Characterization of Chromophoric Water-Soluble ... - ACS Publications

Aug 11, 2016 - ... and Marine Aerosols by HR-ToF-AMS Analysis and Excitation–Emission ... View: ACS ActiveView PDF | PDF | PDF w/ Links | Full Text ...
7 downloads 0 Views 2MB Size
Subscriber access provided by Northern Illinois University

Article

Characterization of chromophoric water-soluble organic matter in urban, forest and marine aerosols by HR-ToFAMS analysis and excitation–emission matrix spectroscopy Qingcai Chen, Yuzo Miyazaki, Kimitaka Kawamura, Kiyoshi Matsumoto, Sean C. Coburn, Rainer Volkamer, Yoko Iwamoto, Sara Kagami, Yange Deng, Shuhei Ogawa, Sathiyamurthi Ramasamy, Shungo Kato, Akira Ida, Yoshizumi Kajii, and Michihiro Mochida Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b01643 • Publication Date (Web): 11 Aug 2016 Downloaded from http://pubs.acs.org on August 12, 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 34

Environmental Science & Technology

1

Characterization of chromophoric water-soluble organic

2

matter in urban, forest and marine aerosols by HR-ToF-

3

AMS analysis and excitation–emission matrix

4

spectroscopy

5

Qingcai Chen1, Yuzo Miyazaki 2, Kimitaka Kawamura2,3, Kiyoshi Matsumoto4, Sean Coburn5,6,7,

6

Rainer Volkamer5,6, Yoko Iwamoto8,9, Sara Kagami1, Yange Deng1, Shuhei Ogawa1,

7

Sathiyamurthi Ramasamy10, Shungo Kato11, Akira Ida10,12, Yoshizumi Kajii10,13, Michihiro

8

Mochida*1

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

1

Graduate School of Environmental Studies, Nagoya University, Nagoya 464-8601, Japan Institute of Low Temperature Science, Hokkaido University, Sapporo 060-0819, Japan 3 Now at Chubu Institute for Advanced Studies, Chubu University, Kasugai 487-8501, Japan 4 Interdisciplinary Graduate School of Medicine and Engineering, University of Yamanashi, Kofu 400-8510, Japan 5 Department of Chemistry & Biochemistry, University of Colorado, Boulder, CO 80309-0215, USA 6 Cooperative Institute for Research in Environmental Sciences (CIRES), University of Colorado, Boulder, CO 80309-0215, USA 7 Now at Department of Mechanical Engineering, University of Colorado, Boulder, CO 803090215, USA 8 Institute for Advanced Research, Nagoya University, Nagoya 464-8601, Japan 9 Now at Faculty of Science Division I, Tokyo University of Science, Tokyo 162-8601, Japan 10 Graduate School of Global Environmental Studies, Kyoto University, Kyoto 606-8501, Japan 11 Department of Applied Chemistry, Tokyo Metropolitan University, Tokyo 192-0397, Japan 12 Now at Seraku Co. Ltd., Tokyo 160-0023, Japan 13 Center for Regional Environmental Research, National Institute for Environmental Studies, Tsukuba 305-8506, Japan 2

27

*Corresponding author e-mail: [email protected]

28

Phone/fax: 052-788-6157

29

Address: Department of Earth and Environmental Sciences, Graduate School of Environmental

30

Studies, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8601, Japan

1 Environment ACS Paragon Plus

Environmental Science & Technology

31

ABSTRACT: Chromophoric water-soluble organic matter in atmospheric aerosols potentially

32

plays an important role in aqueous reactions and light absorption by organics. The fluorescence

33

and chemical-structural characteristics of the chromophoric water-soluble organic matter in

34

submicron aerosols collected in urban, forest and marine environments (Nagoya, Kii Peninsula,

35

and the tropical Eastern Pacific) were investigated using excitation-emission matrices (EEMs)

36

and a high-resolution aerosol mass spectrometer. Three types of water-soluble chromophores,

37

two with fluorescence characteristics similar to those of humic-like substances (HULIS-1 and

38

HULIS-2) and one with fluorescence characteristics similar to those of protein compounds

39

(PLOM), were identified in atmospheric aerosols by parallel factor analysis (PARAFAC) for

40

EEMs. We found that the chromophore components of HULIS-1 and -2 were associated with

41

highly and less-oxygenated structures, respectively, which may provide a clue to understanding

42

the chemical formation or loss of organic chromophores in atmospheric aerosols. Whereas

43

HULIS-1 was ubiquitous in water-soluble chromophores over different environments, HULIS-2

44

was abundant only in terrestrial aerosols and PLOM was abundant in marine aerosols. These

45

findings are useful for further studies regarding the classification and source identification of

46

chromophores in atmospheric aerosols.

