Hygroscopicity of Organic Compounds as a Function of Carbon Chain

Jun 16, 2017 - The albedo and microphysical properties of clouds are controlled in part by the hygroscopicity of particles serving as cloud condensati...
0 downloads 1 Views 2MB Size
Article

Hygroscopicity of Organic Compounds as a Function of Carbon Chain Length, Carboxyl, Hydroperoxy, and Carbonyl Functional Groups Sarah Suda Petters, Demetrios Pagonis, Megan Suzanne Claflin, Ezra J. T. Levin, Markus D. Petters, Paul J. Ziemann, and Sonia M. Kreidenweis J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b04114 • Publication Date (Web): 16 Jun 2017 Downloaded from http://pubs.acs.org on June 19, 2017

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.

The Journal of Physical Chemistry A 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 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

1

Hygroscopicity of Organic Compounds as a

2

Function of Carbon Chain Length, Carboxyl,

3

Hydroperoxy, and Carbonyl Functional Groups

4

Sarah Suda Petters1,†,*, Demetrios Pagonis2, Megan S. Claflin2, Ezra J. T. Levin1, Markus D.

5

Petters3, Paul J. Ziemann2, and Sonia M. Kreidenweis1

6

1

Department of Atmospheric Science, Colorado State University, Fort Collins, Colorado 80523-

7

1371, United States 2

8

Department of Chemistry and Biochemistry, University of Colorado at Boulder, Boulder,

9

Colorado 80309-0216, United States 3

10 11 12

Department of Marine Earth and Atmospheric Sciences, North Carolina State University, Raleigh, North Carolina 27695-8208, United States

Draft for submission to the Journal of Physical Chemistry A: June 13, 2017.

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

13

Abstract. The albedo and microphysical properties of clouds are controlled in part by the

14

hygroscopicity of particles serving as cloud condensation nuclei (CCN). Hygroscopicity of

15

complex organic mixtures in the atmosphere varies widely and remains challenging to predict.

16

Here we present new measurements characterizing the CCN activity of pure compounds in

17

which carbon chain length and the number of hydroperoxy, carboxyl, and carbonyl functional

18

groups were systematically varied to establish the contributions of these groups to organic

19

aerosol apparent hygroscopicity. Apparent hygroscopicity decreased with carbon chain length

20

and increased with polar functional groups in the order carboxyl > hydroperoxy > carbonyl.

21

Activation diameters at different supersaturations deviated from the -3/2 slope in log-log space

22

predicted by Köhler theory, suggesting that water solubility limits CCN activity of particles

23

composed of weakly functionalized organic compounds. Results are compared to a functional

24

group contribution model that predicts CCN activity of organic compounds. The model

25

performed well for most compounds but under-predicted the CCN activity of hydroperoxy

26

groups. New best-fit hydroperoxy group/water interaction parameters were derived from the

27

available CCN data. These results may help improve estimates of the CCN activity of ambient

28

organic aerosols from composition data.

2

ACS Paragon Plus Environment

Page 2 of 43

Page 3 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

29

The Journal of Physical Chemistry

Introduction

30

Atmospheric aerosols participate in the formation of clouds by serving as cloud condensation

31

nuclei (CCN),1-2 affect the global radiative balance through direct and indirect effects,3-4 and can

32

be detrimental to the wellbeing of plants and animals.5-6 Numerous organic compounds

33

contribute to the aerosol mass in the atmosphere,7-8 many of which are formed through secondary

34

processes. Secondary organic aerosol (SOA) forms when volatile organic compounds are

35

oxidized in the atmosphere, yielding low-volatility compounds that can condense.9-10 Low-

36

volatility compounds also form in aqueous phase reactions as dissolved organic compounds are

37

oxidized in cloud droplets11-12 or wet aerosols.13-14 Due to the chemical complexity of SOA, the

38

chemical evolution and properties of SOA in the atmosphere are often described by simplified

39

metrics such as the elemental oxygen to carbon ratio (O:C),15-16 the fraction of mass spectral

40

fragment at mass-to-charge ratio 44 (f44),17 the change in H:C ratio relative to the O:C ratio,18 or

41

the average oxidation state of carbon.19 The O:C ratio and f44 are both good predictors of the

42

hygroscopicity of the organic fraction of aerosols in the atmosphere,15-17 but the relationship with

43

hygroscopicity is not linear and may vary greatly under different conditions,20-23 demonstrating

44

the need to include more detailed information such as molecular structure.

45

Many organic compounds present in SOA contribute to the hygroscopicity of the whole

46

particle.24-25 The hygroscopicity of each compound depends on the length of the carbon chain

47

and the number and type of attached functional groups.23 Aerosol water content can be predicted

48

by chemical equilibrium models such as UNIFAC for arbitrary mixtures of organic functional

49

groups in solution.26 UNIFAC predictions of aerosol properties are widely used in the aerosol

50

community27-31 and also serve as the core component of a model we have developed to predict

51

CCN activity from functional group composition and particle dry diameter30 (hereafter referred

3

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 43

52

to as the U-CCN model). UNIFAC is semi-empirical and thus measurements are necessary to

53

constrain predictions.28-29 Most single-component studies characterizing the CCN activity of

54

organic aerosols have been limited to chemicals that are commercially available. Suitable

55

datasets for fitting empirical UNIFAC interaction parameters remain limited due to the difficulty

56

of synthesizing a large suite of compounds with targeted functional groups. Although the U-CCN

57

model was previously validated against a database of published laboratory CCN measurements,

58

that database had limited coverage for many functional groups that are important in atmospheric

59

organic aerosols. This work aims to close the gap for carboxyl, hydroperoxy, and carbonyl

60

functional groups, which are investigated by using organic synthesis to systematically vary the

61

type and location of the functional group on the carbon backbone. The U-CCN model uses a

62

UNIFAC core to predict water activity,26 a functional group prediction of molar volume,32 and a

63

liquid-liquid phase equilibrium algorithm.33 These are combined with traditional Köhler theory

64

to predict CCN activity. The efficacy by which individual functional groups promote CCN

65

activity is strongly tied to the water activity and solubility/miscibility for highly dilute

66

solutions,30 meaning that validation and tuning of UNIFAC group interaction parameters for the

67

U-CCN model must be performed against actual CCN measurements. Therefore this work

68

primarily focuses on measurements of CCN activity, expressed in terms of the compound’s

69

hygroscopicity parameter, to improve the fidelity of the U-CCN model.

70

Theoretical Here the hygroscopicity parameter κ is used together with Köhler theory to describe CCN

71 72

activation (eq 1) and aerosol water uptake in sub-saturated relative humidity (eq 2):24,34  −    ,   = max      1   −  − 

4

ACS Paragon Plus Environment

(1)

Page 5 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

 − 1  ,   =        − 1 − 

73 74 75 76 77

(2)

where  is the critical water vapor supersaturation at CCN activation,  is the droplet diameter,

 is the particle dry diameter,  = 4 ⁄



is the Kelvin diameter,  is the surface

tension of the air/solution interface,  is the molar mass of pure water,  is the universal gas constant,  is temperature,



is the density of pure water,  is the water vapor saturation ratio

(relative humidity expressed as a fraction), and  = ⁄ is the diameter growth factor.

78

Throughout this work  is expressed as a percent: " =  − 1 × 100%. By convention κ

79

values are normalized to a standard state by using the surface tension of pure water and room

80

temperature ( = 0.072 N m-1,  = 298.15 K) in calculations. This κ is denoted as “apparent κ”

81

because it subsumes surface tension and other physicochemical processes to express the

82

observed CCN activation or water uptake. Specifying any two of the variables κ, " ,  or the

83

variables κ, ,  will uniquely determine the third, facilitating intercomparison of values across

84

different studies. Hereafter ‘hygroscopicity’ and ‘κ’ refer to apparent κ unless otherwise

85

specified.

86

Mismatch between κ derived from CCN measurements and κ derived from sub-saturated

87

measurements has been observed35-36 but remains difficult to interpret. Contributing factors

88

include reduced surface tension,36-37 solubility limitations,38 and particle irregularity and surface

89

restructuring during measurements.39 To account for these factors we further differentiate

90

between apparent κ and intrinsic κ, ,-./ , which can be expressed by fully dissolved compounds

91

in the absence of surface tension reduction or particle irregularities during the measurement.40

92

For a mixture of compounds of limited solubility,  = ∑2 12 ,-./,2 32 , where 12 is the volume

93

fraction of the ith component in the dry particle, ,-./,2 is the intrinsic κ of the ith component, and

5

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

94 95 96 97 98

32  is the dissolved fraction of the solute. This fraction is found using 2 = 42 5 /57,2 , where

42 is the aqueous solubility of the ith component expressed as volume solute per volume water at

saturation, 5 is the volume of water, and 57,2 is the total dry volume of the ith component. The

fraction dissolved is then 32  = 2 for partial dissolution (2 < 1) and 32  = 1 for

complete dissolution (2 ≥ 1). For pure compounds, κ is given by  = 3,-./

99 100

Page 6 of 43

(3)

Based on eq 3, for dissolved compounds (3 = 1), measurements of log"  as a function of

varying log  follow κ isolines with a slope of ~–3/2. For compounds with limited solubility

101

(3 < 1), the measurements will exhibit off-isoline behavior at the onset of solubility

102

limitation, when the particles are not fully dissolved at the point of activation.34,41-43

103

CCN activity was modeled using the open-source U-CCN model of Petters et al.30 In this study

104

improvements to the UNIFAC core (Fredenslund et al.26 eq 1–7 with the parameters table given

105

by Petters et al.30 containing values from Hansen et al.,44 Raatikainen and Laaksonen,28 and

106

Compernolle et al.29) are investigated for the hydroperoxide functional group. We specifically

107

seek interactions parameters that are optimal for CCN predictions only. As we have shown

108

previously,30,45 the transition from CCN inactive (κ ~ 0) to CCN active (κ consistent with

109

predictions based on molar volume) requires the removal of solubility/miscibility limitations.

