High-resolution datasets unravel the effects of sources and

Subscriber access provided by MIDWESTERN UNIVERSITY

Environmental Processes

High-resolution datasets unravel the effects of sources and meteorological conditions on nitrate and its gas-particle partitioning Xurong Shi, Athanasios Nenes, Zhimei Xiao, Shaojie Song, Haofei Yu, Guoliang Shi, Qianyu Zhao, Kui Chen, Yinchang Feng, and Armistead G. Russell Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b06524 • Publication Date (Web): 22 Feb 2019 Downloaded from http://pubs.acs.org on February 23, 2019

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 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 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.

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 35

Environmental Science & Technology

1 2

High-resolution datasets unravel the effects of sources

3

and meteorological conditions on nitrate and its

4

gas-particle partitioning

5 6

Xurong Shi †, Athanasios Nenes‡,§,||, Zhimei Xiao⊥, Shaojie Song#, Haofei Yu∇,

7

Guoliang Shi†,*,Qianyu Zhao†, Kui Chen⊥, Yinchang Feng†, Armistead G. Russell○

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

†State

Environmental Protection Key Laboratory of Urban Ambient Air Particulate Matter Pollution Prevention and Control, College of Environmental Science and Engineering, Nankai University, Tianjin, 300071, China. ‡ Laboratory of Atmospheric Processes and their Impacts, School of Architecture, Civil and Environmental Engineering, École Polytechnique Fédérale de Lausanne, Lausanne, CH-1015, Switzerland § Institute of Chemical Engineering Sciences, Foundation for Research and Technology Hellas, Patras, Greece, GR-26504 || Institute for Environmental Research and Sustainable Development, National Observatory of Athens, Palea Penteli, Greece GR-15236 ⊥ Tianjin Eco-Environmental Monitoring Center. # School of Engineering and Applied Sciences, Harvard University, Cambridge, Massachusetts 02138, USA. ∇ Department of Civil, Environmental and Construction Engineering, University of Central Florida. ○ School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0512 * Corresponding author: Guoliang Shi ([email protected])

1

ACS Paragon Plus Environment


29

ABSTRACT

30

Nitrate is one of the most abundant inorganic water-soluble ions in fine particulate

31

matter (PM2.5). However, the formation mechanism of nitrate in the ambient

32

atmosphere, especially the impacts of its semi-volatility and the various existing

33

forms of “N”, remain under-investigated. In this study, hourly ambient observations

34

of speciated PM2.5 components (NO3-, SO42-, etc.) were collected in Tianjin, China.

35

Source contributions were analyzed by PMF/ME2 (Positive Matrix Factorization

36

using the Multilinear Engine 2) program, and pH were estimated by ISORROPIA-II,

37

to investigate the relationship between pH and nitrate. Five sources (factors) were

38

resolved: secondary sulfate (SS), secondary nitrate (SN), dust, vehicle and coal

39

combustion. SN and pH showed a triangle-shaped relationship. When SS were high,

40

the fraction of nitrate partitioning into the aerosol phase exhibits a characteristic

41

“S-curve” relationship with pH for different seasons. An index (ITL) is developed and

42

combined with pH to explore the sensitive regions of “S-curve”. Controlling the

43

emissions of anions (SO42-. Cl-), cations (Ca2+, Mg2+, etc.) and gases (NOx, NH3, SO2,

44

etc.) will change pH, potentially reducing or increasing SN. The findings of this work

45

provide an effective approach for exploring the formation mechanisms of nitrate

46

under different influencing factors (sources, pH and IRL).

2


Page 2 of 35

Page 3 of 35


TOC art

47

0.6

3

-

ε (NO ε(NO )3 )

nitrate (μg/m3)

Secondary sulfate 1.0 contribution (μg/m3) 0.8

0.4

0.2

0.0 -2

0

2

pH

4

pH pH

48 49

3


6

8


50

INTRODUCTION

51

Nitrate is a dominant inorganic component of PM2.5 during intense haze periods in

52

China,1-4 and arises from secondary formation.5-6 Nitrate is known to form through

53

different chemical pathways during daytime and nighttime:7-9 NO2 homogeneously

54

reacting with OH radical is the most important HNO3 formation pathway during

55

daytime; heterogeneous hydrolysis of NO3 and N2O5 are the most significant

56

pathways of particulate nitrate during nighttime (Eq. S1-S6 of the Supporting

57

Information). Owing to its extremely low volatility, acidic H2SO4 resides in aerosol

58

and tends to associate with NH3/NH4+ from gas-phase to form secondary ammonium

59

sulfate/bisulfate. When sufficient amounts of ammonia are present, HNO3 can react

60

with NH3 to form ammonium nitrate.10-11

61

Nitrate concentrations in particulate matter (PM) are influenced by meteorological

62

conditions (temperature (T), relative humidity (RH)), precursor source emissions (i.e.,

63

NOx), the presence of other ionic species and complex gas and condensed phase

64

chemical reactions. When relative humidity is very low, aerosol nitrate tends to be in

65

the form of solid NH4NO3 according to the reaction𝑁𝐻4𝑁𝑂3(s)↔𝐻𝑁𝑂3(𝑔) + 𝑁𝐻3(

66

𝑔). Increases in temperature shifts the equilibrium to the right hand side, meaning that

67

nitrate remains in the gas-phase and particulate nitrate concentration is low.12 When

68

an aqueous phase exists in the aerosol (promoted by higher RH and lower T), it tends

69

to completely dissolve inorganic nitrate; the volatility of nitrate then becomes strongly

70

influenced by the acidity, or pH, of the aerosol aqueous phase.12-17 Aerosol acidity is 4


Page 4 of 35

Page 5 of 35


71

not only important for aerosol nitrate, but also drives the gas-particle partitioning of

72

other acidic/basic semi-volatile species.15-17 Acidity can also play an important role in

73

the formation of secondary organic aerosols,13-14 and the solubility of trace nutrients

74

that drives the toxicity and nutrient availability in atmospheric aerosol.18-21

75

Thermodynamics and analysis of ambient data show that, under conditions of constant

76

temperature and aerosol liquid water, the fraction of nitrate in deliquesced aerosol,

77

ε(NO3-) (NO3- /(NO3-+HNO3(g))) exhibits an “S-curve” response to aerosol pH (Figure

78

S1).10, 22-23 Sufficiently high pH values, usually above 2.5 to 3,10, 24 ensures that most

79

of the inorganic nitrate condenses onto the aerosol phase,17, 23, 25 but a pH below 1

80

drives all nitrate back into the gas phase. pH in turn is determined by the relative

81

amounts of aerosol water, NH4+, SO42-, NO3- and non-volatile cations (NVC) from

82

numerous sources,26 and to a secondary degree organic species and mixing state.10, 27

83

Source emission patterns, RH, and T are known to vary considerably over space and

84

time, which makes it challenging to unravel the drivers for ambient nitrate, ε(NO3-),

85

and aerosol pH.

86

Here, we study the impact of source emissions and meteorological parameters (RH,

87

T) on aerosol pH, aerosol nitrate concentrations and ε(NO3-). We first collected

88

datasets of hourly ambient observations of speciated PM2.5 during summer in Tianjin,

89

a megacity in northern China, and employed PMF/ME2 (Positive Matrix Factorization

90

using the Multilinear Engine 2, program) source apportionment tools to explore

91

source emission patterns; the ISORROPIA-II thermodynamic model28 was applied to

92

estimate hourly aerosol pH, and to understand its role on nitrate partitioning and 5



93

source. We chose summertime because of the more prominent sensitivity of aerosol

94

nitrate to pH than in other seasons due to the favorable conditions of high temperature

95

and RH for semi-volatile species in the gas phase. Additionally, summertime

96

meteorological conditions also favor the presence of a metastable aqueous phase,

97

which greatly facilitates the interpretation of data and thus further enhance the

98

robustness of pH estimates. Furthermore, the number of pollutant source categories

99

are fewer in summer than other seasons, which simplifies the attributions of nitrate to

100

its formation pathways. Few field-based studies exist that explore how aerosol pH,

101

RH, T, and source contributions affect nitrate concentrations and ε(NO3-),10

102

particularly in China.