47

KEYWORDS: chromophores, organic aerosol, chemical characteristics, aerosol mass

48

spectrometry, excitation-emission matrices

49 50 51 52

2 Environment ACS Paragon Plus

Page 2 of 34

Page 3 of 34

53

Environmental Science & Technology

INTRODUCTION

54

Water-soluble organic matter (WSOM) constitutes a large fraction of the organics in

55

submicron aerosols in the atmosphere and can affect the role of aerosols in terms of climate

56

processes and human health.1-6 Chromophoric water-soluble organic matter is a component of

57

WSOM that has the ability to absorb sunlight at visible and near-ultraviolet (UV) wavelengths

58

and takes part in photoinitiated reactions. The chromophoric water-soluble organic matter

59

therefore potentially plays important roles in some aqueous reactions of organics in aerosols and

60

thus in the radiative balance of Earth.7-10 Although the origin, composition and chemistry of

61

WSOM in aerosols have attracted great attention over the past decade, the characteristics of the

62

chromophoric fraction of water-soluble organic matter are poorly understood. 1-6

63

Brown carbon in atmospheric aerosols, which contain chromophoric water-soluble organic

64

matter, is primarily emitted from biomass burning and is formed secondarily.11-13 Chromophoric

65

organic matter in aerosol can be formed via a variety of reaction pathways. Nitrophenols are

66

major chromophores in chamber-generated secondary organic aerosols from toluene

67

photooxidation in the presence of NOx.14 The reactions of carbonyl compounds with amines and

68

ammonia can form nitrogen-containing chromophoric organics.15-20 In view of its role in the

69

atmosphere, an important characteristic of chromophoric organic matter is its sensitivity to light.

70

Photoinitiated reactions with chromophoric organic matter include both degradation and

71

polymerization.21-23 Substantial photodegradation of chromophoric organic matter in rainwater

72

and laboratory-generated solutions containing brown carbon occurs when the matter is exposed

73

to simulated sunlight.21,

74

excited states through the capture of the excitation energy from photons and thus plays a

24, 25

Chromophoric organic matter has the potential to be in triplet

3 Environment ACS Paragon Plus

Environmental Science & Technology

75

significant role in aqueous-phase oxidation of organics.8, 10 The triplet state of 1-nitronaphthalene

76

reacts with phenol to form humic-like substances (HULIS).26

77

Excitation-emission matrix (EEM) fluorescence spectroscopy is a commonly used technique

78

to investigate the optical and structural characteristics of chromophores that are responsible for

79

the light absorption and fluorescence by complex organic matter. Although the method has been

80

widely applied to the characterization of chromophoric water-soluble organic matter in terrestrial

81

and oceanic systems,27,

82

Chromophore components associated with HULIS and protein-like organic matter (PLOM) have

83

been statistically determined for complex organic matter in terrestrial and oceanic systems as

84

well as atmospheric aerosols by parallel factor analysis (PARAFAC) for EEMs. Although

85

chromophoric water-soluble organic matter in terrestrial and oceanic systems are relatively well

86

characterized by EEMs, the results of the origins and chemical structures of PARAFAC

87

components may not apply to organic matter in the atmosphere because the sources and

88

underlying physicochemical processes are different.30, 32 The organic chromophores in different

89

types of aerosols (e.g., urban, forest and ocean aerosols) may also be different, which must result

90

in differences in the optical properties and the photochemical reactivity of the organic

91

components. Investigating the relationships between the optical properties and structural

92

characteristics of WSOM in different types of aerosols is also important to better understand the

93

structures, sources and chemistry of chromophoric water-soluble organic matter in atmospheric

94

aerosols.