110

Thus the prediction for compounds with solubility limitations is highly sensitive to the water

111

activity and liquid-liquid phase separation prediction at water activities very near unity. Three

112

approaches were taken: (1) original parameters were used (“default approach”); (2) the

113

CH(OOH) subgroup was replaced with a CH(O) and an OH group (“replacement

114

approach”);29,46-47 and (3) the CH(OOH) parameters were adjusted based on an exhaustive grid

115

search (“adjusted approach”). A brute force grid search was performed seeking a new pair of

6

ACS Paragon Plus Environment

Page 7 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

116

UNIFAC interaction parameters26 (>?→A ,>A→? ) for hydroperoxide (m) and water (n). This

117

procedure produced a line in parameter space for which the mean error was ~0. These values are

118

summarized in the Supporting Information (SI). Along this line, SSE asymptotically approached

119

a low value. The variation of modeled apparent κ as a function of carbon chain length showed

120

that some parameters result in unphysical behavior, and these parameters are discarded.

121

Unphysical behavior is diagnosed by plotting simulated homologous series of functionalized n-

122

alkanes.30 The physical basis for the unphysical regions is that, due to its reduced oxidation state,

123

the water affinity of the hydroperoxide group should not exceed that of the hydroxyl group. The

124

primary effects on CCN activity are molecular size and water solubility/miscibility and

125

secondary effects include activity coefficients and surface tension; thus the U-CCN model

126

captures the majority of the physical processes. Specifics about the procedure to constrain the

127

parameter space to a physical solution are provided in the SI. Adjusted parameters were selected

128

from within the set of reasonable parameters based on lowest available mean error and SSE.

129

These parameters are >?→A = −667.3 and >A→? = 963.6. In discussion to follow, UNIFAC

130

refers to the model with adjusted parameters unless otherwise specified. For the adjusted

131

L parameters the sum square error (SSE): ∑logDE FGHI  − logDE GJ7H/KH  and the mean

132

error: meanlogDE FGHI  − logDE GJ7H/KH  were both assessed. The modeled behavior of

133

hypothetical C6–C35 hydroperoxides was evaluated.

134

CCN activity was also estimated independently based on modeled molar volume and Flory-

135

Huggins-Köhler theory, which assumes an athermal mixture of molecules of different sizes,

136

considering only the entropy of mixing.48 Because it ignores chemical interactions, Flory-

137

Huggins-Köhler theory describes the CCN activity for compounds that are completely dissolved

138

at compositions corresponding to water contents at the critical supersaturation.45,49 Flory-

7

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

139

Huggins-Köhler theory has been shown to model the CCN activity of effectively soluble

140

molecules with molecular weights ranging from 80 to 2000 Da.45

141

Experimental

142

Two types of organic compounds were synthesized for this study: compounds containing

143

hydroperoxy groups and compounds containing carbonyl groups. Hydroperoxides included three

144

isomers each of C10, C12, C14, and C16 hydroperoxyacids (alkoxy hydroperoxyalkanoic acids) and

145

the same number of dihydroperoxides (dialkoxy dihydroperoxyalkanes). Carbonyls included a

146

C16 dione, C18 diketoaldehyde, and a C19 ketoester. Hydroperoxides were synthesized in solution

147

by ozonolysis of either an alkenoic acid or diene in the presence of excess alcohol.50-51 Details of

148

the synthesis are reported in the SI and the reaction mechanism is reproduced in Scheme S1.

149

Precursor compounds, targeted molecular structures, molecular weights, molar volumes

150

estimated from the U-CCN model, computer readable SMILES strings (simplified molecular

151

input line-entry system), and UNIFAC representations are summarized in Tables S2 and S3. The

152

alcohols and alkenes were selected to determine the carbon chain length of the hydroperoxide

153

product as well as the placement of the functional groups. A total of 24 hydroperoxides were

154

synthesized.

155

Synthesized hydroperoxides were purified using reverse-phase high-performance liquid

156

chromatography (HPLC) as described in previous work.23,25 Compound fractionation was

157

achieved using an Agilent 1100 series HPLC equipped with an Agilent Zorbax Eclipse Plus

158

XDB-C18 column (250 × 4.6 mm, 5 µm particle size) at room temperature and a Zorbax

159

Reliance cartridge guard column (4.6 × 12.5 mm with 5 µm particle size) with a flow rate of 0.8

160

mL min-1. A gradient elution method was employed starting with a 5:95 acetonitrile:water

161

mixture for 5 min, then ramping to 100:0 over 50 min and holding at this composition for 15

8

ACS Paragon Plus Environment

Page 8 of 43

Page 9 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

162

min. The column eluent was routed directly to a Collison atomizer, where the solution was

163

atomized, passed through silica gel and charcoal diffusion driers to remove the solvent, and

164

residual hydroperoxide particles were routed to the Scanning Flow-CCN measurement setup

165

described below.

166

Carbonyls were synthesized and purified using a method we have employed previously,52 and

167

which is described in the SI. 1H and 13C NMR were used to verify the structure and purity of the

168

synthesized compounds. Molecular structures, SMILES strings, molecular weights, modeled

169

molar volumes from the U-CCN model, and UNIFAC representations are summarized in Table

170

S4. The purified carbonyls were dissolved in acetonitrile, atomized, dried, and the residual

171

carbonyl particles were then routed to a Scanning Mobility Particle Sizer (SMPS)-CCN

172

measurement and a Hygroscopicity Tandem Differential Mobility Analyzer (HTDMA), both of

173

which are described below.

174

The CCN measurement setup had two modes of operation: Scanning Flow-CCN23,25,53 and

175

SMPS-CCN.54-55 Schematics of these previously described setups are provided in Figures S1 and

176

S2 in the SI. Scanning Flow-CCN mode had a faster duty cycle and was used to capture eluting

177

HPLC peaks; SMPS-CCN mode had a larger dynamic range and was used for selected

178

dihydroperoxides and for the already-purified carbonyls. In all cases aerosol was size-selected by

179

a differential mobility analyzer (DMA; TSI 3071A) operated at 9:1.5 L min-1 sheath-to-sample

180

flow ratio. The flow was split between a condensation particle counter (CPC; TSI 3010, 1

181

L min-1) and a streamwise gradient continuous flow CCN counter56 (DMT, 0.5 L min-1). CCN

182

instrument supersaturation was determined using size-selected ammonium sulfate aerosol and the

183

Aerosol Inorganic Model57-58 to relate dry diameter and critical supersaturation.25,41,59 For

184

Scanning Flow-CCN measurements, the diameter was held constant while the CCN flow was

9

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

185

scanned linearly from ~0.3 to 1.0 L min-1 in 44 s, translating to a supersaturation scan of 0.2 to

186

1.1%. For SMPS-CCN measurements the DMA voltage was exponentially ramped from 9000 to

187

50 V in 220 s, translating to a diameter ramp from 330 to 20 nm.55,60 SMPS-CCN data were

188

inverted to correct for multiply charged particles as described previously.61 Supersaturation and

189

diameter at 50% activation were used to calculate apparent κ using eq 1.

Page 10 of 43

190

The HTDMA instrument has been described previously62-64 and measured the humidified

191

diameter of wet aerosols at a controlled relative humidity.65 A schematic of the setup is provided

192

in Figure S3 in the SI. Dry 100 nm particles were selected by a DMA (TSI 3080) before

193

humidification by passage through a Nafion membrane immersed in water. The water

194

temperature dictates the dewpoint of the sample,66 which is then passed to a thermally isolated

195

high-flow DMA67 whose temperature inhomogeneities are actively held within ±0.02 °C.62 The

196

instrument was operated in comparative hygroscopicity mode. In this mode the aerosol flow to

197

the instrument alternates between sample aerosol and ammonium sulfate. The ammonium sulfate

198

growth factor is used to determine the relative humidity inside the instrument.62-64 Two important

199

modifications were made to the HTDMA instrument. First, the high flow DMA was placed in a

200

secondary temperature controlled enclosure that was held at ~19 °C, in addition to the heat

201

exchangers mounted to the column.62 The slight cooling below room temperature (22–30 °C)

202

allowed for stable long-term humidification at a relative humidity of 99% while avoiding

203

condensation of vapor in lines outside the secondary enclosure. Second, the humidified size

204

distributions were characterized using 10 Hz data acquisition, scanning from 3000 to 50 V in 50

205

s using a sheath:sample flow ratio of 14.5:1. The fast scans were implemented to interface the

206

instrument with the HPLC system (data not shown here). The rapid scanning leads to smearing

207

of the peak. However, it was verified experimentally that the accuracy of the retrieved mode

10

ACS Paragon Plus Environment

Page 11 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

208

diameter was not affected by the scan rate. Due to limited instrument availability, only carbonyls

209

were measured using the HTDMA instrument. Relative humidities and growth factors were used

210

to calculate apparent κ using eq 2.

211

Results and Discussion

212

Figure 1 and Table S5 summarize results for the hygroscopicity of hydroperoxyacids.

213

Hygroscopicity decreased with increasing carbon chain length and ranged from κ = 0.1 to 0.03.

214

The data fall between the Flory-Huggins prediction, which over-predicted κ for some but not all

215

of the isomers, and the adjusted U-CCN model prediction, which greatly under-predicted κ. This

216

under-prediction is discussed below in the context of the U-CCN model adjustment. Molecules

217

with the hydroperoxy group near the end of the carbon chain tend to have larger κ and were

218

predicted better by Flory-Huggins-Köhler theory. Compounds that fall below the Flory-Huggins

219

prediction undergo solubility- (referring to solubility of crystalline solids) or miscibility-

220

(referring to liquid-liquid phase separation) controlled droplet activation and are expected to be

221

only sparingly soluble/miscible in water.

