E.; Pilinis, C.; Denier Van Der Gon, H. A. C.; Crippa, M.; Canonaco, F.; Mohr, C.; Prévôt, A. S. H.; et al. Organic Aerosol Concentration and Composition Over Europe: Insights from Comparison of Regional Model Predictions With Aerosol Mass Spectrometer Factor Analysis. Atmos. Chem. Phys. 2014, 14 (17), 9061–9076.

681 682 683 684

(35) Kuwayama, T.; Collier, S.; Forestieri, S.; Brady, J. M.; Bertram, T. H.; Cappa, C. D.; Zhang, Q.; Kleeman, M. J. Volatility of Primary Organic Aerosol Emitted from Light Duty Gasoline Vehicles. Environ. Sci. Technol. 2015, 49, 1569–1577.

685 686 687 688

(36) Biswas, S.; Ntziachristos, L.; Moore, K. F.; Sioutas, C. Particle Volatility in the Vicinity of a Freeway With Heavy-duty Diesel Traffic. Atmos. Environ. 2007, 41 (16), 3479–3493.

689 690 691 692

(37) Koo, B.; Knipping, E.; Yarwood, G. 1.5-Dimensional Volatility Basis Set Approach for Modeling Organic Aerosol in CAMx and CMAQ. Atmos. Environ. 2014, 95, 158–164.

693 694 695 696 697

(38) Tkacik, D. S.; Lambe, A. T.; Jathar, S.; Li, X.; Presto, A. A.; Zhao, Y.; Blake, D.; Meinardi, S.; Jayne, J. T.; Croteau, P. L.; et al. Secondary Organic Aerosol Formation from In-use Motor Vehicle Emissions Using a Potential Aerosol Mass Reactor. Environ. Sci. Technol. 2014, 48 (19), 11235–11242.

698 699 700 701 702

(39) Ng, N. L.; Herndon, S. C.; Trimborn, A.; Canagaratna, M. R.; Croteau, P. L.; Onasch, T. B.; Sueper, D.; Worsnop, D. R.; Zhang, Q.; Sun, Y. L.; et al. An Aerosol Chemical Speciation Monitor (ACSM) for Routine Monitoring of the Composition and Mass Concentrations of Ambient Aerosol. Aerosol Sci. Technol. 2011, 45 (7), 780–794.


28

Page 29 of 37


703 704 705 706 707 708

(40) Huffman, J. A.; Ziemann, P. J.; Jayne, J. T.; Worsnop, D. R.; Jimenez, J. L. Development and Characterization of a Fast-Stepping/Scanning Thermodenuder for Chemically-Resolved Aerosol Volatility Measurements. Aerosol Sci. Technol. 2008, 42 (5), 395–407.

709 710 711 712

(41) Riipinen, I.; Pierce, J. R.; Donahue, N. M.; Pandis, S. N. Equilibration Time Scales of Organic Aerosol Inside Thermodenuders: Evaporation Kinetics Versus Thermodynamics. Atmos. Environ. 2010, 44 (5), 597–607.

713 714 715

(42) Saleh, R.; Shihadeh, A.; Khlystov, A. On Transport Phenomena and Equilibration Time Scales in Thermodenuders. Atmos. Meas. Tech. 2011, 4 (3), 571–581.

716 717 718 719

(43) Julin, J.; Winkler, P. M.; Donahue, N. M.; Wagner, P. E.; Riipinen, I. Near-Unity Mass Accommodation Coefficient of Organic Molecules of Varying Structure. Environ. Sci. Technol. 2014, No. 48, 12083–12089.

720 721 722 723 724

(44) Chow, J. C.; Watson, J. G.; Chen, L. W. A.; Chang, M. C. O.; Robinson, N. F.; Trimble, D.; Kohl, S. The IMPROVE_A Temperature Protocol for Thermal/optical Carbon Analysis: Maintaining Consistency With a Long-term Database. J. Air Waste Manag. Assoc. 2007, 57 (9), 1014–1023.

725 726 727 728

(45) Turpin, B. J.; Saxena, P.; Andrews, E. Measuring and Simulating Particulate Organics in The Atmosphere: Problems and Prospects. Atmos. Environ. 2000, 34 (18), 2983–3013.

729 730 731 732

(46) Subramanian, R.; Khlystov, A. Y.; Cabada, J. C.; Robinson, A. L. Positive and negative artifacts in particulate organic carbon measurements with denuded and undenuded sampler configurations. Aerosol Sci. Technol. 2004, 38, 27–48.