28

it has not been extensively used in atmospheric aerosol research.29-31

95

Chromophoric water-soluble organic matter in atmospheric aerosols may contain an array of

96

chemical species, as in the case of WSOM.33 For the chemical characterization of such complex

97

matter, high-resolution aerosol mass spectrometers (HR-AMSs) can be used in combination with

4 Environment ACS Paragon Plus

Page 4 of 34

Page 5 of 34

Environmental Science & Technology

98

the nebulization technique,34,

35

99

complex organic matter, e.g., elemental composition and ion groups of the mass spectra. In this

100

study, EEMs and HR-AMS analyses were performed for WSOM in urban, forest and marine

101

aerosols to characterize the chromophoric water-soluble organic matter therein. Statistical

102

analyses of PARAFAC and non-negative matrix factorization (NMF) were applied to EEMs and

103

HR-AMS spectra to resolve the different components of WSOM that exhibit different chemical-

104

structural and fluorescence characteristics. The chemical structures and possible sources of

105

chromophoric water-soluble organic matter in different types of submicron aerosols are analyzed

106

and discussed based on hierarchical cluster analysis for the WSOM samples, backward air mass

107

trajectory analysis, and correlation analysis using the relative intensity of ion groups from the

108

HR-AMS spectra. Application of the results to future studies is suggested.

109

EXPERIMENTAL SECTION

110

Collection of aerosol samples and WSOM extraction

which provides the overall structural characteristics of the

111

This study used sixty-three PM0.95 samples (50% cutoff: 0.95 µm) collected on 8 × 10 inch

112

quartz fiber filters using high-volume samplers in three different environments. Twenty-seven

113

samples were collected at an urban site in Nagoya, Japan (Nagoya University, 35.15°N,

114

136.97°E) from 11 August to 14 September 2013 (sample ID: 1–17), and from 4 to 24 March

115

2014 (sample ID: 18–27). Sixteen samples were collected at a forest site on the Kii Peninsula,

116

Japan (Wakayama Forest Research Station (WFRS), Kyoto University, 34.06°N, 135.52°E), five

117

from 20 to 30 August 2010 (sample ID: 28–32), and eleven from 28 July to 8 August 2014

118

(sample ID: 33–43). Twenty samples were collected aboard the NOAA RV Ka’imimoana as part

119

of the Tropical Ocean tRoposphere Exchange of Reactive halogens and Oxygenated VOCs

120

(TORERO) field experiment over the eastern equatorial Pacific ocean during the TORERO/KA-

5 Environment ACS Paragon Plus

Environmental Science & Technology

121

12-01 cruise in February 2012 (sample ID: 44–63). Blank filters were also collected at different

122

sites and periods. More information on the samples and the environment of the sampling sites are

123

given in Table S1 and section S2 of the Supporting Information (SI), respectively.

124

WSOM for the EEM analysis and HR-AMS analysis was extracted via the ultrasonication of

125

filter punches (diameter: 34 mm) with Milli-Q water for 1 h, followed by filtration through 0.2

126

µm PTFE filters (Millex). Different extraction conditions, i.e., the number of filter punches

127

(diameter: 34 mm) and the mass of Mill-Q water, were applied to different samples, as

128

summarized in Table S1. Quantification of WSOC was performed using a TOC analyzer (Model

129

TOC-VCHS, Shimadzu, Japan), as detailed in the SI (section S3). Analyses of amino acids and

130

biogenic secondary organic aerosol (BSOA) tracers were also performed for urban and forest

131

samples, respectively (see section S3, SI).

132

HR-AMS measurements

133

Each extract was atomized with argon (purity: 99.9999%). The generated aerosol was passed

134

through two diffusion scrubbers to remove water vapors. Then, the aerosol was introduced to the

135

HR-AMS (HR-ToF-AMS, Aerodyne Research Inc., Billerica, MA, USA). The HR-AMS spectra

136

were acquired in V-mode, and the data were analyzed using Squirrel v.1.56A and Pika v1.15A

137

Software (http://cires.colorado.edu/jimenez-group/ToFAMSResources/ToFSoftware/). In the

138

analysis using Pika, all of the CO+ ion signals were considered to have originated from organics

139

because pure argon was used as the carrier gas (Figure S1a). All of the organic ions were

140

classified into CH, CHN, CHO1, CHOgt1, CHO1N, CHOgt1N, CS, Cx, HO, CO, and CO2 groups.

141

Elemental analyses were performed to determine the atomic ratios (O/C, H/C, and N/C), and the

142

mass ratio of organic matter (OM) to organic carbon (OC). N/C was derived according to Aiken

143

et al,36 and O/C, H/C, and OM/OC were derived according to Canagaratna et al.37

6 Environment ACS Paragon Plus

Page 6 of 34

Page 7 of 34

144 145

Environmental Science & Technology

UV-visible absorption spectra and EEM fluorescence spectra The extracts were placed in a 1-cm path-length quartz cell and subjected to analysis using a

146

UV-visible

147

spectrophotometer (Model FP-6600, JASCO, Japan). A UV-visible absorption spectrum was