103

104

METHODS

105

Sampling. Online aerosol and gaseous measurements with 1-hour time

106

resolution were conducted in Tianjin, a megacity in northern China, from August 1st,

107

2015 to August 23th, 2015. The sampling site was located in a residential area, about

108

200 m away from a major roadway with dense automobile traffic. Mass

109

concentrations of major water-soluble ions (NH4+, Na+, K+, Ca2+, Mg2+, SO42-, NO3-,

110

Cl-, F-, NO2-) and semi-volatile species in the gas phase (HCl, HNO3, NH3) were

111

measured by URG9000B samplers (AIM, URG Corporation) with two ICs.

112

Methanesulfonic acid (20 mM) was used for cation analysis and 0.08 mM

113

Na2CO3/0.01 mM NaHCO3 was used for anion detection. Both ICs were operated in 6


Page 6 of 35

Page 7 of 35


114

isocratic elution at a flow rate of 0.5 mL min-1. Concentrations of trace gases (NO2,

115

SO2, CO, O3, NH3) were measured in hourly temporal resolution by Thermo Fisher

116

Instruments Model 42i, Model 43i, Model 48i, Model 49i, Model 17i instrument,

117

respectively.

118

pH estimation. Aerosol pH is a fundamental property determined by the

119

amount of hydronium ion, H+, and aerosol water content (Eq.1).17, 29 Owing to the

120

lack of an established direct measurement of aerosol pH, thermodynamic analysis of

121

ambient composition data is used to constrain acidity, a method which has been

122

widely used recently.24, 28, 30-36 In this study, the ISORROPIA-II inorganic model was

123

employed to estimate gas-particle partitioning, liquid water content and pH for the

124

Na+-K+-Ca2+-Mg2+-NH4+-SO42--NO3--Cl--H2O aerosol system.20,

125

hourly samples were introduced into ISORROPIA-II to obtain aerosol pH and model

126

concentrations of chemical species (NH3, NO3-, SO42-, Cl-, etc.) in the gas and

127

aerosol phase at thermodynamic equilibrium. Input concentration data included

128

water-soluble ions in PM2.5 and semi-volatile components in the gas phase (HCl,

129

HNO3, NH3) along with RH and T. In this sampling campaign, RH ranged from

130

47.2% to 79.4% with an average value of 60.1%. The forward (in which known

131

quantities are temperature, relative humidity and the total concentrations of NH3,

132

H2SO4, Na, HCl, HNO3, and non-volatile cations Na, K, Ca, Mg) and metastable

133

modes of ISORROPIA-II were selected for the simulations (CaSO4 do precipitate

134

here).

135

calculated as follows:

39-40

36-39

In total, 387

The detailed information was presented in SI. Aerosol pH was

7



𝑝𝐻 = - 𝑙𝑜𝑔10(𝛾𝐻 + 𝑚𝐻 + )

136

Page 8 of 35

(𝐸𝑞 . 1)

137

where 𝑚𝐻 + and 𝛾𝐻 + are the molality (mol∙kg-1 water) and the molality-based

138

activity coefficient of hydrogen ions, respectively.17, 29 𝛾𝐻 + is assumed to be unity in

139

ISORROPIA-II when single-ion activities for H+ are required,28 which introduces

140

only minor uncertainties in pH calculations.40

141

Source apportionment. Source contribution was calculated with a 1-h time

142

solution by PMF/ME2, a widely used receptor model developed by Paatero.41-42 Prior

143

information was incorporated into the model in the forms of auxiliary equations,

144

which were included as additional terms 𝑸∂𝒖𝒙 in an enhanced object function

145

𝑸𝒆𝒏𝒉.43-44 The equation can be written as follows:

146

𝑄𝑒𝑛 h = 𝑄𝑚𝑎𝑖𝑛 + 𝑄∂𝑢𝑥

(𝐸𝑞 . 2)

147

One of the common auxiliary equations is a “pulling equation” that “pulls” 𝑓𝑝𝑘 (for

148

instance) toward the specific target value ∂𝑝𝑘:

149

𝑄∂𝑢𝑥 = 2

(𝑓𝑝𝑘 - ∂𝑝𝑘)2 𝜎∂𝑢𝑥 𝑝𝑘

(𝐸𝑞 . 3)

2

150

where 𝜎∂𝑢𝑥 𝑝𝑘

151

element of factor loading. More description of ME2 was presented in SI. In this study,

152

a data set with 387 rows (number of samples, with 1-h temporal resolution) by 12

153

columns (number of species) was introduced into PMF/ME2 to apportion pollutant

154

source contributions. The species included NO3-, SO42-, NH4+, Cl-, F-, Ca2+, Mg2+, K+,

155

Na+, OC, and EC. Source profiles from previous studies

156

pulling. Ca2+ is the marker for dust46; EC and OC are the markers of vehicle

is the uncertainty associated with the pulling equation; 𝑓𝑝𝑘 is the

8


26, 45

were used for factor

Page 9 of 35


157

(exhaust)47; OC, EC and Cl- are the source markers of coal combustion48; the markers

158

of secondary sulfate (SS) and secondary nitrate (SN) are ammonium, sulfate and

159

nitrate49.

160

161

RESULTS AND DISCUSSION

162

Water-soluble ions, pH, and sources. Time series of the major

163

water-soluble ions (SNA: SO42-, NO3-, NH4+), and meteorological conditions are

164

presented in Figure 1. The mean ambient temperature was 301.4K (298.2K-303.8K),

165

and mean RH was 60.1% (47.2%-79.4%). The highest concentrations of NO3-, SO42-,

166

and NH4+ were 49.9, 40.9, 27.5 μg m-3, while the lowest were 3.8, 4.0, 3.2 μg m-3,

167

respectively. Concentrations of semi-volatile species including NO3- and NH4+ were

168

negatively correlated with temperature and positively correlated with RH. SO42- was

169

positively correlated with T and RH (Table S1). SO42- was positive correlated with T

170

and RH (Table S1). Total ammonia (molar concentration of gas-phase ammonia and

171

aerosol ammonium, [TA]) largely exceeded the 2:1 stoichiometric ratio for

172

(NH4)2SO4, [TS] (i.e., [TA]-2[TS] > 0).

173

The calculated aerosol pH ranged from 2.6 to 4.6, with an average of 3.4±0.5 (mean

174

± standard deviation) (Figure 1); the correlations (r) between measured and modeled

175

concentrations were 0.98 for NO3-, 0.98 for NH4+and 0.97 for NH3 (g); and the

176

regression slopes were 1.02, 1.08, 1.00 for NO3-, NH4+ and NH3 (g) (see Figure S2,

177

black point) which has been used in the past to support that the pH values are 9



178

representative.11, 50 Additionally, the root mean squared error (RMSE) and normalized

179

mean bias (NMB) between measured and modeled concentrations were 2.62, 3.63,

180

4.03 and -0.09, -0.27, 0.16 for NO3-, NH4+ and NH3 (g) (see Table S2). More detailed

181

information are presented in SI. In our data, the agreement between estimated and

182

observed gas-phase NH3 values does not in itself ensure that the pH values are

183

well-constrained – because most of the total ammonia resides in the gas phase, and

184

hence its vapor pressure is insensitive to pH errors.25 Concentrations of inorganic

185

species in the aerosol and gas-phase are reproduced well by the thermodynamic

186

calculations, including NH4+, suggesting that the partitioning fractions of the

187

semi-volatile inorganics NH3/ NH4+ and HNO3/NO3- are well captured (Figure S2).