733 734 735 736

(47) Turpin, B. J.; Lim, H.-J. Species Contributions to PM2.5 Mass Concentrations: Revisiting Common Assumptions for Estimating Organic Mass. Aerosol Sci. Technol. 2001, 35 (1), 602–610.

737 738 739 740 741

(48) Ma, J.; Li, X.; Gu, P.; Dallmann, T. R.; Presto, A. A.; Donahue, N. M. Estimating Ambient Particulate Organic Carbon Concentrations and Partitioning Using Thermal Optical Measurements and the Volatility Basis Set. Aerosol Sci. Technol. 2016.


29


Page 30 of 37

742 743 744 745 746

(49) Zhang, Q.; Rami Alfarra, M.; Worsnop, D. R.; Allan, J. D.; Coe, H.; Canagaratna, M. R.; Jimenez, J. L. Deconvolution and Quantification of Hydrocarbon-like and Oxygenated Organic Aerosols Based on Aerosol Mass Spectrometry. Environ. Sci. Technol. 2005, 39 (13), 4938–4952.

747 748 749 750 751 752 753

(50) Mohr, C.; Huffman, J. A.; Cubison, M. J.; Aiken, A. C.; Docherty, K. S.; Kimmel, J. R.; Ulbrich, I. M.; Hannigan, M.; Jimenez, J. L. Characterization of Primary Organic Aerosol Emissions from Meat Cooking, Trash Burning, and Motor Vehicles With High-resolution Aerosol Mass Spectrometry and Comparison With Ambient and Chamber Observations. Environ. Sci. Technol. 2009, 43 (7), 2443– 2449.

754 755 756 757 758

(51) Zhang, Q.; Worsnop, D. R.; Canagaratna, M. R.; Jimenez, J.-L. Hydrocarbon-like and Oxygenated Organic Aerosols in Pittsburgh: Insights into Sources and Processes of Organic Aerosols. Atmos. Chem. Phys. Discuss. 2005, 5 (5), 8421– 8471.

759 760 761 762

(52) An, W. J.; Pathak, R. K.; Lee, B. H.; Pandis, S. N. Aerosol Volatility Measurement Using an Improved Thermodenuder: Application to Secondary Organic Aerosol. J. Aerosol Sci. 2007, 38 (3), 305–314.

763 764 765 766 767

(53) Faulhaber, A. E.; Thomas, B. M.; Jimenez, J. L.; Jayne, J. T.; Worsnop, D. R.; Ziemann, P. J. Characterization of a Thermodenuder-particle Beam Mass Spectrometer System for the Study of Organic Aerosol Volatility and Composition. Atmos. Meas. Tech. 2009, 2 (1), 15–31.

768 769 770

(54) Cappa, C. D. A Model of Aerosol Evaporation Kinetics in a Thermodenuder. Atmos. Meas. Tech. 2010, 3 (3), 579–592.

771 772 773 774

(55) Gkatzelis, G. I.; Papanastasiou, D. K.; Florou, K.; Kaltsonoudis, C.; Louvaris, E.; Pandis, S. N. Measurement of Non-volatile Particle Number Size Distribution. Atmos. Meas. Tech. Discuss. 2015, 8 (6), 6355–6393.

775 776 777 778 779

(56) Zhao, Y.; Nguyen, N. T.; Presto, A. A.; Hennigan, C. J.; May, A. A.; Robinson, A. L. Intermediate Volatility Organic Compound Emissions from On-Road Gasoline Vehicles and Small O ff -Road Gasoline Engines. Environ. Sci. Technol. 2016, 50, 4554–4563.

780 781 782

(57) Zhao, Y.; Nguyen, N. T.; Presto, A. A.; Hennigan, C. J.; May, A. A.; Robinson, A. L. Intermediate Volatility Organic Compound Emissions from On-Road Diesel Vehicles: Chemical Composition, Emission Factors, and Estimated Secondary


30

Page 31 of 37

783 784


Organic Aerosol Production. Environ. Sci. Technol. 2015, 49 (19), 11516–11526.

785 786 787 788

(58) Gentner, D. R.; Harley, R. A.; Miller, A. M.; Goldstein, A. H. Diurnal and Seasonal Variability of Gasoline-related Volatile Organic Compound Emissions in Riverside, California. Environ. Sci. Technol. 2009, 43 (12), 4247–4252.