148

recorded in the range of 200 to 800 nm. Before the scans for a series of extracts, Milli-Q water

149

was also analyzed for a baseline correction. The mass absorption efficiency (MAE, m2 g–1 OC)

150

of the organics in the extracts at λ nm (MAEλ) was calculated using the following equation: 38

151

spectrophotometer

(Model

V-570,

JASCO,

Japan)

and

MAEλ = Absλ ln(10)/CWSOC,

a

fluorescence

(1)

152

where Absλ is the light absorption coefficients (m–1) of the extract solutions at λ, and CWSOC is the

153

concentration of water-soluble organic carbon (WSOC, g m-3) in the extracts. Because the MAE

154

of the extracts from the forest and marine samples were low in the visible range (Figure S2), the

155

MAE at 266 nm (MAE266) was used as a measure of absorptivity of the extracts.

156

The EEMs were measured in the range of 200 to 400 nm for excitation and in the range of 250

157

to 650 nm for emission at a scan speed of 1000 nm min–1. The band-passes for excitation and

158

emission were both set as 10 nm.22 The EEMs were calibrated according to the procedures

159

described by Lawaetz and Stedmon, and Murphy. 39, 40 The EEMs were normalized according to

160

the Raman peak of water, so the data are reported in Raman units (RU).39 The fluorescence

161

index, biological index, and humification index, which are the ratios of the fluorescence intensity

162

at excitation/emission wavelengths (Ex/Em) of 370/470 nm to that at 370/520 nm, the intensity

163

at 310/380 nm to that at 310/430 nm, and the intensity at 255/434–480 nm (average of the values

164

on the line between 255/434 and 255/480 nm) to that at 255/300–344 nm (average of the values

165

on the line between 255/300 and 255/344 nm), respectively, were obtained from the EEMs.41, 42

166

To compare the fluorescence efficiency of different water-extract samples, the fluorescence

7 Environment ACS Paragon Plus

Environmental Science & Technology

167

intensity over excitation wavelengths of 200–400 nm and emission wavelengths of 250–650 nm

168

were normalized by the OC concentration of the extracts (abbreviated to NFV, RU-nm2 [mg L−1

169

OC]–1).

170

Factor analysis for EEMs and HR-AMS spectra

171

PARAFAC analysis with non-negativity constraints of EEMs was performed with the drEEM

172

toolbox version 0.2.0 for MATLAB (http://www.models.life.ku.dk/dreem).27 For the analysis,

173

the EEM data at the excitation wavelengths longer than or equal to those on the line between

174

Ex/Em of 200/570 nm and that of 235/650 nm were omitted. The three-component PARAFAC

175

model was selected from two- to ten-component PARAFAC models because the residual errors

176

decreased noticeably when the number of components increased from two to three (Figure S3).

177

In addition, the three-component model has reasonable spectra of fluorescent components, and it

178

passes the assessment of a split-half analysis with the split style of ‘S4C6T3’ based on the

179

PARAFAC tutorial (Figure S4).27 The seven-component PARAFAC solution, which also

180

exhibits a noticeable decrease in residual errors with the increase in the number of components

181

among four- to ten-component PARAFAC solutions (Figures S3), was used to obtain more

182

detailed information on the chromophores. The seven-component PARAFAC solution passes the

183

split analysis with the split style of ‘S3C3T3’ (Figure S5), indicating that the solution is stable.

184

Non-negative matrix factorization (NMF) analysis was performed for the HR-AMS dataset

185

using the NMF toolbox version 1.4 in MATLAB (https://sites.google.com/site/nmftool/).43 First,

186

a gradient-descent-based multiplicative algorithm was used to find solutions from multiple

187

random starting values, and then a least-squares active-set algorithm was used to find final

188

solutions from the first solutions. To find a final solution, the model was run 100 times with a

8 Environment ACS Paragon Plus

Page 8 of 34

Page 9 of 34

Environmental Science & Technology

189

different starting point each time. The four-factor NMF model was validated by comparisons of

190

residual errors, split analysis and visualization of spectral loadings (Figures S6, S7 and S8).

191

Hierarchical cluster analysis

192

A hierarchical cluster method was performed to classify aerosol samples based on the relative

193

contributions of PARAFAC components and NMF factors to the respective samples. The

194

Euclidean distance was used to evaluate the distances between samples, and the centroid method

195

was chosen for the hierarchical cluster analysis by comparing the multiple correlation

196

coefficients for nearest neighbor, farthest neighbor, median neighbor and Ward’s method. Six

197

clusters were determined for EEMs and HR-AMS data and were named clusters a-f and clusters

198

A-F, respectively (Figures S9 and S10).