188

Additionally, sensitivity tests were also performed. One set of tests involved

189

excluding NVCs (all NVC concentrations were set to 0, and using the observed NH4+)

190

for calculating pH (identified as “pH*”), to show the importance of using NVCs in

191

ISORROPIA-II for the observed conditions. In Figure S2, regression slopes of

192

measured vs. predicted values of NH4+, NO3-, and NH3(g) were from 1.00 to 1.13 and

193

R2’s were from 0.97 to 0.99, suggesting that ISORROPIA II captured the partitioning

194

of nitrate and ammonium and that, in this case, the presence of NVCs had a limited

195

impact on the modeled partitioning. Besides, fitted regression equations between pH

196

estimated by E-AIM and by ISORROPIA were shown in Figure S2 (b). The estimated

197

pH were 3.59 (ISORROPIA) and 3.03 (E-AIM), and the correlation (r) of these two

198

models’ results was 0.74, indicated that the pH from two models were similar. Given

199

above information, and that for the RH, T values here a single (metastable) aqueous 10


Page 10 of 35

Page 11 of 35


200

phase is present suggest the summertime pH values calculated are well constrained by

201

the data.

202

A matrix of 387 rows (number of samples, with 1-h temporal resolution) by 12

203

columns (number of chemical species) was introduced into ME2. Five factors were

204

resolved by the model and identified by the source markers. Qmain was about 2750

205

(theoretical Q=387×12-5×(387+12)=2649). The secondary source factors including

206

secondary sulfate (SS) and secondary nitrate (SN) are formed through complex

207

chemical reactions involving gas precursors. SS and SN were characterized by the

208

presence of ammonium, sulfate, and nitrate.49 In this work, SN extracted by

209

PMF/ME2 include only condensed phase nitrate, but not nitric acid vapor. The dust

210

factor was determined by relatively high Ca2+ abundance;46 the vehicle factor was

211

distinguished by the high weights of carbonaceous species (EC and OC);47 and the

212

coal factor was characterized by large proportions of OC, EC, and Cl- .48 Detailed

213

source profiles were provided in Supporting Information (Figure S3). In addition, the

214

hourly time series of contributions from different sources/factor were also calculated

215

by ME2 (Figure S3), with average contributions ranked as SS (contributed 25.2% to

216

PM2.5), SN (contributed 23.0%), dust (contributed 22.7%), coal (contributed 16.5%),

217

and vehicle (contributed 12.6%). The fitted regression of modeled against measured

218

hourly concentrations of PM2.5 are shown in Figure S3(d). The regression slope (0.89)

219

and R2 (0.89) indicate that the performance of source apportionment is satisfactory.

220

(Figure 1)

11



221

Diurnal variations of NO3-, sources, and pH. The observed diurnal

222

variations of NO3- and pH are presented in Figure 2. To better explore the underlying

223

causes of such variation, we compared the diurnal variations of NO3- with those for

224

pH, sources, and other ions (Figure 2). The temporal patterns of NO3- exhibited more

225

similarities with pH compared to SO42- and NH4+. It was partially because low pH can

226

lead to the loss of semi-volatile NO3-, while high pH can enhance nitrate partitioning

227

into particle phase.17 Higher SO42- fraction in the aerosol, which is also associated

228

with [TS] in the region, tends to decrease pH.26

229

Figure. 2b shows that aerosol pH, NO3-, HNO3 (g) concentrations and ε(NO3-)

230

(calculated as [NO3-]/([HNO3]+[NO3-]), with [HNO3] and [NO3-] being the mole

231

concentration of gas-phase HNO3 and aerosol NO3, respectively (in mole m-3) all have

232

distinct diurnal variations. Aerosol pH was relatively lower during daytime, and

233

higher during nighttime (Figure 2a). HNO3 (g) concentration (μg m-3) was higher in

234

daytime than nighttime, while NO3- concentration (μg m-3) and ε(NO3-) were higher

235

during nighttime than daytime. Such diurnal patterns of HNO3 (g) can be explained as

236

follows. O3 concentration and atmospheric temperature were higher during daytime

237

(Figure S4) and the diurnal patterns of HNO3 (g) concentration was opposite to that of

238

NOx (NO+NO2, reactive precursors of nitrate, Figure S5), suggesting that high

239

gaseous HNO3 during daytime was photochemically driven.51-52 The high nighttime

240

RH and lower T caused nocturnal NO3- to be much higher than daytime (Figure S4),

241

because of the nitrate equilibrium constant and aerosol liquid water increases so that

242

equilibrium is shifted to particle phase. Compared with the aerosol pH led by 12


Page 12 of 35

Page 13 of 35


243

(NH4)2SO4, the formation of deliquesced NH4NO3 in the aerosol phase tends to

244

increase pH, which in turn favors even more condensation of nitrate and associated

245

ammonium from the gas phase.10-11 Higher RH in the nighttime than in daytime also

246

promotes higher water content in the aerosol, which drives much of the diurnal

247

variability in aerosol pH.24,

248

partitioning of nitrate, and it was higher during nighttime and lower during daytime

249

(Figure 2b). The diurnal variation of ε(NO3-) suggested that most of the nitrate was in

250

particle phase, especially during morning and nighttime. Most of the nitrate volatilizes

251

during the day, closely following the trends of T, RH and aerosol pH. In this study,

252

the mean ε(NO3-) was 76.7%, which is lower than that of Song et al. (99.6%, Beijing,

253

China)40 and higher than that of Guo et al. (45.67%, Pasadena, CA. USA).25 This

254

result was likely because the sampling campaign of this study was conducted in

255

August (versus winter for Song et al.), and the mean estimated pH in this study was

256

3.4 vs. 4.2 in Song et al.40 Higher temperature and more acidic environment are

257

expected to enhance nitrate volatilization and its partitioning to gas phase, leading to

258

lower ε(NO3-) in this study.10-11, 17, 20

53

Additionally, ε(NO3-) reflects the particle-gas

259

(Figure 2)

260

Emissions from different factors can directly influence aerosol pH.26 Figure S6

261

shows diurnal variations of factor contributions and pH. SS is negatively correlated

262

with pH (especially during daytime) (Figure S6c), as sulfuric acid is a strong acid

263

with a very low vapor pressure, thus increased temperatures and lowered RH are

264

expected to lead to reduced aerosol water, consequently concentrating acid. This in 13



265

turn leads to a persistently acidic aerosol11,

266

when concentrations of other aerosol components decrease while the relative

267

proportion of sulfate increases.15,

268

contributions and pH were found to be positive (Figure S6c) because the pH of pure

269

deliquesced ammonium nitrate (and other nitrate salts with NVCs) is much higher

270

than the pH associated with an aerosol dominated by SS.25

26

15

which becomes increasingly acidic

Unlike SS, the correlations between SN

271

Figure S6 also presented diurnal variations of contributions from other

272

sources/factors, and pH. Aerosol pH starts decreasing at 7:00 am, and reaches a

273

minimum at about 2:00 pm, a pattern similar to those for dust and coal contributions,

274

but different from vehicle contribution. The heating and expansion of the boundary

275

layer generally leads to the diurnal cycle of RH – that generates a corresponding

276

diurnal cycle in aerosol liquid water. These temporal changes induce a pH diurnal

277

cycle,17, 24, 53 which together with the diurnal temperature cycle, generates the diurnal

278

cycle of nitrate partitioning as seen in the data seen here (Figure 2b). However,

279

cations that originate from dust and coal combustion (NH3/NH4+, Ca2+, etc)54 can

280

readily increase aerosol pH,26 so a part of the diurnal pH cycle may be driven by

281

emissions – especially from NVCs (Figure S6e). Coal combustion and vehicular

282

traffic also contribute acidic precursors (SO2, NOx) which modulate acidity as

283

described above in terms of contributing SS and SN.