222 223

Figure 1. Hygroscopicity of hydroperoxyacids (alkoxy-hydroperoxyalkanoic

224

acids) as a function of carbon chain length. Selected molecules are pictured at

225

right. Each point is the average of several experiments and error bars indicate the

11

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

226

min and max of all results. Colors indicate the length of the hydrophobic alkyl

227

tail. The black line shows the Flory-Huggins prediction and the purple line shows

228

the adjusted approach U-CCN model prediction. For Cn ≥ C11, the U-CCN

229

prediction is below the lower limit of the ordinate.

Page 12 of 43

230

Figure 2 and Table S6 summarize the hygroscopicity measurements of the dihydroperoxides.

231

Hygroscopicity decreased with increasing carbon chain length and ranged from κ = 0.05 to

232

0.007. The rate of decrease is more rapid and reaches lower values than for the

233

hydroperoxyacids. All data fell below the Flory-Huggins prediction, suggesting that the

234

dihydroperoxides are sparingly soluble in water. The data are reasonably well described by the

235

adjusted U-CCN model prediction, although several isomers have κ values that are significantly

236

lower than the adjusted U-CCN model. With the exception of C10 compounds, the

237

dihydroperoxides with the COOH group positioned closer to the center of the molecule were

238

more hygroscopic, consistent with our previous results for similar dihydroperoxides.23

239 240

Figure 2. Hygroscopicity of dihydroperoxides (dialkoxy-dihydroperoxyalkanes)

241

as a function of carbon chain length. Each point is the average of several

242

experiments and error bars indicate the min and max of all results. Colors indicate

12

ACS Paragon Plus Environment

Page 13 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

243

the length of the hydrophobic alkyl tail. The black line shows the Flory-Huggins

244

prediction and the purple line shows the adjusted approach U-CCN model

245

prediction.

246

Although at present the observed isomeric differences evident in the data in Figures 1 and 2 are

247

within the uncertainty of the HPLC-CCN technique (i.e. the error bars overlap), it is known that

248

the position of organic functional groups can affect compound properties including the

249

hygroscopicity,68 solubility,69 vapor pressure,70-71 melting point,72-73 reaction mechanisms,74-76

250

viscosity,77 and glass-transition temperatures.78 The behavior of dihydroperoxides is broadly

251

consistent with the hypothesis that polar functional groups are more efficient in promoting

252

solubility if they are positioned near the center of the molecule.69 The hydroperoxyacids do not

253

show a strong correlation between the length of the hydrophobic alkyl tail and isomer

254

hygroscopicity. Competing effects may be at play. The hydrophobic tail may fold as solvating

255

water molecules attract polar functional groups to the solvation shell of the molecule, and since

256

folded isomers tend to have a reduced effective molecular size69 they may be more

257

hygroscopic.45,49 However, if the tail instead forms a hydrophobic branch it may disrupt the first

258

solvation layer of water molecules, reducing hygroscopicity. These effects are influenced by the

259

specific geometries of molecules, and standard UNIFAC is indifferent toward the location of

260

functional groups. Therefore, isomer effects cannot be resolved with the U-CCN model. Marcolli

261

and Peter68 modified the UNIFAC functional group definitions to separately account for

262

molecular alkyl tails, and although this approach is attractive, caveats include the need for more

263

fitted parameters and the absence of a large volume of observational data.

264

Figure 3 and Table S7 summarize the additional tests of hygroscopicity and supersaturation at

265

CCN activation of C10 and C12 dihydroperoxides as a function of particle dry diameter.

13

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 43

266

Specifically, isomers #2.1 and #2.4 (Table S3) are shown (the C10 square and C12 triangle in

267

Figure 2). Each isomer was measured twice and differences between the first and the repeated

268

experiments are consistent with the ±10% variability in the supersaturation of the CCN

269

instrument. The relationship between critical supersaturation and dry diameter shows an

270

inflection point (panels A and C) where the C10 dihydroperoxide observation departs from the κ

271

isoline. No inflection point is observed for the C12-dihydroperoxide. The off-isoline behavior is

272

usually attributed to sparing solubility of the compound.34,41-43 The pure-compound κ model

273

including solubility34 (eq 3) describes the observations well in both cases. Based on the best fit of

274

eq 3 to the data, the solubility of the C10 compound was 4 = 0.31 with ,-./ = 0.042 and the

275

solubility of the C12 compound was 4 = 0.52 with ,-./ = 0.04. The solubility-controlled

276

transition in apparent κ (i.e. the inflection point) began for C10 dihydroperoxide particles near

277

 = 150 nm (panel A). The C10 particles were not fully dissolved at the point of activation for

278

 < 150 nm due to their solubility. The undissolved fraction increased at lower diameters and

279

could not contribute water uptake to the growing droplet. The C12 compounds did not exhibit

280

significant solubility limitations in the range measured.

281 282

Figure 3. Hygroscopicity of dihydroperoxides as a function of particle dry

283

diameter (A and B) and supersaturation at activation as a function of particle dry

14

ACS Paragon Plus Environment

Page 15 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

284

diameter (C and D). Panels A and C show the C10 dihydroperoxide (tail length 3)

285

and panels B and D show the C12 dihydroperoxide (tail length 4). Open symbols

286

indicate a repeated experiment. Each point is the average of ~3 experiments. Error

287

bars indicate the range of measured values. Grey lines of constant hygroscopicity

288

are shown in panels C and D. Black lines show the κ solubility model for pure

289

compounds (eq 3).34

290

The success of the pure-compound κ solubility framework in describing the observations in

291

Figure 3 is consistent with the near unity yields expected for the target hydroperoxides produced

292

in the synthesis of these standards.50,79-81 It is reassuring that these compounds behave as pure

293

compounds, given that for this set of solubility experiments the HPLC purification step was

294

skipped (due to its very low throughput). The HPLC-CCN measurements of the C10

295

dihydroperoxide isomers in Figure 2 were collected in a diameter range (123, 123, and 177 nm;

296

Table S6) close to the onset of solubility limitations shown in Figure 3 (panel A). Thus, it is

297

unsurprising that their apparent κ = 0.011 fell short of the fitted ,-./ = 0.042 in Figure 3. The

298

C12 dihydroperoxide apparent κ = 0.029 was nearer the fitted ,-./ = 0.04. Many sparingly

299

soluble compounds that should show inflection points do not in practice.34,42-43,82-83 Reasons for

300

this behavior frequently cited in the literature are solubilization of the analyte by trace impurities

301

present in the particle and uncertainties about the dried particle phase state.42,82

302

We now explore if effects other than solubility, such as reduced surface tension, viscosity, or

303

volatility, could contribute to the observed CCN data. Solubility, surface tension reduction,

304

viscosity, volatility and hygroscopicity are all linked to molecular structure. The surface tension

305

of C12 and C10 dihydroperoxides in aqueous solution is expected to lie between that of sodium

306

fatty acid salts and dicarboxylic acids (c.f. Figure 1 of Petters and Petters55), based on functional

15

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 43

307

group composition and carbon chain length. Thus, the surface tension reduction may affect CCN

308

activation. However, figure 3A shows a sharp reduction in κ at smaller diameters, contrary to

309

model predictions that show larger κ at smaller diameters regardless of the extent to which the

310

surfactant partitions to the surface.36,55,84 Facchini et al.85 proposed that surface tension reduction

311

during CCN activation greatly enhances the CCN activity of organic compounds. This prediction

312

was made using bulk measurements of surface tension from environmental samples. Bulk

313

measurements have also been used to predict strong CCN enhancement in more recent work

314

(e.g., Giordano et al.,86 Nozière et al.87). The problem with this approach is that hygroscopic

315

growth and CCN activation are controlled by both surface tension and water activity, and the

316

partitioning of surfactants to the surface of small droplets perturbs the water activity within the

317

droplet.88-90 Measurements of CCN activity for surfactant compounds have repeatedly shown that

318

strong CCN enhancement does not occur,55,90-91 and delayed CCN activation tests show that the

319

partitioning of surfactants is likely not time dependent within the measurement timescale.92

320

Measurement of surface tension is typically done in the bulk. Although measurements have been

321

recently extended to sizes as small as 5 microns, this is still much larger than the growing and

322

activating CCN droplet.93 The surface tension of droplets is undoubtedly perturbed by the solute;

323

however, few authors are willing to attribute their discrepancies entirely to surface tension

324

reduction.36-37 Several theories predicting surface partitioning in nanoscale droplets have been

325

developed89,94-96 and may be included in future versions of the U-CCN model.

326

Based on the molecular structures, the viscosities of the compounds in this study are well

327

below the 100 Pa·s delineation between liquid and semi-solid,97 making it unlikely that water

328

uptake was kinetically limited by water diffusion through the condensed phase organic matrix. It

329

is surprising, however, that the C10 compound exhibited a lower solubility than the C12

16

ACS Paragon Plus Environment

Page 17 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

330

compound, because the larger molecule would be expected to be less soluble due to the addition

331

of the hydrophobic group. One potential explanation is volatility. If the C10 particles partially

332

evaporated in the CCN instrument, a lower κ would be measured. The thermal gradient in the

333

CCN instrument is larger for larger supersaturation, and thus the decrease in κ with increasing

334

supersaturation might be also due to volatilization.98 Volatilization experiments that measure

335

aerosol diameter changes at ambient99-100 or chilled temperatures64 or with liquid water to inhibit

336

evaporation60,101-102 could probe this hypothesis.