199

RESULTS AND DISCUSSION

200

EEMs and PARAFAC components

201

To identify and quantify the underlying chromophore components in aerosols, the PARAFAC

202

model was applied to separate independently varying fluorescent components. Figure 1 presents

203

the EEMs of the three- (3CM-C1~3) and seven- (7CM-C1~7) component PARAFAC solutions.

204

The EEMs of 3CM-C1~3 are similar to those identified from water and alkaline soluble matter in

205

urban aerosols in Granada, Spain and in Aveiro, Portugal.29, 30 Based on a comparison with the

206

EEMs of three fluorescent components defined for terrestrial and oceanic organic matter by Yu

207

et al., 28 the chemical components corresponding to 3CM-C1, -C2 and -C3 are named HULIS-1,

208

HULIS-2 and PLOM, respectively, in this study. Note that the origins and chemical structures of

209

HULIS and PLOM chromophores studied are not necessarily similar to those of the

9 Environment ACS Paragon Plus

Environmental Science & Technology

210

chromophores with same names in terrestrial and oceanic organic matter. The origins and

211

chemical structures of the chromophores in atmospheric aerosols are discussed below.

Page 10 of 34

212

The seven-component PARAFAC solution provides more detailed information on the

213

chromophores. By comparing the emission wavelengths with the maximum fluorescence of the

214

three-component PARAFAC solution and based on a correlation analysis of the relative

215

abundance of the three and seven components (Table S2), chemical components corresponding

216

to 7CM-C1 and 7CM-C3 are assigned to HULIS-1, whereas 7CM-C2, 7CM-C4 and 7CM-C5 are

217

assigned to HULIS-2, and 7CM-C6 and 7CM-C7 are assigned to PLOM. Among the seven

218

components, 7CM-C1, 7CM-C3 and 7CM-C5 have peaks at emission wavelengths of >400 nm

219

(Figure 1); these emissions probably originate from conjugated systems, e.g., via quinonoid n-π*

220

transitions.44 This claim is supported by the result that the emission wavelength of 7CM-C3

221

(~470 nm) is similar to that of the quinone-like component identified in aerosol WSOM in

222

Granada, Spain.29 Based on the definition of fluorescent components of terrestrial and oceanic

223

organic matter, 7CM-C6 and 7CM-C7 correspond to tryptophan-like and tyrosine-like

224

components, respectively.27,

225

major portion of the chromophores of 7CM-C6 and 7CM-C7 in atmospheric aerosols, as

226

discussed below.

28

However, non-nitrogen-containing species may account for a

227

The relative intensities of different fluorescent components in different types of aerosols were

228

highly variable depending on the sampling site and period, as shown in Figure 2a. This result

229

suggests that the chemical composition of chromophoric water-soluble organic matter differs

230

markedly from sample to sample. HULIS-1was the substance with most intense fluorescence in

231

all samples, and, on average, accounted for 46% of the total fluorescence intensity of all samples.

232

High mass fractions of HULIS were observed in an area similar to that of the marine aerosol

10 Environment ACS Paragon Plus

Page 11 of 34

Environmental Science & Technology

233

sampling, supporting the particularly intense fluorescence of HULIS-1 in the marine aerosols

234

studied.47 The fluorescence of HULIS-2 was intense in the terrestrial aerosols (urban and forest

235

aerosols) (mean: 47% of the total fluorescence intensity of the extracts), particularly in forest

236

aerosols (mean: 51%). Conversely, the fluorescence of HULIS-2 in marine aerosols was weak

237

(mean: 23%). This result suggests that HULIS-2 primarily has a terrestrial origin. The marine

238

aerosols were enriched in the PLOM (mean: 19%), which can be attributed to marine biological

239

materials.28, 46 The fluorescent components assigned to PLOM were also present in the urban

240

aerosols, among which the fluorescence of tryptophan-like substances (7CM-C6) was the

241

strongest. Accordingly, these aerosols differ from the marine aerosols, in which the relative

242

strengths of the fluorescences of tryptophan-like and tyrosine-like components (7CM-C7) were

243

nearly equal. This result indicates that PLOM in urban and marine aerosols can be attributed to

244

different chemical structures and sources.