284

Impacts of aerosol pH on secondary nitrate formation under

285

different sources contributions. When SN contribution (calculated from ME2)

286

is plotted against pH, an interesting “triangle” relationship can be observed (Figure 3); 14


Page 14 of 35

Page 15 of 35


287

four sub-regions (identified as A to D) are identified, based on time period,

288

meteorological regime, and source/factor contributions (Figure 3, 4, S7).

289

Region A contains relatively few points, mostly observed at night, and is

290

characterized by high PM2.5, SNA concentrations and moderate aerosol pH (~ 3.2).

291

Samples in Region B were mostly collected between 0:00-8:00, and were

292

characterized by moderate RH, NO3-, NO2, as well as low SO42-, O3, and temperature.

293

Hence, SN in this region were associated with liquid-phase reactions. It should be

294

noted that aerosol pH in Region B was higher than in other regions (Figure 3) because

295

of its higher liquid water content (Figure S8 and S9). Samples in regions B have

296

considerable dust contributions and low contributions from SS (Figure 4), which

297

further increases pH. Also, from Figure 4, as aerosol pH increases, the contributions

298

of SN decreases in Region B (Figure 3), owing to the reductions in total nitrate –

299

which in turn is linked to reductions in NOx emissions. SS concentrations were

300

relatively lower in region B (Figure 4), which could help promoting aerosol pH

301

increase, as the importance of NVC tends to promote pH to higher values. The large

302

amounts of NH3 present, which tends to maintain pH high enough to readily promote

303

NO3 condensation, implies that reductions in SS also decreases SN, because

304

reductions in aerosol liquid water associated with SS loss promotes volatilization of

305

nitrate back to the gas-phase.

306

Samples in Region C were collected between 13:00-00:00, and mostly between

307

19:00-00:00. Temperature, O3, and RH were all at moderate levels in this region

308

(Figure S7), suggesting that secondary nitrate in Region C was likely formed through 15



309

both liquid- and gas-phase and photochemical reactions. Samples in Region D were

310

mostly collected between 9:00-17:00, and were characterized by high O3, temperature

311

(306.0K), and SO42- concentrations, in conjunction with low aerosol pH, NO3-

312

concentration, ε(NO3-), and RH (42.1%) (Figure S7). The secondary nitrate in this

313

region was likely formed through photochemical reactions. In Region D, T was

314

relatively higher and RH was relatively lower, which would drive nitrate into its gas

315

phase and decrease the nitrate concentrations in particle phase. The aerosols in this

316

region were very acidic (pH from 2-2.5) as a result of high liquid H+ concentration

317

and low water content in the same time period (Figure S8, S9). Also, dust contributed

318

poorly in this region, and thus had little influence on elevating aerosol pH.

319

(Figure 3, Figure 4)

320

Nitrate partitioning “S-curve” in actual atmospheric environment.

321

For a NO3-- HNO3 (g) system, under a given constant T and RH, an ideal S-curve

322

relationship between particle-gas partitioning of semi-volatile species and aerosol pH

323

(see Figure S1) would be found,12 which can be applied to estimate aerosol pH and

324

aqueous fractions of semi-volatile species.11, 15 There are three regions in Figure S1.

325

Region (1) and (3) are pH-insensitive bands, where the ε(NO3-) changes slightly with

326

respect to pH change. Region (2) is a pH-sensitive band, where the ε(NO3-) changes

327

drastically with pH. For the ambient dataset, the pH-ε(NO3-) curve might be less

328

obvious because T, liquid water content, and ionic strength (activity coefficients) do

329

not remain constant. Here, we plotted ε(NO3-) (calculated from the measured mole

330

concentration of NO3- and HNO3(g)) ratio against aerosol pH (Figure S10). At first 16


Page 16 of 35

Page 17 of 35


331

glance, no obvious S-curve between ε(NO3-) and pH can be observed. However, after

332

conditional sampling of the data (i.e., when SS contributions are relatively higher, i.e.,

333

greater than 25 ug m-3; Figure S10(b)), the ambient samples do captured the

334

pH-ε(NO3-) sensitive band of the S-curve (Figure S1).

335

To empirically describe the relationships between ε(NO3-) ratio and pH, a

336

Boltzmann equation was applied to fit the curve of pH-ε(NO3-). The Boltzmann

337

equation can be used to fit a sigmoid shaped curve and has performed well in the

338

previous works55-56.

339

1

-

𝜀 (𝑁𝑂3 ) = 1 -

(𝑝𝐻 - 1.99)

1+𝑒

(𝐸𝑞 . 4)

0.65

340

Origin 8.5 software was employed to perform the curve fitting (χ² was 0.007 and R2

341

was 0.70) and the result was shown in Figure 5. The fitted curve in Figure 5a is only a

342

portion (sensitive region) of the S-curve, as the range in ambient aerosol pH covers

343

only a fraction of the pH range as encompassed by the entire S-curve (Figure S1). Eq.

344

4 was extrapolated to outside of the conditions as found in this study, specifically

345

from 0-2 and 4-8. Values of the simulated ε(NO3-) ratios (red points in Figure 5b)

346

were also calculated accordingly using (Eq. 4). Results of the extrapolation are

347

provided in Figure 5b, where the extrapolated data are shown as red points and the

348

actual ambient sample data are shown as black points. The results show that the

349

collected ambient samples (black points) were in the pH-ε(NO3-) sensitive region,

350

indicating that a small change of aerosol pH can have a considerable impact on

351

ε(NO3-) (Figure 5b), which reaches maximum at ε(NO3-) ~ 0.5. Nonetheless, there are

17



Page 18 of 35

352

still a few black points that deviate from the fitted sigmoid curve, possibly due to

353

uncertainties involved in curve fitting and analytical solution. The above statistical

354

analysis shows that an S-curve can be found for the ambient dataset in summer. The existence of an S-curve can also be found in thermodynamic estimations. As

355 356

shown in Eq.5- Eq.6:12

357

𝜀(

358

where HA is the Henry Law Constant for HNO3 in M atm-1; R is the ideal-gas constant

359

equal to 0.08205 atm L mol-1 K-1; T is the temperature in K; and L is the aerosol liquid

𝑁𝑂3-

10 -6𝐻𝐴𝑅𝑇𝐿

) = 1 + 10

360

water content in g

361

12,

362

𝜀(𝑁𝑂3- ) ≈ [𝐻 + ] + 𝐼

(𝐸𝑞. 5)

-6𝐻 𝑅𝑇𝐿 𝐴

m-3

3.2 × 106

(as estimated by ISORROPIA-II). Given that 𝐻𝐴 ≈ [𝐻 + ]

and defining 3.2RTL as the index ITL, we obtain: 𝐼𝑇𝐿

(𝐸𝑞. 6)

𝑇𝐿

363

Combining the index ITL, pH, and Eq(6), we can explain the shape of the

364

pH-ε(NO3-) relationship and help identifying the sensitive band of the S-curve. If [H+]

365

is in the range of 0.0001 to 0.01 (pH from 2 to 4), ε(NO3-) is sensitive to [H+]. In this

366

work, ITL ranges from 0.0001 to 0.0175, with an average of 0.0026, predominantly in

367

the sensitive range. When pH is less than 2, [H+] is higher than the maximum value of

368

ITL (0.0175), and the effects of varying T and L on ε(NO3-) can be ignored. When pH

369

is great than 3.3 (especially > 4), [H+] is lower than the minimum value of ITL

370

(0.0001), hence the variation of ambient T and L would greatly influence ε(NO3-)

371

(according to Figure 6, the number of scattered points increased when pH > 3.3).

372

Additionally, when pH is between 2 and 3.3, both [H+] and ITL can impact ε(NO3-). As

18


Page 19 of 35


373

observed from Figure 6, there are only limited amounts of scattered points between

374

pH 2-3.3.