337

Figure 4 shows the hygroscopicity of C14 hydroperoxide compounds as a function of the

338

number of carboxyl (panel A) or hydroperoxy (panel B) groups. Hygroscopicity increased with

339

the number of functional groups, consistent with previous results.23 Hydroperoxyacids measured

340

here (panel A) were slightly more CCN active (κ ~ 0.05) than previous results (κ ~ 0.02), but lie

341

within the experimental uncertainty shown in Figure 1. Note that only the isomer with a C5 tail

342

was measured in both studies. The observed slope d(κ)/d(carboxyl) was roughly equivalent to the

343

slope of the adjusted U-CCN model prediction. Current dihydroperoxide (panel B) observations

344

are consistent with previous results and range from κ = 0.007 to 0.03. The observed slope

345

d(κ)/d(hydroperoxy) was shallower than the U-CCN model prediction. The U-CCN prediction of

346

κ was consistent with observations for dihydroperoxides but under-predicted κ for the

347

hydroperoxyacids and monofunctional hydroperoxides. All compounds were less CCN active

348

than the Flory-Huggins model prediction.

17

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 43

349

350 351

Figure 4. Hygroscopicity of C14 hydroperoxides as a function of the number of

352

carboxyl (A) and hydroperoxy (B) groups. Data from Suda et al.23 are included

353

for comparison (open circles). Each point is the average of several experiments

354

and error bars indicate range. The length of the hydrophobic alkyl tail is indicated

355

with a number (Suda et al.23) or a color (this work). The black line shows the

356

Flory-Huggins prediction, the dotted grey line shows the observed slope, and the

357

purple line shows the adjusted U-CCN model.

358

Figure 5 and Table S8 summarize the hygroscopicity of carbonyls as a function of carbon

359

chain length. Hygroscopicity ranged from 0.0001 to 0.03, generally decreasing with carbon

360

number and increasing with the number of keto groups. Between compounds 1 and 2, the

361

addition of a keto group raises κ despite the addition of two carbons. Between compounds 2 and

362

3, the loss of a keto group renders compound 3 effectively CCN inactive. All observations fall

363

below the Flory-Huggins prediction, likely due to solubility limitations. The U-CCN model

364

prediction was κ ~ 0. Nevertheless, the increase in κ between compounds 1 and 2 shows that the

365

keto group promotes CCN activity. This trend is qualitatively consistent with predictions for

366

carbonyls of lower molar mass.30

18

ACS Paragon Plus Environment

Page 19 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

367 368

Figure 5. Hygroscopicity of carbonyls as a function of carbon chain length. All

369

compounds are pictured at right. Averages (horizontal lines) and individual

370

measurements (circles) are both shown. Sub-saturated measurements (using

371

HTDMA technique) are shown if they were above the detection limit (square).

372

The black line shows the Flory-Huggins prediction. The U-CCN model prediction

373

was κ ~ 0.

374

Figure 5 also shows HTDMA measurements of carbonyl compounds (summarized in Table

375

S9). The diameter growth factor,  from eq 2, was 0.809, 1.007, and 0.877 for carbonyl

376

compounds 1, 2, and 3 at 98–99% relative humidity. The C18 diketoaldehyde showed

377

significantly lower κ for measurements taken at sub-saturated relative humidity than for those

378

taken at supersaturated relative humidity. Growth factors less than one could be indicative of

379

slight evaporation or surface restructuring upon humidification.39,103 The growth factor of the C18

380

diketoaldehyde corresponds to κ = 0.0006 at 98.7% relative humidity; compounds with negative

381

κ are not included because water uptake was below the detection limit.

382

Figure 6 shows the relationship between the measured and modeled CCN activity for

383

hydroperoxides using the three different model approaches. The under-prediction of κ using the

19

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 43

384

default approach is a prominent feature (panel A). The C12 hydroperoxyacids (all isomers;

385

labeled “2”) and C10 dihydroperoxides (all isomers; labeled “5”) have larger error because some

386

of the diameters corresponded to sizes where κ varies strongly due to miscibility limitations. In

387

panel B, the replacement approach performed well for the dihydroperoxides (green circles 5–8

388

and purple square 12) and under-predicted κ for other compounds (hydroperoxyacids and C14

389

hydroperoxides from previous work.23,30

390 391

Figure 6. Hydroperoxide hygroscopicity modeled by the U-CCN model as

392

compared with the observed hygroscopicity. Panel A shows the default approach;

393

Panel B shows the replacement approach; and Panels C and D show the adjusted

394

approach. Panel D uses different axis scaling. Error bars indicate the range of

395

observed values for different isomers (x-error) and the range of modeled values

396

for different dry diameters (y-error). Grey shading indicates where compounds are

397

inactive as CCN under typical stratus conditions. Blue triangles numbered 1–4

398

refer to C10, C12, C14, and C16 hydroperoxyacids (Tables S2 and S5); green circles

399

numbered 5–8 refer to C10, C12, C14, and C16 dihydroperoxides (Tables S3 and

400

S6); and purple squares numbered 9–12 refer to C14 hydroperoxides from

401

previous work.23,30

20

ACS Paragon Plus Environment

Page 21 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

402

Predictions for the adjusted approach are superimposed in Figures 1, 2 and 4. These figures

403

show that the U-CCN model captured the absolute κ value and the relationship between κ and

404

carbon number well for the dihydroperoxides (Figure 2). In contrast, predictions are poor for the

405

hydroperoxyacids (Figure 1), and C14 monofunctional hydroperoxides (Figure 4). Figure 6 panels

406

C and D summarize the same data. The adjusted approach performed slightly better than the

407

replacement approach. The observed ranges in κ were generally larger than the modeled ranges

408

likely in part due to the different isomers measured, which UNIFAC cannot distinguish.

409

The closeness between the Flory-Huggins prediction and the data (Figure 1), suggests that the

410

hydroperoxyacids were sufficiently soluble, i.e. that the 3 term in eq 3 approaches unity.

411

This is clearly not captured by U-CCN model for this specific functional group pair. Potential

412

explanations for this behavior include strongly reduced surface tension, presence of trace

413

impurities, interactions between the hydroperoxide and acid groups that render the molecule

414

soluble but are not be captured in the functional group framework, or acid-water group

415

interaction parameters that are not tuned well enough to exactly model miscibility/solubility. At

416

this time the data generated for this work and the compounds analyzed in previous work23 are

417

insufficient to conclusively attribute the source of the error. Additional CCN studies targeting the

418

carboxyl group will be needed.

419

The grey regions in Figure 6 show κ values that are too low to be considered CCN active under

420

typical warm cloud (e.g., stratus) conditions. Compounds in these regions do not participate in

421

water uptake and contribute to CCN activity only by virtue of the effect of their dry volumes on

422

the overall diameter of a growing droplet. As compounds evolve by photochemical oxidation in

423

the atmosphere they generally become more hygroscopic,15-16 leaving the CCN-inactive region of

424

Figure 6. This solubility-controlled transition is aided by the addition of hygroscopic functional

21

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 43

425

groups to the carbon backbone23 and will be better estimated as more hygroscopicity

426

measurements for different functional groups become available.

427

To our knowledge, this and our previous work23 constitute the first measurements of the CCN

428

activity of pure organic hydroperoxides available in the literature. Compernolle et al.29 derived

429

hydroperoxide interaction parameters based on the SPARC (Sparc Performs Automated

430

Reasoning in Chemistry)104 model. These parameters have since been implemented in several

431

UNIFAC models that predict water uptake.30,47 Why does SPARC under-predict interactions

432

between hydroperoxides and water? The Compernolle parameters were intended for use in

433

predicting vapor-liquid equilibrium of α-pinene reaction products, and the authors point out that

434

as water uptake increases the accuracy of these interaction parameters for UNIFAC will decrease

435

because hydroperoxides participate in hydrogen bonding. The ability of polar organics to form

436

associated solutions or to be solvated by water generally complicates their treatment in models.

437

Multifunctional hydroperoxides are known to form intra- and inter-molecular hydrogen

438

bonds.48,105-107 For example, hydrogen bonding can draw the molecules closer together or

439

contract them, leading to misestimated molar volume in solutions.32,108-109 Similarly,

440

hydroperoxides can form dimers or larger clusters via hydrogen bonds,48,106-107 thereby acting as

441

larger molecules in solution.48 These complex interactions may contribute to the difficulty of

442

uniquely representing hydroperoxides within the UNIFAC modeling framework. Petters et al.30

443

used the U-CCN model with homologous series to quantify how effective functional groups are

444

in removing solubility limitations to CCN activation. These sensitivities can be expressed by

445

comparison to effect of changing carbon chain length (CHn). For example, to maintain constant

446

CCN activity, one hydroxyl group (promoting solubility) compensates for the addition of four

447

CHn groups (reducing solubility) and this is expressed as ∆PCHA S⁄∆POHS ~ − 4. The sensitivity

22

ACS Paragon Plus Environment

Page 23 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

448

for

the

hydroperoxy

group

using

the

U-CCN

model

default

approach

is

449

∆PCHA S⁄∆PCHA OOHS ~ − 2. Conversely, the sensitivity for the hydroperoxy group using the

450

adjusted approach is ∆PCHA OOHS⁄∆PCHA S ~ − 4, similar to that of a hydroxyl group. This is

451

physically reasonable given that both functional groups form hydrogen bonds with water

452

molecules.

453

Conclusions

454

Twenty-seven compounds were synthesized to systematically investigate how the number and

455

location of functional groups affect CCN activity. Measurements of CCN activity were

456

performed

457

(hydroperoxyacids) dialkoxy-dihydroperoxyalkanes (dihydroperoxides), and carbonyls. For most

458

compounds the sample was purified with HPLC and the HPLC eluent was interfaced with

459

Scanning Flow-CCN analysis to determine the CCN activity expressed as an apparent κ value.

for

three

main

compound

groups:

alkoxy-hydroperoxyalkanoic

acids

460

Hydroperoxyacids were more CCN active than dihydroperoxides of the same carbon chain

461

length. The keto functional group promoted CCN activity. CCN activity generally decreased

462

with increasing carbon chain length. Polar functional groups closer to the center of the molecule

463

were more effective in promoting CCN activity except for the hydroperoxyacids that did not

464

show this pattern. This may be due to folding of the hydrophobic alkane tail of the molecules,

465

which can simultaneously reduce molecular size (promoting hygroscopicity) and disrupt the first

466

solvation layer of water (reducing hygroscopicity).