245

HR-AMS spectra and the NMF factors

246

The chemical characteristics of WSOM in aerosols were obtained on the basis of the HR-AMS

247

spectra of the extracts in various samples. Four factors were determined from the HR-AMS

248

spectra of the NMF model. The mass spectra and relative abundances of these factors in different

249

types of aerosols are presented in Figures 3 and 2b, respectively. Based on the degree of

250

oxidation (O/C ratio) of organics, the four factors were identified as highly oxygenated organic

251

aerosols (HOOA-1 and HOOA-2) and low oxygenated organic aerosols (LOOA-1 and LOOA-2).

252

Note that each organic aerosol may represent a group of molecules and/or their moieties.

253

The HOOA-1 factor has a high O/C ratio (1.67) and a low H/C ratio (1.1). Approximately half

254

of the total HOOA-1 signal was contributed by CO2+ ions, suggesting a large fraction of

255

carboxylic acids in HOOA-1.37, 47 The mass fraction of HOOA-1 in the marine aerosols (mean:

11 Environment ACS Paragon Plus

Environmental Science & Technology

Page 12 of 34

256

48%) was obviously higher than those in the urban (mean: 20%) and forest aerosols (mean: 9%),

257

indicating that WSOM in marine aerosols was dominated by more processed oxygenated

258

products.

259

The HOOA-2 factor contains abundant CS ions (16%), primarily CHSO+ (m/z 61), CH2SO+

260

(m/z 62), CH2SO2+ (m/z 78), CH3SO2+ (m/z 79) and CH4SO3+ (m/z 96) ions (Figures 3 and S1).

261

Among these ions, the most abundant CS ion in the mass spectrum of HOOA-2 is CH2SO2+,

262

which likely originated from methanesulfonic acid (CH3SO3H, MSA).48, 49 HOOA-2 contributes

263

largely to the WSOM of marine aerosols (mean: 30%) and slightly to the WSOM of the urban

264

(mean: 2%) and forest samples (mean: 5%). The relative abundance of HOOA-2 is particularly

265

high in marine samples collected over the eastern tropical Pacific Ocean in the Southern

266

Hemisphere (mean: 46%), which is consistent with the high mass concentration ratios of MSA to

267

WSOC.50 The above results suggest that HOOA-2 should be representative of the WSOM of

268

marine aerosols. The characteristics of the low abundances of CxHyOz+ (10%) ions and high

269

abundances of CxOy+ (31%) and CxHy+ (31%) ions in HOOA-2 differ from those of the PMF-

270

derived marine organic aerosols in Paris, France (CxHyOz+, CxOy+ and CxHy+ were 41%, 17% and

271

32%,

272

difference may be because marine aerosols are enriched in water-insoluble organic matter

273

(WISOM), which may consist of hydrophobic sugar-like compounds.52 In fact, the WSOC/TC

274

ratio was low (mean: 0.4), thus supporting the high abundance of WISOM in the marine

275

aerosols.50

respectively)

(URL:

http://cires.colorado.edu/jimenez-group/HRAMSsd/).51

This

276

The LOOA-1 factor was dominated by CxHy+ (38%) and CxHyOz+ (24%) ion groups. In

277

contrast, CO2+ was very low in the LOOA-1 spectra, but abundant CO+ (15%) was observed.

278

This result suggests that LOOA-1 is associated with alcohols and ketones or aldehydes, but not

12 Environment ACS Paragon Plus

Page 13 of 34

Environmental Science & Technology

279

with carboxylic acids and carboxylate esters.37 One of the remarkable characteristics of LOOA-1

280

in comparison with the other factors is the higher mass fraction of the CxHyNq+ ion group (6%),

281

suggesting that nitrogen-containing compounds were abundant. The relative abundances of

282

LOOA-1 in certain urban aerosol samples, particularly in the late winter and early spring

283

samples (mean: 47%), were obviously higher than those in the forest and marine aerosol samples

284

(Figure 2b). Hence, LOOA-1 is probably anthropogenic.

285

The mass spectrum of the LOOA-2 is largely composed of CxHy+ (44%) and CxHyOz+ (31%)

286

ion groups and is characterized by the higher signal of C2H3O+ (11%). The profile of the LOOA-

287

2 factor was similar to that of the factor of semi-volatile oxygenated organic aerosols,

288

determined on the basis of the PMF analysis of the online HR-AMS data at the WFRS forest site,

289

which was interpreted as locally formed BSOA.53 The relative abundance of LOOA-2 was

290

positively correlated with the ratios of total OC concentrations of BSOA tracers (supporting

291

information S3) to WSOC in the forest samples (r = 0.80, p