375

Figure 6 also shows the relationships between pH and ε(NO3-) under different

376

meteorological conditions, and for different source categories (in μg m-3, as estimated

377

by ME2). The collected samples in the pH-ε(NO3-) sensitive band were characterized

378

by high SS contribution, moderate vehicle and dust contributions, and low coal

379

contribution. As shown in our previous work, SS dominates in low pH (2-3).26

380

Moderate dust contribution would provide non-volatile cations to associate with

381

sulfate and elevate pH. Additionally, moderate vehicle contribution can provide NH3

382

and NOx precursors for nitrate aerosol. The low coal contribution suggests relatively

383

low emissions from coal combustion, thus making the S curve more prominent. We

384

also performed sensitivity tests and found that if the concentrations of sulfate

385

decreases, the shape of ε(NO3-) can be “distorted” because aerosol liquid water is

386

appreciably changed as nitrate condenses and evaporates (Figure S11), something that

387

does not occur when SS dominates the liquid water uptake.

388

Figure 12 shows the relationship between ε(NO3-)-pH in other seasons (spring:

389

2015.4.30-2015.5.30 and winter: 2015.1.4-2015.2.10). Similarly, parts of the S-curves

390

were observed in spring and winter datasets that have higher SS contributions (Figure

391

7). For spring, the pH-sensitive band was fit to a sigmoid curve: ε(NO3 ) = 1 -

-

1

392

(𝑝𝐻 - 2.8)

1+𝑒

393

, (R2=0.67). For winter, the pH-insensitive band (right band, Fig 1) was

0.71

found, likely due to the considerable variability in aerosol liquid water, temperature,

19



394

and emissions patterns over the time period examined. ITL ranged from

395

0.0000-0.0094, with an average of 0.00067 for spring samples. So, the range of pH for

396

sensitive band of the S-curve should be from 2-4, which can be confirmed by Figure 7

397

(left side, spring). Additionally, ITL ranged from 0.0000-0.012, with an average of

398

0.0013 for winter samples, suggesting that the range of pH for sensitive band should

399

from 1-4. In Guo et al. (2017), the S-curves were found in the range of pH from -1 to

400

3 in winter. But in this work, pHs typically were higher than 3 in winter (see Figure

401

S12), so nitrate partitioning tends to reside in the upper “plateau” of the S-curve,

402

which are modulated by variations in T and L. Similar as in the summer, TN

403

(NO3-+HNO3 (g)) vs. pH for winter and spring seasons yield “triangles,” indicating

404

that similar emissions and chemical domains are responsible for the behavior. A

405

detailed discussion can be found in Supporting Information. (Figure 5, Figure 6, Figure 7)

406

407

IMPLICATIONS

408

The use of highly time resolved (hourly) observations of aerosol compositions over

409

multiple seasons allowed more detailed examination of the aerosol thermodynamics in

410

different seasons, and the relationships between aerosol composition, sources, pH, and

411

meteorological factors.

412

average pH of 3.4±0.5. The gas-particle partitioning of the observed semi-volatile

413

inorganic species (ammonia-ammonium and nitric acid-nitrate) are consistent with the

414

thermodynamics of the ammonia-sulfate-nitrate-NVC system. An interesting

In this study, aerosol pH ranges from 2.6 to 4.6, with an

20


Page 20 of 35

Page 21 of 35


415

“triangle” relationship was found between the SN contribution and pH. This was

416

found to primarily be linked to the emissions of nitrate precursors (NOx and NH3) and

417

physicochemical processes leading to SN formation. A triangular shape relationship

418

between SN and pH is found.

419

favorable for SN formation (i.e., lower temperatures, higher RH) the pH is pushed

420

towards the value of pure ammonium nitrate’s; as SS increases, pH decreases (left

421

side of triangle), while a higher fraction of NVCs raises pH because TN and SS

422

decrease (right side of triangle) when the air masses are cleaner.

In more polluted environments, in conditions

423

The use of hourly samples further support that the “S-curve” nitrate-nitric acid gas

424

and ammonium-ammonia gas partitioning relationships, particularly when SS levels

425

are higher, leading to greater aerosol water levels. The index (ITL) developed in this

426

work, can help determining when nitrate partitioning is in the pH-sensitive region.

427

This region is also where ammonia and NOx emissions control strategies are both

428

efficient in reducing the nitrate aerosol load, though sulfate reductions will be less

429

efficient. But at very high or low ITL’s, this substitution is less important. Given that

430

the ITL’s vary throughout the day, the use of hourly observations can be important as

431

24-hour samples can mis-identifying times when controls are effective.

432

433

SUPPORTING INFORMATION

434

Other figures needed to evaluate the conclusions in the paper are present in the

435

Supporting Information. 21



436

Figure S1 – S13

437

AUTHOR INFORMATION

438

Corresponding Author

439

* E-mail: [email protected]

440

Note

441

The authors declare that they have no competing interests.

Page 22 of 35

442

443

ACKNOWLEDGEMENTS

444

This study was supported by the National Natural Science Foundation of China

445

(41775149), National Key Research and Development Program of China

446

(2016YFC0208500, 2016YFC0208505), Special Scientific Research Funds for

447

Environment Protection Commonweal Section (Nos. 201509020), the Tianjin

448

Research

449

(14JCQNJC0810),

450

16JCQNJC08700), and the Blue Sky Foundation, Fundamental Research Funds for

451

the Central Universities. This publication was developed under Assistance Agreement

452

No. EPA834799 awarded by the U.S. Environmental Protection Agency to Emory

453

University and Georgia Institute of Technology. It has not been formally reviewed by

454

EPA. The views expressed in this document are solely those of the authors and do not

455

necessarily reflect those of the Agency. EPA does not endorse any products or

Program

of

Application

Tianjin

Natural

Foundation Science

and

Advanced

Foundation

22


Technology

(17JCYBJC23000,

Page 23 of 35


456

commercial services mentioned in this publication. We also acknowledge support

457

from the project PyroTRACH (ERC-2016-COG) funded from H2020-EU.1.1. -

458

Excellent Science - European Research Council (ERC), project ID 726165. We also

459

thank Ms. G. Pavur for her help in preparing the manuscript.