467

Size

and

supersaturation-resolved

CCN

experiments

were

performed

with

two

468

dihydroperoxide compounds and showed that solubility limitations were likely responsible for

469

measured CCN activity below the Flory-Huggins prediction from molecular size. The C10

23

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

470

dihydroperoxide appeared to have limited solubility at CCN activation for dry diameters less

471

than 150 nm, while the C12 dihydroperoxide did not.

Page 24 of 43

472

The data collected in this work were used to evaluate the performance of a group contribution

473

model that predicts CCN activity. This model combines a UNIFAC core, a molar volume

474

prediction, and a liquid-liquid phase separation algorithm with Köhler theory to model the

475

activation diameter for organic compounds. Improved hydroperoxide-water group interaction

476

parameters were derived. The adjusted parameters result in excellent agreement between the data

477

and the dihydroperoxide compounds. However, the adjusted parameter model still

478

underestimates CCN activity for the monfunctional hydroperoxides and hydroperoxyacids.

479

Potential explanations for this behavior include strongly reduced surface tension, trace impurities

480

that remove solubility or miscibility limitations, hydroperoxide-acid group interactions that

481

promote solubility but are not captured in the functional group framework, or acid-water group

482

interaction parameters that do not capture miscibility/solubility near the point of droplet

483

activation. Additional studies are required to better understand the cause of these discrepancies.

484

The net result of the adjusted parameters is that the hydroperoxide group is approximately as

485

effective as the hydroxyl group in removing solubility/miscibility limitations from otherwise

486

non-polar molecules. This similarity between OH and CHn(OOH) is unsurprising due to the

487

expected underlying hydrogen bonding driving solvation and dissolution processes of the

488

molecule.

24

ACS Paragon Plus Environment

Page 25 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

489 490

The Journal of Physical Chemistry

Supporting Information Synthesis of compounds; instrument schematics; synthesized compounds, experimental results,

491

and modeling results; and exhaustive grid search for adjusted parameters.

492

Corresponding Author

493

*Sarah Suda Petters ([email protected]), Colorado State University, Department of Atmospheric

494

Science, Fort Collins, Colorado 80523-1371, United States.

495



Present address: Department of Environmental Sciences and Engineering, University of North

496

Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-7400, United States.

497

Acknowledgements

498

This research was funded by the United States Department of Energy, Office of Science grant

499

DE-SC 0012043. The authors thank Patrick Chuang for the use of the high flow DMA and

500

Timothy Wright for construction of the HTDMA enclosure and high-resolution scanning

501

software.

25

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

502

References

503 504

(1)

Köhler, H., The nucleus in and the growth of hygroscopic droplets. Trans. Faraday Soc. 1936, 32, 1152-1161, doi 10.1039/TF9363201152.

505 506 507

(2)

Farmer, D. K.; Cappa, C. D.; Kreidenweis, S. M., Atmospheric processes and their controlling influence on cloud condensation nuclei activity. Chem. Rev. 2015, 115 (10), 4199-4217, doi 10.1021/cr5006292.

508 509

(3)

Ångström, A., On the atmospheric transmission of sun radiation and on dust in the air. Geogr. Ann. 1929, 11, 156-166, doi 10.2307/519399.

510 511 512 513

(4)

Yu, H.; Kaufman, Y. J.; Chin, M.; Feingold, G.; Remer, L. A.; Anderson, T. L.; Balkanski, Y.; Bellouin, N.; Boucher, O.; Christopher, S. A.; et al., A review of measurement-based assessments of the aerosol direct radiative effect and forcing. Atmos. Chem. Phys. 2006, 6 (3), 613-666, doi 10.5194/acp-6-613-2006.

514 515 516

(5)

Dockery, D. W.; Pope, C. A.; Xu, X.; Spengler, J. D.; Ware, J. H.; Fay, M. E.; Ferris, B. G.; Speizer, F. E., An association between air pollution and mortality in six U.S. Cities. N. Engl. J. Med. 1993, 329 (24), 1753-1759, doi 10.1056/NEJM199312093292401.

517 518 519 520

(6)

West, J. J.; Cohen, A.; Dentener, F.; Brunekreef, B.; Zhu, T.; Armstrong, B.; Bell, M. L.; Brauer, M.; Carmichael, G.; Costa, D. L.; et al., "What we breathe impacts our health: Improving understanding of the link between air pollution and health". Environ. Sci. Technol. 2016, 50 (10), 4895-4904, doi 10.1021/acs.est.5b03827.

521 522 523 524 525

(7)

Zhang, Q.; Jimenez, J. L.; Canagaratna, M. R.; Allan, J. D.; Coe, H.; Ulbrich, I.; Alfarra, M. R.; Takami, A.; Middlebrook, A. M.; Sun, Y. L.; et al., Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-influenced northern hemisphere midlatitudes. Geophys. Res. Lett. 2007, 34 (13), L13801, doi 10.1029/2007GL029979.

526 527 528

(8)

Goldstein, A. H.; Galbally, I. E., Known and unexplored organic constituents in the earth's atmosphere. Environ. Sci. Technol. 2007, 41 (5), 1515-1521, doi 10.1021/es072476p.

529 530 531

(9)

Kroll, J. H.; Seinfeld, J. H., Chemistry of secondary organic aerosol: Formation and evolution of low-volatility organics in the atmosphere. Atmos. Environ. 2008, 42 (16), 3593-3624, doi 10.1016/j.atmosenv.2008.01.003.

532 533

(10)

Ziemann, P. J.; Atkinson, R., Kinetics, products, and mechanisms of secondary organic aerosol formation. Chem. Soc. Rev. 2012, 41 (19), 6582-6605, doi 10.1039/C2CS35122F.

534 535 536

(11)

Blando, J. D.; Turpin, B. J., Secondary organic aerosol formation in cloud and fog droplets: A literature evaluation of plausibility. Atmos. Environ. 2000, 34 (10), 16231632, doi 10.1016/S1352-2310(99)00392-1.

26

ACS Paragon Plus Environment

Page 26 of 43

Page 27 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

537 538 539 540

(12)

Carlton, A. G.; Turpin, B. J.; Altieri, K. E.; Seitzinger, S. P.; Mathur, R.; Roselle, S. J.; Weber, R. J., CMAQ model performance enhanced when in-cloud secondary organic aerosol is included: Comparisons of organic carbon predictions with measurements. Environ. Sci. Technol. 2008, 42 (23), 8798-8802, doi 10.1021/es801192n.

541 542 543

(13)

Ervens, B.; Turpin, B. J.; Weber, R. J., Secondary organic aerosol formation in cloud droplets and aqueous particles (aqSOA): A review of laboratory, field and model studies. Atmos. Chem. Phys. 2011, 11 (21), 11069-11102, doi 10.5194/acp-11-11069-2011.

544 545 546

(14)

McNeill, V. F., Aqueous organic chemistry in the atmosphere: Sources and chemical processing of organic aerosols. Environ. Sci. Technol. 2015, 49 (3), 1237-1244, doi 10.1021/es5043707.

547 548 549

(15)

Jimenez, J. L.; Canagaratna, M. R.; Donahue, N. M.; Prévôt, A. S. H.; Zhang, Q.; Kroll, J. H.; DeCarlo, P. F.; Allan, J. D.; Coe, H.; Ng, N. L.; et al., Evolution of organic aerosols in the atmosphere. Science 2009, 326 (5959), 1525-1529, doi 10.1126/science.1180353.

550 551 552 553 554

(16)

Chang, R. Y.-W.; Slowik, J. G.; Shantz, N. C.; Vlasenko, A.; Liggio, J.; Sjostedt, S. J.; Leaitch, W. R.; Abbatt, J. P. D., The hygroscopicity parameter (κ) of ambient organic aerosol at a field site subject to biogenic and anthropogenic influences: Relationship to degree of aerosol oxidation. Atmos. Chem. Phys. 2010, 10 (11), 5047-5064, doi 10.5194/acp-10-5047-2010.

555 556 557

(17)

Mei, F.; Setyan, A.; Zhang, Q.; Wang, J., CCN activity of organic aerosols observed downwind of urban emissions during CARES. Atmos. Chem. Phys. 2013, 13 (24), 1215512169, doi 10.5194/acp-13-12155-2013.

558 559 560 561

(18)

Heald, C. L.; Kroll, J. H.; Jimenez, J. L.; Docherty, K. S.; DeCarlo, P. F.; Aiken, A. C.; Chen, Q.; Martin, S. T.; Farmer, D. K.; Artaxo, P., A simplified description of the evolution of organic aerosol composition in the atmosphere. Geophys. Res. Lett. 2010, 37 (8), L08803, doi 10.1029/2010GL042737.

562 563 564 565

(19)

Kroll, J. H.; Donahue, N. M.; Jimenez, J. L.; Kessler, S. H.; Canagaratna, M. R.; Wilson, K. R.; Altieri, K. E.; Mazzoleni, L. R.; Wozniak, A. S.; Bluhm, H.; et al., Carbon oxidation state as a metric for describing the chemistry of atmospheric organic aerosol. Nat. Chem. 2011, 3 (2), 133-139, doi 10.1038/nchem.948.

566 567 568 569 570

(20)

Lambe, A. T.; Onasch, T. B.; Massoli, P.; Croasdale, D. R.; Wright, J. P.; Ahern, A. T.; Williams, L. R.; Worsnop, D. R.; Brune, W. H.; Davidovits, P., Laboratory studies of the chemical composition and cloud condensation nuclei (CCN) activity of secondary organic aerosol (SOA) and oxidized primary organic aerosol (OPOA). Atmos. Chem. Phys. 2011, 11 (17), 8913-8928, doi 10.5194/acp-11-8913-2011.