789 790 791 792

(59) Lough, G. C.; Schauer, J. J.; Lonneman, W. A.; Allen, M. K. Summer and Winter Nonmethane Hydrocarbon Emissions from On-road Motor Vehicles in the Midwestern United States. J. Air Waste Manag. Assoc. 2005, 55 (5), 629–646.

793

Figures:


31


Page 32 of 37

794 795 796

(b)

(a) 797 798 799 800 801 802 803 804 805 806 807 808

Figure 1. a) Organic concentrations (CQ) and organic particle phase mass fraction (Xp) determined from quartz filter samples. Organic concentrations are sorted from the lowest to the highest CQ. Data are colored based on the sampling periods. Filled areas represent the particle phase organics (bareQ-QBT) and open areas represent the gas phase organics (QBT). All CQ are OC concentrations multiplied by an organic-mass-to-organic-carbon ratio of 1.2. b) Boxwhisker plot of OC concentrations desorbed at different temperature stages during the OCEC analysis. OC1- OC4 represent OC desorbed at 140 °C, 280 °C, 480 °C and 580 °C, respectively. Results from bare-Q and QBT filters are in black and blue, respectively. The ends of the box represent the first and third quartiles. The center line inside the box is the median. The length of the whiskers covers 99.3 percent assuming the data are normally distributed. For OC2 and OC3 measured from QBT filters the median values overlap with the 25th percentile.

809


32

Page 33 of 37


810 OOA HOA ∆NOx

3

OA(µg/m )

20 15

1000 800 600

10 400

∆NOx(ppb)

25

5 200 0

811 812 813 814 815 816 817

0:00 5/24 FRI

12:00

0:00 5/25 SAT

12:00

0:00 5/26 SUN

12:00

0:00 5/27 MON

12:00

0:00 5/28 TUE

Date and time

Figure 2. Organic aerosol (OA) time series measured by the Aerosol Chemical Speciation Monitor (ACSM). OA has been classified into Hydrocarbon-like Organic Aerosol (HOA, blue areas) and Oxygenated Organic Aerosol (OOA, red areas). The background corrected NOx data (green dash line) are also presented. All data in the figure are 1-hour averages. The dates colored with red were weekends. The Monday during the sampling period was the Memorial Day holiday so we considered it as a weekend day here.

818


33


Page 34 of 37

819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834

Figure 3. Thermodenuder (TD) measurements of the HOA mass fraction remaining (MFR) as a function of TD temperature. We use the box-whisker plot to present tunnel measurements. Each box represents the HOA MFR measured under a certain TD temperature stage. The length of the whiskers covers 99.3 percent assuming the data are normally distributed. Data outside of this range are considered as outliers, which are presented with red crosses. Brown dash line shows the HOA MFR under non-volatile POA assumption. Blue lines and green lines are the TD model results based on the gasoline and the diesel POA volatility distributions measured during the dynamometer testing.22,23 The black line is the modeled HOA MFR based on the volatility distribution measured in this tunnel study. Red lines are modeled HOA MFR based on the parameterization from Kuwayama et al.35 The solid red line represents the higher motor oil fraction case while the dashed red line represents the lower motor oil fraction case. All lines in the figure are simulations based on median volatility distribution, averaged OA concentration (COA) and averaged particle mass-median diameter (dp). The shaded region represents the variation of modeled MFR caused by the variation of both gasoline and diesel volatility distributions and the variations of the COA and the dp in the tunnel.

835


34

Page 35 of 37


836

837 838 839 840 841 842 843

Figure 4. The volatility distribution derived from all bare-Q filters. Each box-whisker represents organic mass fractions (fi) measured from all bare-Q filters in each C* bin. The center line is the median, and the top and bottom of the box are 25th and 75th percentile, respectively. The blue and green symbols represent the gasoline and the diesel POA volatility distributions, respectively, from dynamometer studies of May et al.22,23 The upper and lower ranges of fi considering the propagated uncertainties are presented in Figure S13 and are not shown here.

844


35


Page 36 of 37

845

846 847 848 849 850

Figure 5. Thermodenuder measurements of HOA evaporation in different time periods with different HDDV fuel fractions. Colors indicate different time periods. Bars represent standard deviation of all measurements under a certain TD temperature and time period.

851


36

Page 37 of 37


852 853

For the table of content use only

854 855 856 857 858 859


37

Gas-Particle Partitioning of Vehicle Emitted ... - ACS Publications

Recommend Documents