460

23



461

REFFERENCES

462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505

(1) Wu, C.; Wang, G. H.; Wang, J. Y.; Li, J. J.; Ren, Y. Q.; Zhang, L.; Cao, C.; Li, J.; Ge, S. S.; Xie, Y. N.; Wang, X. P.; Xue, G. Y. Chemical characteristics of haze particles in Xi'an during Chinese Spring Festival: Impact of fireworks burning. J. Environ. Sci. 2018, 71, 179-187. (2) Yao, X.; Chan, C. K.; Fang, M.; Cadle, S.; Chan, T.; Mulawa, P.; He, K.; Ye, B. The water-soluble ionic composition of PM2.5 in Shanghai and Beijing, China. Atmos. Environ. 2002, 36, 4223−4234. (3) Qiao, T.; Zhao, M.; Xiu, G.; Yu, J. Seasonal variations of water soluble composition (WSOC, Hulis and WSIIs) in PM1 and its implications on haze pollution in urban Shanghai, China. Atmos. Environ. 2015, 123, 306-314. (4) Hua, Y.; Cheng, Z.; Wang, S. X.; Jiang, J. K.; Chen, D. R.; Cai, S. Y.; Fu, X.; Fu, Q. Y.; Chen, C. H.; Xu, B. Y.; Yu, J. Q. Characteristics and source apportionment of PM2.5 during a fall heavy haze episode in the Yangtze River Delta of China. Atmos. Environ. 2015, 123, 380-391. (5) Tian, Y. Z.; Liu, J. Y.; Han S. Q.; Shi, X. R.; Shi, G. L.; Xu, H.; Yu, H. F.; Zhang, Y. F.; Feng, Y. C.; Russell, A. G. Spatial, seasonal and diurnal patterns in physicochemical characteristics and sources of PM2.5 in both inland and coastal regions within a megacity in China. J. Hazard. Mater. 2018, 342, 139-149. (6) Gao, J.; Peng, X.; Chen, G.; Xu, J.; Shi, G. L.; Zhang, Y. F.; Feng, Y. C. Insights into the chemical characterization and sources of PM2.5 in Beijing at a 1-h time resolution. Sci. Total Environ. 2016, 532, 162-171. (7) Wang, X. F.; Zhang, Y. P.; Chen, H.; Yang, X.; Chen, J. M. Particulate nitrate formation in a highly polluted urban area: a case study by single-particle mass spectrometry in Shanghai. Environ. Sci. Technol. 2009, 43, 3061-3066. (8) Fu, X. X.; Guo, H.; Wang, X. M.; Ding, X.; He, Q. F.; Liu, T. Y.; Zhang, Z. PM2.5 acidity at a background site in the Pearl River Delta region in fall-winter of 2007-2012. J. Hazard. Mater. 2015, 286, 484-492. (9) Xue, J.; Yuan, Z. B.; Lau, A. K. H.; Yu, J. Z. Insights into factors affecting nitrate in PM2.5 in a polluted high NOx environment through hourly observations and size distribution measurements. J. Geophys. Res. Atmos. 2014, 119 (8), 4888-4902. (10) Guo, H. Y.; Otjes, R.; Schlag, P.; Kiendler-Scharr, A.; Nenes, A.; Weber, R. J. Effectiveness of ammonia reduction on control of fine particle nitrate. Atmos. Chem. Phys. 2018, 18, 12241-12256. (11) Guo, H. Y.; Weber, R. J.; Nenes, A. High levels of ammonia do not raise fine particle pH sufficiently to yield nitrogen oxide-dominated sulfate production. Sci. Rep. 2017a, 7, 12109; DOI 10.1038/s41598-017-11704-0. (12) Seinfeld, J. H.; Pandis, S. N. Atmospheric chemistry and physics: from air pollution to climate change, Third ed.; John Wiley & Sons: Hoboken, New Jersey, 2016. (13) Cheng, Y.; Zheng, G.; Wei, C.; Mu, Q.; Zheng, B.; Wang, Z.; Gao, M.; Zhang, Q.; He, K.; Carmichael, G.; Pöschl, U.; Su, H. Reactive nitrogen chemistry in aerosol water as a source of sulfate during haze events in China. Sci. Adv. 2016, 2, e1601530; DOI 10.1126/sciadv.1601530. 24


Page 24 of 35

Page 25 of 35

506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553


(14) Xu, L.; Guo, H.; Boyd, C. M.; Klein, M.; Bougiatioti, A.; Cerully, K. M.; Hite, J. R.; Isaacman-VanWertz, G.; Kreisberg, N. M.; Knote, C.; Olson, K.; Koss, A.; Goldstein, A. H.; Hering, S. V.; de Gouw, J.; Baumann, K.; Lee, S. -H.; Nenes, A.; Weber, R. J.; Ng, N. L. Effects of anthropogenic emissions on aerosol formation from isoprene and monoterpenes in the southeastern United States. P. Natl. Acad. Sci. USA. 2015, 112, 37-42. (15) Weber, R. J.; Guo, H.; Russell, A. G.; Nenes, A. High aerosol acidity despite declining atmospheric sulfate concentrations over the past 15 years. Nat. Geosci. 2016, 9, 282-285. (16) Keene, W. C.; Pszenny, A. A. P.; Maben, J. R.; Stevenson, E.; Wall, A. Closure evaluation of size-resolved aerosol pH in the New England coastal atmosphere during summer. J. Geophys. Res. Atmos. 2004, 109, D23307; DOI 10.1029/2004JD004801. (17) Guo, H. Y.; Sullivan, A. P.; Campuzano-Jost, P.; Schroder, J. C.; Lopez-Hilfiker, F. D.; Dibb, J. E.; Jimenez, J. L.; Thornton, J. A.; Brown, S. S.; Nenes, A.; Weber, R. J. Fine particle pH and the partitioning of nitric acid during winter in the northeastern United States. J. Geophys. Res. Atmos. 2016, 121, 10355-10376. (18) Longo, A.F.; Feng, Y.; Lai, B.; Landing, W.M.; Shelley R.U.; Nenes, A.; Mihalopoulos, N.; Violaki, K.; Ingall, E.D. Influence of Atmospheric Processes on the Solubility and Composition of Iron in Saharan Dust. Environ. Sci. Technol. 2016, 50, 6912–6920. (19) Nenes, A.; Krom, M.D.; Mihalopoulos, N.; Van Cappellen, P.; Shi, Z.; Bougiatioti, A.; Zarmpas, P.; Herut, B. Atmospheric acidification of mineral aerosols: A source of bioavailable phosphorus for the oceans. Atmos. Chem. Phys. 2011, 11, 6265-6272 (20) Meskhidze, N.; Chameides, W.; Nenes, A.; Chen, G. Iron Mobilization in Mineral Dust: Can Anthropogenic SO2 Emissions Affect Ocean Productivity?. Geophys. Res. Lett. 2003, 30(21), 2085; DOI 10.1029/2003GL018035. (21) Fang, T., Guo, H., Zeng, L., Verma, V., Nenes, A., Weber, R.J. Highly acidic ambient particles, soluble metals and oxidative potential: A link between sulfate and aerosol toxicity. Environ. Sci. Technol. 2017, 51 (5), 2611–2620. (22) Nah, T.; Guo, H. Y.; Sullivan, A. P.; Chen, Y. L.; Tanner, D. J.; Nenes, A.; Russell, A. G.; Ng, N. L.; Huey, L. G.; Weber, R. J. Characterization of aerosol composition, aerosol acidity and organic acid partitioning at an agriculture-intensive rural southeastern U. S. Site. Atmos. Chem. Phys. 2018, 18, 11471-11491. (23) Vasilakos, P.; Russell, A. G.; Weber, R. J.; Nenes, A. Understanding nitrate formation in a world with less sulfate. Atmos. Chem. Phys. 2018, 18, 12765-12775. (24) Bougiatioti, A.; Nikolaou, P.; Stavroulas, I.; Kouvarakis, G.; Weber, R.; Nenes, A.; Kanakidou, M.; Mihalopoulos, N. Particle water and pH in the eastern Mediterranean: source variability and implications for nutrient availability. Atmos. Chem. Phys. 2016, 16, 4579-4591. (25) Guo, H. Y.; Liu, J.; Froyd, K. D.; Roberts, J. M.; Veres, P. R.; Hayes, P. L.; Jimenez, J. L.; Nenes, A.; Weber, R. J. Fine particle pH and gas–particle phase partitioning of inorganic species in Pasadena, California, during the 2010 CalNex campaign. Atmos. Chem. Phys. 2017b, 17, 5703-5719. (26) Shi, G. L.; Xu, J.; Peng, X.; Xiao, Z. M.; Chen, K.; Tian, Y. Z.; Guan, X. B.; 25