571 572 573 574

(21)

Zhang, F.; Li, Z.; Li, Y.; Sun, Y.; Wang, Z.; Li, P.; Sun, L.; Wang, P.; Cribb, M.; Zhao, C.; et al., Impacts of organic aerosols and its oxidation level on CCN activity from measurement at a suburban site in China. Atmos. Chem. Phys. 2016, 16 (8), 5413-5425, doi 10.5194/acp-16-5413-2016.

27

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

575 576 577

(22)

Rickards, A. M. J.; Miles, R. E. H.; Davies, J. F.; Marshall, F. H.; Reid, J. P., Measurements of the sensitivity of aerosol hygroscopicity and the κ parameter to the O/C ratio. J. Phys. Chem. A 2013, 117 (51), 14120-14131, doi 10.1021/jp407991n.

578 579 580 581

(23)

Suda, S. R.; Petters, M. D.; Yeh, G. K.; Strollo, C.; Matsunaga, A.; Faulhaber, A.; Ziemann, P. J.; Prenni, A. J.; Carrico, C. M.; Sullivan, R. C.; et al., Influence of functional groups on organic aerosol cloud condensation nucleus activity. Environ. Sci. Technol. 2014, 48 (17), 10182-10190, doi 10.1021/es502147y.

582 583 584

(24)

Petters, M. D.; Kreidenweis, S. M., A single parameter representation of hygroscopic growth and cloud condensation nucleus activity. Atmos. Chem. Phys. 2007, 7 (8), 19611971, doi 10.5194/acp-7-1961-2007.

585 586 587

(25)

Suda, S. R.; Petters, M. D.; Matsunaga, A.; Sullivan, R. C.; Ziemann, P. J.; Kreidenweis, S. M., Hygroscopicity frequency distributions of secondary organic aerosols. J. Geophys. Res. -Atmos. 2012, 117, D04207, doi 10.1029/2011JD016823.

588 589 590

(26)

Fredenslund, A.; Jones, R. L.; Prausnitz, J. M., Group-contribution estimation of activity coefficients in nonideal liquid mixtures. AIChE J. 1975, 21 (6), 1086-1099, doi 10.1002/aic.690210607.

591 592

(27)

Ming, Y.; Russell, L. M., Thermodynamic equilibrium of organic-electrolyte mixtures in aerosol particles. AIChE J. 2002, 48 (6), 1331-1348, doi 10.1002/aic.690480619.

593 594 595

(28)

Raatikainen, T.; Laaksonen, A., Application of several activity coefficient models to water-organic-electrolyte aerosols of atmospheric interest. Atmos. Chem. Phys. 2005, 5 (9), 2475-2495, doi 10.5194/acp-5-2475-2005.

596 597

(29)

Compernolle, S.; Ceulemans, K.; Müller, J.-F., Influence of non-ideality on condensation to aerosol. Atmos. Chem. Phys. 2009, 9 (4), 1325-1337, doi 10.5194/acp-9-1325-2009.

598 599 600

(30)

Petters, M. D.; Kreidenweis, S. M.; Ziemann, P. J., Prediction of cloud condensation nuclei activity for organic compounds using functional group contribution methods. Geosci. Model Dev. 2016, 9 (1), 111-124, doi 10.5194/gmd-9-111-2016.

601 602 603

(31)

Zuend, A.; Marcolli, C.; Luo, B. P.; Peter, T., A thermodynamic model of mixed organicinorganic aerosols to predict activity coefficients. Atmos. Chem. Phys. 2008, 8 (16), 4559-4593, doi 10.5194/acp-8-4559-2008.

604 605 606

(32)

Girolami, G. S., A simple "back of the envelope" method for estimating the densities and molecular volumes of liquids and solids. J. Chem. Educ. 1994, 71 (11), 962-964, doi 10.1021/ed071p962.

607 608 609

(33)

Eubank, P. T.; Elhassan, A. E.; Barrufet, M. A.; Whiting, W. B., Area method for prediction of fluid-phase equilibria. Ind. Eng. Chem. Res. 1992, 31 (3), 942-949, doi 10.1021/ie00003a041.

28

ACS Paragon Plus Environment

Page 28 of 43

Page 29 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

610 611 612

(34)

Petters, M. D.; Kreidenweis, S. M., A single parameter representation of hygroscopic growth and cloud condensation nucleus activity - Part 2: Including solubility. Atmos. Chem. Phys. 2008, 8 (20), 6273-6279, doi 10.5194/acp-8-6273-2008.

613 614 615

(35)

Prenni, A. J.; Petters, M. D.; Kreidenweis, S. M.; DeMott, P. J.; Ziemann, P. J., Cloud droplet activation of secondary organic aerosol. J. Geophys. Res. 2007, 112, D10223, doi 10.1029/2006JD007963.

616 617 618 619

(36)

Irwin, M.; Good, N.; Crosier, J.; Choularton, T. W.; McFiggans, G., Reconciliation of measurements of hygroscopic growth and critical supersaturation of aerosol particles in central Germany. Atmos. Chem. Phys. 2010, 10 (23), 11737-11752, doi 10.5194/acp-1011737-2010.

620 621 622 623 624

(37)

Wex, H.; Petters, M. D.; Carrico, C. M.; Hallbauer, E.; Massling, A.; McMeeking, G. R.; Poulain, L.; Wu, Z.; Kreidenweis, S. M.; Stratmann, F., Towards closing the gap between hygroscopic growth and activation for secondary organic aerosol: Part 1 - Evidence from measurements. Atmos. Chem. Phys. 2009, 9 (12), 3987-3997, doi 10.5194/acp-9-39872009.

625 626 627 628

(38)

Petters, M. D.; Wex, H.; Carrico, C. M.; Hallbauer, E.; Massling, A.; McMeeking, G. R.; Poulain, L.; Wu, Z.; Kreidenweis, S. M.; Stratmann, F., Towards closing the gap between hygroscopic growth and activation for secondary organic aerosol – Part 2: Theoretical approaches. Atmos. Chem. Phys. 2009, 9 (12), 3999-4009, doi 10.5194/acp-9-3999-2009.

629 630 631 632

(39)

Mikhailov, E.; Vlasenko, S.; Martin, S. T.; Koop, T.; Pöschl, U., Amorphous and crystalline aerosol particles interacting with water vapor: Conceptual framework and experimental evidence for restructuring, phase transitions and kinetic limitations. Atmos. Chem. Phys. 2009, 9 (24), 9491-9522, doi 10.5194/acp-9-9491-2009.

633 634 635 636

(40)

Sullivan, R. C.; Moore, M. J. K.; Petters, M. D.; Kreidenweis, S. M.; Roberts, G. C.; Prather, K. A., Effect of chemical mixing state on the hygroscopicity and cloud nucleation properties of calcium mineral dust particles. Atmos. Chem. Phys. 2009, 9 (10), 3303-3316, doi 10.5194/acp-9-3303-2009.

637 638

(41)

Christensen, S. I.; Petters, M. D., The role of temperature in cloud droplet activation. J. Phys. Chem. A 2012, 116 (39), 9706-9717, doi 10.1021/jp3064454.

639 640 641

(42)

Hori, M.; Ohta, S.; Murao, N.; Yamagata, S., Activation capability of water soluble organic substances as CCN. J. Aerosol Sci. 2003, 34 (4), 419-448, doi 10.1016/S00218502(02)00190-8.

642 643 644 645

(43)

Hings, S. S.; Wrobel, W. C.; Cross, E. S.; Worsnop, D. R.; Davidovits, P.; Onasch, T. B., CCN activation experiments with adipic acid: Effect of particle phase and adipic acid coatings on soluble and insoluble particles. Atmos. Chem. Phys. 2008, 8 (14), 3735-3748, doi 10.5194/acp-8-3735-2008.

29

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

646 647 648

(44)

Hansen, H. K.; Rasmussen, P.; Fredenslund, A.; Schiller, M.; Gmehling, J., Vapor-liquid equilibria by UNIFAC group contribution. 5. Revision and extension. Ind. Eng. Chem. Res. 1991, 30 (10), 2352-2355, doi 10.1021/ie00058a017.

649 650 651

(45)

Petters, M. D.; Kreidenweis, S. M.; Prenni, A. J.; Sullivan, R. C.; Carrico, C. M.; Koehler, K. A.; Ziemann, P. J., Role of molecular size in cloud droplet activation. Geophys. Res. Lett. 2009, 36 (22), L22801, doi 10.1029/2009GL040131.

652 653 654 655 656

(46)

Bilde, M.; Barsanti, K.; Booth, M.; Cappa, C. D.; Donahue, N. M.; Emanuelsson, E. U.; McFiggans, G.; Krieger, U. K.; Marcolli, C.; Topping, D.; et al., Saturation vapor pressures and transition enthalpies of low-volatility organic molecules of atmospheric relevance: From dicarboxylic acids to complex mixtures. Chem. Rev. 2015, 115 (10), 4115-4156, doi 10.1021/cr5005502.

657 658 659

(47)

Zuend, A.; Seinfeld, J. H., Modeling the gas-particle partitioning of secondary organic aerosol: The importance of liquid-liquid phase separation. Atmos. Chem. Phys. 2012, 12 (9), 3857-3882, doi 10.5194/acp-12-3857-2012.

660 661 662

(48)

Prausnitz, J. M.; Lichtenthaler, R. N.; de Azevedo, E. G., Molecular thermodynamics of fluid-phase equilibria. Prentice Hall: Upper Saddle River, New Jersey, USA, 1999; Vol. 3rd Ed.

663 664 665

(49)

Petters, M. D.; Kreidenweis, S. M.; Snider, J. R.; Koehler, K. A.; Wang, Q.; Prenni, A. J.; DeMott, P. J., Cloud droplet activation of polymerized organic aerosol. Tellus B 2006, 58 (3), 196-205, doi 10.1111/j.1600-0889.2006.00181.x.