554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601

Feng, Y. C.; Yu, H. F.; Nenes, A.; Russell, A. G. pH of aerosols in a polluted atmosphere: source contributions to highly acidic aerosol. Environ. Sci. Technol. 2017, 51, 4289-4296. (27) Pye, H. O. T.; Zuend, A.; Fry, J. L.; Isaacman-VanWertz, G.; Capps, S. L.; Appel, K. W.; Foroutan, H.; Xu, L.; Ng, N. L.; Goldstein, A. H. Coupling of organic and inorganic aerosol systems and the effect on gas-particle partitioning in the southeastern United States. Atmos. Chem. Phys. 2018, 18, 357-370. (28) Fountoukis, C.; Nenes, A. ISORROPIA II: a computationally efficient thermodynamic equilibrium model for K+-Ca2+-Mg2+-NH4+-Na+-SO42--NO3--Cl--H2O aerosols. Atmos. Chem. Phys. 2007, 7(17), 4639-4659. (29) Battaglia, M. A.; Douglas, S.; Hennigan, C. J. Effect of the urban heat island on aerosol pH. Environ. Sci. Technol. 2017, 51, 13095-13103. (30) Clegg, S. L.; Brimblecombe, P.; Wexler, A. S. Thermodynamic model of the system H+-NH4+-SO42--NO3--H2O at tropospheric temperatures. J. Phys. Chem. A. 1998, 102(12), 2137-2154. (31) Guo, H. Y.; Liu, J.; Froyd, K. D.; Roberts, J. M.; Veres, P. R.; Hayes, P. L.; Jimenez, J. L.; Nenes, A.; Weber, R. J. Fine particle pH and gas–particle phase partitioning of inorganic species in Pasadena, California, during the 2010 CalNex campaign. Atmos. Chem. Phys. 2017c, 17, 5703-5719. (32) Parworth, C. L.; Young, D. E.; Kim, H.; Zhang, X.; Cappa, C. D.; Collier, S.; Zhang, Q. Wintertime water-soluble aerosol composition and particle water content in Fresno, California. J. Geophys. Res. Atmos. 2017, 122, 3155-3170. (33) Murphy, J. G.; Gregoire, P. K.; Tevlin, A. G.; Wentworth, G. R.; Ellis, R. A.; Markovic, M. Z.; VandenBoer, T. C. Observational constraints on particle acidity using measurements and modelling of particles and gases. Faraday Discuss. 2017, 200, 379-395. (34) Yao, X.; Ling, T. Y.; Fang, M.; Chan, C. K. Size dependence of in situ pH in submicron atmospheric particles in Hong Kong. Atmos. Environ. 2007, 41, 382-393. (35) Zhang, Q.; Jimenez, J. L.; Worsnop, D. R.; Canagaratna, M. A case study of urban particle acidity and its influence on secondary organic aerosol. Environ. Sci. Technol. 2007, 41, 3213-3219. (36) Nenes, A.; Pandis, S. N.; Pilinis, C. ISORROPIA: A new thermodynamic equilibrium model for multiphase multicomponent inorganic aerosols. Aquat. Geochem. 1998, 4(1), 123-152. (37) Fountoukis, C.; Nenes, A.; Sullivan, A.; Weber, R.; Van Reken, T.; Fischer, M.; Matias, E.; Moya, M.; Farmer, D.; Cohen, R. C. Thermodynamic characterization of Mexico City aerosol during MILAGRO 2006. Atmos. Chem. Phys. 2009, 9(6), 2141- 2156. (38) Nowak, J. B.; Huey, L. G.; Russell, A. G.; Tian, D.; Neuman, J. A.; Orsini, D.; Sjostedt, S. J.; Sullivan, A. P.; Tanner, D. J.; Weber, R. J.; Nenes, A.; Edgerton, E.; Fehsenfeld, F. C. Analysis of urban gas phase ammonia measurements from the 2002 Atlanta Aerosol Nucleation and Real-Time Characterization Experiment (ANARChE). J. Geophys. Res. 2006, 111(D17), D17308; DOI 10.1029/2006jd007113. (39) Hennigan, C. J.; Izumi, J.; Sullivan, A. P.; Weber, R. J.; Nenes, A. A critical evaluation of proxy methods used to estimate the acidity of atmospheric 26


Page 26 of 35

Page 27 of 35

602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650


particles. Atmos. Chem. Phys. 2015, 15(5), 2775-2790. (40) Song, S. J.; Gao, M.; Xu, W. Q.; Shao, J. Y.; Shi, G. L.; Wang, S. X.; Wang, Y. X.; Sun, Y. L.; McElroy, M. B. Fine-particle pH for Beijing winter haze as inferred from different thermodynamic equilibrium models. Atmos. Chem. Phys. 2018, 18, 7423-7438. (41) Paatero, P. The multilinear engine-a table-driven, least squares program for solving multilinear problems, including the n-way parallel factor analysis model. J. Comput. Graph. Stat. 1999, 9, 854–888. (42) Paatero, P. End User’s Guide to Multilinear Engine Applications. 2007. (43) Amato, F.; Pandolfi, M.; Escrig, A.; Querol, X.; Alastuey, A.; Pey, J.; Perez, N.; Hopke, P. K. Quantifying road dust resuspension in urban environment by Multilinear Engine: A comparison with PMF2. Atmos. Environ. 2009, 43, 2770-278. (44) Amato, F.; Hopke, P. K. Source apportionment of the ambient PM2.5 across St. Louis using constrained positive matrix factorization. Atmos. Environ. 2012, 46, 329-337. (45) Shi, G. L.; Liu, J. Y.; Wang, H. T.; Tian, Y. Z., Wen, J.; Shi, X. R.; Feng, Y. C.; Ivey, C. E.; Russell, A. G. Source apportionment for fine particulate matter in a Chinese city using an improved gas-constrained method and comparison with multiple receptor models. Environ.Pollut. 2018, 233, 1058-1067. (46) Pant, P.; Harrison, R. M. Critical review of receptor modelling for particulate matter: a case study of India. Atmos. Environ. 2012, 49, 1-12. (47) Tian, Y. Z.; Liu, J. Y.; Han, S. Q.; Shi, X. R.; Shi, G. L.; Xu, H.; Yu, H. F.; Zhang, Y. F.; Feng, Y. C.; Russell, A. G. Spatial, seasonal and diurnal patterns in physicochemical characteristics and sources of PM2.5 in both inland and coastal regions within a megacity in China. J. Hazard. Mater. 2018, 342, 139-149. (48) Tian, Y. Z.; Shi, G. L.; Han, B.; Wu, J. H.; Zhou, X. Y.; Zhou, L. D.; Zhang, P.; Feng, Y. C. Using an improved source directional apportionment method to quantify the PM2.5 source contributions from various directions in a megacity in China. Chemosphere. 2015, 119, 750-756. (49) Waked, A.; Favez, O.; Alleman, L. Y.; Piot, C.; Petit, J. E.; Delaunay, T.; Verlinden, E.; Golly, B.; Besombes, J. L.; Jaffrezo, J. L.; Leoz-Garziandia, E. Source apportionment of PM10 in a north-western Europe regional urban background site (Lens, France) using positive matrix factorization and including primary biogenic emissions. Atmos. Chem. Phys. 2015, 14, Waked-3346. (50) Wang, G.; Zhang, R.; Gomez, M. E.; Yang, L.; Levy Zamora, M.; Hu, M.; Lin, Y.; Peng, J.; Guo, S.; Meng, J.; Li, J.; Cheng, C.; Hu, T.; Ren, Y.; Wang, Y.; Gao, J.; Cao, J.; An, Z.; Zhou, W.; Li, G.; Wang, J.; Tian, P.; Marrero-Ortiz, W.; Secrest, J.; Du, Z.; Zheng, J.; Shang, D.; Zeng, L.; Shao, M.; Wang, W.; Huang, Y.; Wang, Y.; Zhu, Y.; Li, Y.; Hu, J.; Pan, B.; Cai, L.; Cheng, Y.; Ji, Y.; Zhang, F.; Rosenfeld, D.; Liss, P. S.; Duce, R. A.; Kolb, C. E.; Molina, M. J. Persistent sulfate formation from London Fog to Chinese haze. P. Natl. Acad. Sci. USA. 2016, 113, 13630-13635. (51) Eddingsaas, N. C.; VanderVelde, D. G.; Wennberg, P. O. Kinetics and products of the acid-catalyzed ring-opening of atmospherically relevant butyl epoxy alcohols. J. Phys. Chem. A. 2010, 114(31), 8106-8113. (52) Sandford, R. C.; Exenberger, A.; Worsfold, P. J. Nitrogen cycling in natural waters using in situ, reagentless UV spectrophotometry with simultaneous determination of nitrate and nitrite. Environ. Sci. Technol. 2007, 41(24), 27