666 667 668

(50)

Tobias, H. J.; Ziemann, P. J., Compound identification in organic aerosols using temperature-programmed thermal desorption particle beam mass spectrometry. Anal. Chem. 1999, 71 (16), 3428-3435, doi 10.1021/ac990056f.

669 670 671

(51)

Tobias, H. J.; Kooiman, P. M.; Docherty, K. S.; Ziemann, P. J., Real-time chemical analysis of organic aerosols using a thermal desorption particle beam mass spectrometer. Aerosol Sci. Technol. 2000, 33 (1-2), 170-190, doi 10.1080/027868200410912.

672 673 674 675

(52)

Ranney, A. P.; Ziemann, P. J., Identification and quantification of oxidized organic aerosol compounds using derivatization, liquid chromatography, and chemical ionization mass spectrometry. Aerosol Sci. Technol. 2016, 1-12, doi 10.1080/02786826.2016.1271108.

676 677 678

(53)

Moore, R. H.; Nenes, A., Scanning flow CCN analysis — a method for fast measurements of CCN spectra. Aerosol Sci. Technol. 2009, 43 (12), 1192-1207, doi 10.1080/02786820903289780.

679 680 681

(54)

Moore, R. H.; Nenes, A.; Medina, J., Scanning mobility CCN analysis—a method for fast measurements of size-resolved CCN distributions and activation kinetics. Aerosol Sci. Technol. 2010, 44 (10), 861-871, doi 10.1080/02786826.2010.498715.

30

ACS Paragon Plus Environment

Page 30 of 43

Page 31 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

682 683 684

(55)

Petters, S. S.; Petters, M. D., Surfactant effect on cloud condensation nuclei for twocomponent internally mixed aerosols. J. Geophys. Res. -Atmos. 2016, 121 (4), 18781895, doi 10.1002/2015JD024090.

685 686 687

(56)

Roberts, G. C.; Nenes, A., A continuous-flow streamwise thermal-gradient CCN chamber for atmospheric measurements. Aerosol Sci. Technol. 2005, 39 (3), 206-221, doi 10.1080/027868290913988.

688 689 690

(57)

Clegg, S. L.; Brimblecombe, P.; Wexler, A. S., Thermodynamic model of the system H+– NH4+– Na+– SO42-– NO3-– Cl-– H2O at 298.15 K. J. Phys. Chem. A 1998, 102 (12), 21552171, doi 10.1021/jp973043j.

691 692 693

(58)

Wexler, A. S.; Clegg, S. L., Atmospheric aerosol models for systems including the ions H+, NH4+, Na+, SO42-, NO3-, Cl-, Br-, and H2O. J. Geophys. Res. 2002, 107 (D14), 4207, doi 10.1029/2001JD000451.

694 695 696 697 698

(59)

Rose, D.; Gunthe, S. S.; Mikhailov, E.; Frank, G. P.; Dusek, U.; Andreae, M. O.; Pöschl, U., Calibration and measurement uncertainties of a continuous-flow cloud condensation nuclei counter (DMT-CCNC): CCN activation of ammonium sulfate and sodium chloride aerosol particles in theory and experiment. Atmos. Chem. Phys. 2008, 8 (5), 1153-1179, doi 10.5194/acp-8-1153-2008.

699 700 701

(60)

Nguyen, T. K. V.; Petters, M. D.; Suda, S. R.; Guo, H.; Weber, R. J.; Carlton, A. G., Trends in particle-phase liquid water during the Southern Oxidant and Aerosol Study. Atmos. Chem. Phys. 2014, 14 (20), 10911-10930, doi 10.5194/acp-14-10911-2014.

702 703 704

(61)

Petters, M. D.; Carrico, C. M.; Kreidenweis, S. M.; Prenni, A. J.; DeMott, P. J.; Collett, J. L., Jr.; Moosmüller, H., Cloud condensation nucleation activity of biomass burning aerosol. J. Geophys. Res. 2009, 114, D22205, doi 10.1029/2009JD012353.

705 706 707 708

(62)

Suda, S. R.; Petters, M. D., Accurate determination of aerosol activity coefficients at relative humidities up to 99% using the hygroscopicity tandem differential mobility analyzer technique. Aerosol Sci. Technol. 2013, 47 (9), 991-1000, doi 10.1080/02786826.2013.807906.

709 710 711 712

(63)

Dawson, K. W.; Petters, M. D.; Meskhidze, N.; Petters, S. S.; Kreidenweis, S. M., Hygroscopic growth and cloud droplet activation of xanthan gum as a proxy for marine hydrogels. J. Geophys. Res. -Atmos. 2016, 121 (19), 11803-11818, doi 10.1002/2016JD025143.

713 714 715

(64)

Wright, T. P.; Song, C.; Sears, S.; Petters, M. D., Thermodynamic and kinetic behavior of glycerol aerosol. Aerosol Sci. Technol. 2016, 50 (12), 1385-1396, doi 10.1080/02786826.2016.1245405.

716 717 718

(65)

Rader, D. J.; McMurry, P. H., Application of the tandem differential mobility analyzer to studies of droplet growth or evaporation. J. Aerosol Sci. 1986, 17 (5), 771-787, doi 10.1016/0021-8502(86)90031-5.

31

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

719 720 721

(66)

Rothfuss, N. E.; Petters, M. D., Coalescence-based assessment of aerosol phase state using dimers prepared through a dual-differential mobility analyzer technique. Aerosol Sci. Technol. 2016, 1-12, doi 10.1080/02786826.2016.1221050.

722 723 724

(67)

Stolzenburg, M.; Kreisberg, N.; Hering, S., Atmospheric size distributions measured by differential mobility optical particle size spectrometry. Aerosol Sci. Technol. 1998, 29 (5), 402-418, doi 10.1080/02786829808965579.

725 726 727

(68)

Marcolli, C.; Peter, T., Water activity in polyol/water systems: New UNIFAC parameterization. Atmos. Chem. Phys. 2005, 5 (6), 1545-1555, doi 10.5194/acp-5-15452005.

728 729

(69)

Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M., Environmental organic chemistry. 1 ed.; John Wiley & Sons: New York, 1993.

730 731 732

(70)

Bilde, M.; Svenningsson, B.; Mønster, J.; Rosenørn, T., Even–odd alternation of evaporation rates and vapor pressures of C3–C9 dicarboxylic acid aerosols. Environ. Sci. Technol. 2003, 37 (7), 1371-1378, doi 10.1021/es0201810.

733 734 735

(71)

Frosch, M.; Zardini, A. A.; Platt, S. M.; Müller, L.; Reinnig, M.-C.; Hoffmann, T.; Bilde, M., Thermodynamic properties and cloud droplet activation of a series of oxo-acids. Atmos. Chem. Phys. 2010, 10 (13), 5873-5890, doi 10.5194/acp-10-5873-2010.

736 737

(72)

Thalladi, V. R.; Boese, R.; Weiss, H.-C., The melting point alternation in α,ωalkanedithiols. J. Am. Chem. Soc. 2000, 122 (6), 1186-1190, doi 10.1021/ja993422l.

738 739 740

(73)

Thalladi, V. R.; Nüsse, M.; Boese, R., The melting point alternation in α,ωalkanedicarboxylic acids. J. Am. Chem. Soc. 2000, 122 (38), 9227-9236, doi 10.1021/ja0011459.

741 742 743

(74)

Matsunaga, A.; Ziemann, P. J., Yields of β-hydroxynitrates, dihydroxynitrates, and trihydroxynitrates formed from OH radical-initiated reactions of 2-methyl-1-alkenes. Proc. Natl. Acad. Sci. U.S.A. 2010, 107 (15), 6664-6669, doi 10.1073/pnas.0910585107.

744 745 746

(75)

Matsunaga, A.; Ziemann, P. J., Yields of β-hydroxynitrates and dihydroxynitrates in aerosol formed from OH radical-initiated reactions of linear alkenes in the presence of NOx. J. Phys. Chem. A 2009, 113 (3), 599-606, doi 10.1021/jp807764d.

747 748 749

(76)

Algrim, L. B.; Ziemann, P. J., Effect of the keto group on yields and composition of organic aerosol formed from OH radical-initiated reactions of ketones in the presence of NOx. J. Phys. Chem. A 2016, 120 (35), 6978-6989, doi 10.1021/acs.jpca.6b05839.

750 751 752

(77)

Diller, D. E.; Van Poolen, L. J., Measurements of the viscosities of saturated and compressed liquid normal butane and isobutane. Int. J. Thermophys. 1985, 6 (1), 43-62, doi 10.1007/BF00505791.

753 754

(78)

Rothfuss, N. E.; Petters, M. D., Influence of functional groups on the viscosity of organic aerosol. Environ. Sci. Technol. 2017, 51 (1), 271-279, doi 10.1021/acs.est.6b04478.

32

ACS Paragon Plus Environment

Page 32 of 43

Page 33 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

755 756

(79)

Bailey, P. S., The reactions of ozone with organic compounds. Chem. Rev. 1958, 58 (5), 925-1010, doi 10.1021/cr50023a005.

757 758

(80)

Murai, S.; Sonoda, N.; Tsutsumi, S., The redox reaction of 1-ethoxy-n-heptyl hydroperoxide. Bull. Chem. Soc. Jpn. 1964, 37 (8), 1187-1190, doi 10.1246/bcsj.37.1187.

759 760 761 762

(81)

Tobias, H. J.; Ziemann, P. J., Thermal desorption mass spectrometric analysis of organic aerosol formed from reactions of 1-tetradecene and O3 in the presence of alcohols and carboxylic acids. Environ. Sci. Technol. 2000, 34 (11), 2105-2115, doi 10.1021/es9907156.