651 652 653 654 655 656 657 658 659 660 661 662 663 664 665

8420-8425. (53) Guo, H. Y.; Xu, L.; Bougiatioti, A.; Cerully, K. M.; Capps, S. L.; Hite Jr., J. R.; Carlton, A. G.; Lee, S. -H.; Bergin, M. H.; Ng, N. L.; Nenes, A.; Weber, R. J. Fine-particle water and pH in the southeastern United States. Atmos. Chem. Phys. 2015, 15, 5211-5228. (54) Marmur, A.; Unal, A.; Mulholland, J. A.; Russell, A. G. Optimization-Based source apportionment of PM2.5 incorporating gas-to-particle ratios. Environ. Sci. Technol. 2005, 39, 3245-3254. (55) Jamieson, L. E.; Jaworska, A.; Jiang, J.; Baranska, M.; Harrison, D.; Campbell, C. Simultaneous intracellular redox potential and pH measurements in live cells using SERS nanosensors. Analyst. 2015, 140(7), 2330-2335. (56) Wei, H. R.; Vejerano, E. P.; Leng, W. N.; Huang, Q. S.; Willner, M. R.; Marr, L. C.; Vikesland, P. J. Aerosol microdroplets exhibit a stable pH gradient. P. Natl. Acad. Sci. USA. 2018, 115(28), 7272-7277.

28


Page 28 of 35

Page 29 of 35


mass concentration(μg/m3)

666 NO3-

120

NH4+

5

(a)

100 4

pH

80 60

3

40 20

2

T(K)

304

300

20

15

/8

/2

/2 /8 15 20

15 20

4

0

6 /8

/1

/1 /8 15 20

20

15

/8

/8

/4 /8 15 20

/3

2

296

/7 15

308

(b)

1

RH(%)

0 90 85 80 75 70 65 60 55 50 45

20

667

SO42-

668

Figure 1. Time series (daily average) of (a) mass concentrations of the major PM2.5

669

species (NO3-, SO42-, NH4+) (μg/m3) and aerosol water pH; (b) temperature and RH.

670

29



25 4.0

16

3.5

14

pH

18

3.0

12 10 8

2.5 0

2

4

6

8

(b)

HNO3

pH

NO3-

ε (NO3-)

1.0 4.0

20

0.8

15

0.6

3.5

10

0.4

3.0

5

0.2

0

0.0

10 12 14 16 18 20 22 24

pH

(a)

ε (NO3-)

SO42pH

20

mass concentration (μg/m3)

mass concentration (μg/m3)

NH4+

Page 30 of 35

2.5 0

2

4

6

8

time

10 12 14 16 18 20 22 24

time RH Water

70

(c)

60

RH

50 40 30 20 10 0

671

5

10

15

20

25

time

672

Figure 2. Diurnal patterns of pH, SNA (sulfate, nitrate, ammonium) and HNO3

673

concentrations, ε(NO3-) (NO3-/(HNO3+ NO3-)), water content and RH. ε(NO3-) was

674

calculated using measured HNO3 (μg/m3) and NO3- concentrations (μg/m3). (a) pH,

675

SO42- and NH4+ concentrations (μg/m3); (b) HNO3 and NO3- concentrations (μg/m3),

676

ε(NO3-). Note that SO42-, NO3-, NH4+, NO3-, and HNO3 were all measured

677

concentrations. (c) RH (%) and water content (μg/m3, estimated by ISORROPIA II).

678

30


Page 31 of 35


secondary nitrate contribution(μg/m3)

time 23 ~ 00 23:00-00:00 22 ~ 23 22:00-23:00 21 ~ 22 21:00-22:00 20 ~ 21 20:00-21:00 19 ~ 20 19:00-20:00 18:00-19:00 18 ~ 19 17 ~ 18 17:00-18:00 16:00-17:00 16 ~ 17 15:00-16:00 15 ~ 16 14:00-15:00 14 ~ 15 13:00-14:00 13 ~ 14 12:00-13:00 12 ~ 13 11:00-12:00 11 ~ 12 10:00-11:00 10 ~ 11 9:00-10:00 09 ~ 10 8:00-9:00 08 ~ 09 7:00-8:00 07 ~ 08 6:00-7:00 06 ~ 07 5:00-6:00 05 ~ 06 4:00-5:00 04 ~ 05 3:00-4:00 03 ~ 04 2:00-3:00 02 ~ 03 1:00-2:00 01 ~ 02 00 ~ 01 0:00-1:00

150

A

100

B

50

0

D

C

2

4

6

pH

8

pH-original

679 680

Figure 3. Relationships between aerosol pH and secondary nitrate (SN) contribution.

681

Samples in A-D were collected from different time periods, and were characterized

682

with

683

meteorological conditions are provided in Figure S7.

different

meteorological

conditions.

Detailed

684

31


information

related

to

Secondary nitrate contribution (μg/m3)


(a)

(b)

(c)

(d)

pH

pH

685

Page 32 of 35

686

Figure 4. The secondary nitrate contribution (SN) as a function of pH. Circles were

687

color-coded by sources contributions: (a) Secondary sulfate contribution (μg/m3), (b)

688

vehicle contribution (μg/m3), (c) coal combustion (μg/m3), (d) dust contribution

689

(μg/m3).

690

32


Page 33 of 35


1.0 1.0 0.8

0.6

0.6

-

e(NO3 )

-

e(NO3 )

0.8

0.4

y  1

0.2

1 x  1.99 1  exp( ) 0.65

0.4 Ambient atmospheric samples

0.2

Modeled samples calculated by Eq. (4)

0.0 0.0

-2

691

0

2

pH

4

6

8

-2

0

2

pH

4

6

8

692

Figure 5. S curve constructed by fitting ε(NO3-) as a function of pH using the

693

Boltzmann equation for summer samples (R2=0.70). Black points are actual

694

atmospheric samples with higher SS contribution (higher than 10 μg m-3); red points

695

are calculated by the Boltzmann equation. ε(NO3-) was calculated using measured

696

NO3- concentration in particle phase and measured HNO3 (g) concentration in gas

697

phase.

698

33



(a)

Page 34 of 35

(c)

ε(NO3-)

(b)

pH

pH

pH (e)

(f)

ε(NO3-)

(d)

699

pH

pH

pH

700

Figure 6. ε(NO3-) as a function of pH under different conditions (a) RH (%); (b)

701

Temperature (T, oC); (c) Water content (μg/m3); (d) Dust contribution (μg/m3); (e)

702

Coal contribution (μg/m3); and (f) Vehicle contribution (μg/m3). ε(NO3-) was

703

calculated using measured NO3- concentration in particle phase and measured HNO3

704

(g) concentration in gas phase.

705

34


Page 35 of 35


0.9

 (NO3-)

0.8

1.0

Spring

Winter

0.9

Sample with SS 3 higher than 25 g/m

Sample with SS 3 higher than 25 g/m

0.8

 (NO3-)

1.0

0.7 0.6

1 x  2.8 ) 0.71

y  1

1  exp(

0.5

0.6 0.5 0.4

R2=0.67

0.4

0.7

0.3 0.2

0.3

0.1

0.2

0.0

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

0

706

2

4

6

8

pH

pH

707

Figure 7. ε(NO3-) vs. pH for spring and winter samples when SS was greater than 25

708

μg m-3

709

35


High-resolution datasets unravel the effects of sources and

Recommend Documents