763 764 765

(82)

Bilde, M.; Svenningsson, B., CCN activation of slightly soluble organics: The importance of small amounts of inorganic salt and particle phase. Tellus B 2004, 56 (2), 128-134, doi 10.1111/j.1600-0889.2004.00090.x.

766 767

(83)

Kristensson, A.; Rosenørn, T.; Bilde, M., Cloud droplet activation of amino acid aerosol particles. J. Phys. Chem. A 2010, 114 (1), 379-386, doi 10.1021/jp9055329.

768 769 770

(84)

Wex, H.; Stratmann, F.; Topping, D.; McFiggans, G., The Kelvin versus the Raoult term in the Köhler equation. J. Atmos. Sci. 2008, 65 (12), 4004-4016, doi 10.1175/2008JAS2720.1.

771 772 773

(85)

Facchini, M. C.; Mircea, M.; Fuzzi, S.; Charlson, R. J., Cloud albedo enhancement by surface-active organic solutes in growing droplets. Nature 1999, 401 (6750), 257-259, doi 10.1038/45758.

774 775 776 777

(86)

Giordano, M. R.; Short, D. Z.; Hosseini, S.; Lichtenberg, W.; Asa-Awuku, A., Changes in droplet surface tension affect the observed hygroscopicity of photochemically aged biomass burning aerosol. Environ. Sci. Technol. 2013, 47 (19), 10980-10986, doi 10.1021/es401867j.

778 779 780

(87)

Nozière, B.; Baduel, C.; Jaffrezo, J.-L., The dynamic surface tension of atmospheric aerosol surfactants reveals new aspects of cloud activation. Nat. Commun. 2014, 5, 3333, doi 10.1038/ncomms4335.

781 782 783

(88)

Li, Z.; Williams, A. L.; Rood, M. J., Influence of soluble surfactant properties on the activation of aerosol particles containing inorganic solute. J. Atmos. Sci. 1998, 55 (10), 1859-1866, doi 10.1175/1520-0469(1998)0552.0.CO;2.

784 785 786

(89)

Sorjamaa, R.; Svenningsson, B.; Raatikainen, T.; Henning, S.; Bilde, M.; Laaksonen, A., The role of surfactants in Köhler theory reconsidered. Atmos. Chem. Phys. 2004, 4 (8), 2107-2117, doi 10.5194/acp-4-2107-2004.

787 788 789 790

(90)

Prisle, N. L.; Raatikainen, T.; Sorjamaa, R.; Svenningsson, B.; Laaksonen, A.; Bilde, M., Surfactant partitioning in cloud droplet activation: A study of C8, C10, C12 and C14 normal fatty acid sodium salts. Tellus B 2008, 60 (3), 416-431, doi 10.1111/j.16000889.2008.00352.x.

33

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

791 792 793 794

(91)

Kristensen, T. B.; Wex, H.; Nekat, B.; Nøjgaard, J. K.; van Pinxteren, D.; Lowenthal, D. H.; Mazzoleni, L. R.; Dieckmann, K.; Koch, C. B.; Mentel, T. F.; et al., Hygroscopic growth and CCN activity of HULIS from different environments. J. Geophys. Res. Atmos. 2012, 117, doi 10.1029/2012JD018249.

795 796 797

(92)

Petters, M. D.; Suda, S. R.; Christensen, S. I., The role of dynamic surface tension in cloud droplet activation. AIP Conf. Proc. 2013, 1527 (1), 801-807, doi 10.1063/1.4803393.

798 799 800

(93)

Bzdek, B. R.; Power, R. M.; Simpson, S. H.; Reid, J. P.; Royall, C. P., Precise, contactless measurements of the surface tension of picolitre aerosol droplets. Chem. Sci. 2016, 7 (1), 274-285, doi 10.1039/C5SC03184B.

801 802

(94)

Raatikainen, T.; Laaksonen, A., A simplified treatment of surfactant effects on cloud drop activation. Geosci. Model Dev. 2011, 4 (1), 107-116, doi 10.5194/gmd-4-107-2011.

803 804 805

(95)

Ruehl, C. R.; Davies, J. F.; Wilson, K. R., An interfacial mechanism for cloud droplet formation on organic aerosols. Science 2016, 351 (6280), 1447, doi 10.1126/science.aad4889.

806 807 808 809

(96)

Petters, M. D.; Kreidenweis, S. M., A single parameter representation of hygroscopic growth and cloud condensation nucleus activity – Part 3: Including surfactant partitioning. Atmos. Chem. Phys. 2013, 13 (2), 1081-1091, doi 10.5194/acp-13-10812013.

810 811 812 813

(97)

Renbaum-Wolff, L.; Grayson, J. W.; Bateman, A. P.; Kuwata, M.; Sellier, M.; Murray, B. J.; Shilling, J. E.; Martin, S. T.; Bertram, A. K., Viscosity of α-pinene secondary organic material and implications for particle growth and reactivity. Proc. Natl. Acad. Sci. U.S.A. 2013, 110 (20), 8014-8019, doi 10.1073/pnas.1219548110.

814 815 816

(98)

Asa-Awuku, A.; Engelhart, G. J.; Lee, B. H.; Pandis, S. N.; Nenes, A., Relating CCN activity, volatility, and droplet growth kinetics of β-caryophyllene secondary organic aerosol. Atmos. Chem. Phys. 2009, 9 (3), 795-812, doi 10.5194/acp-9-795-2009.

817 818 819 820

(99)

Ortiz-Montalvo, D.; Lim, Y. B.; Perri, M. J.; Seitzinger, S. P.; Turpin, B. J., Volatility and yield of glycolaldehyde SOA formed through aqueous photochemistry and droplet evaporation. Aerosol Sci. Technol. 2012, 46 (9), 1002-1014, doi 10.1080/02786826.2012.686676.

821 822 823 824

(100) Ortiz-Montalvo, D.; Häkkinen, S. A. K.; Schwier, A. N.; Lim, Y. B.; McNeill, V. F.; Turpin, B. J., Ammonium addition (and aerosol pH) has a dramatic impact on the volatility and yield of glyoxal secondary organic aerosol. Environ. Sci. Technol. 2014, 48 (1), 255-262, doi 10.1021/es4035667.

825 826 827

(101) Topping, D.; Connolly, P.; McFiggans, G., Cloud droplet number enhanced by cocondensation of organic vapours. Nat. Geosci. 2013, 6 (6), 443-446, doi 10.1038/ngeo1809.

34

ACS Paragon Plus Environment

Page 34 of 43

Page 35 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

828 829 830

(102) Nakao, S.; Suda, S. R.; Camp, M.; Petters, M. D.; Kreidenweis, S. M., Droplet activation of wet particles: Development of the wet CCN approach. Atmos. Meas. Tech. 2014, 7 (7), 2227-2241, doi 10.5194/amt-7-2227-2014.

831 832 833

(103) Lee, A. K. Y.; Ling, T. Y.; Chan, C. K., Understanding hygroscopic growth and phase transformation of aerosols using single particle Raman spectroscopy in an electrodynamic balance. Faraday Discuss. 2008, 137 (0), 245-263, doi 10.1039/B704580H.

834 835 836

(104) Hilal, S. H.; Karickhoff, S. W.; Carreira, L. A., Prediction of the vapor pressure boiling point, heat of vaporization and diffusion coefficient of organic compounds. QSAR Comb. Sci. 2003, 22 (6), 565-574, doi 10.1002/qsar.200330812.

837 838

(105) Minkoff, G. J., The infra-red absorption spectra of organic peroxides. Proc. Roy, Soc. Lond. A 1954, 224 (1157), 176-191, doi 10.1098/rspa.1954.0150.

839 840

(106) Johnson, R. M.; Siddiqi, I. W., The determination of organic peroxides. Pergamon: Oxford, 1970; Vol. 4, p 119.

841 842 843

(107) Yablonskii, O. P.; Belyaev, V. A.; Vinogradov, A. N., Association of hydrocarbon hydroperoxides. Russ. Chem. Rev. 1972, 41 (7), 565-573, doi 10.1070/RC1972v041n07ABEH002077.

844 845 846

(108) Barley, M. H.; Topping, D. O.; McFiggans, G., Critical assessment of liquid density estimation methods for multifunctional organic compounds and their use in atmospheric science. J. Phys. Chem. A 2013, 117 (16), 3428-3441, doi 10.1021/jp304547r.

847 848 849 850

(109) Kurtén, T.; Tiusanen, K.; Roldin, P.; Rissanen, M.; Luy, J.-N.; Boy, M.; Ehn, M.; Donahue, N., α-pinene autoxidation products may not have extremely low saturation vapor pressures despite high O:C ratios. J. Phys. Chem. A 2016, 120 (16), 2569-2582, doi 10.1021/acs.jpca.6b02196.

35

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

851 852

TOC Graphic

36

ACS Paragon Plus Environment

Page 36 of 43

Hydroperoxyacids

Apparent κ (−)

OOH Page 37 The of Journal 43 of Physical Chemistry 1 1 O -1 10 3 OOH

3 O 1 5 2 OOH 5 O 3 4 tail length 5 ACS Paragon Plus Environment 6 1 2 3 4 5 6 7 8 7 10-2 10 12 14 16 8 Carbons (−)

O OH O OH O OH

9 10 11

Apparent κ (−)

10-1

Dihydroperoxides The Journal of Physical Chemistry Page 38 of 43 1

OOH

1

O

O HOO

1 HOO 2 2 2 O OOH 3 OOH O 4 3 O 3 HOO -2 5 10 6 tail length 7 ACS Paragon Plus Environment 1 2 3 4 5 6 8 10 12 14 16 9 Carbons (−) O

Apparent κ (−) PageA39 of 43 0.1

B

D

0. 0.1

1

ACS Paragon Plus Environment 0.1 100 200 300 50 100 200 300 50 Dd (nm) Dd (nm)

01

01

0.1

100 200 300 50 Dd (nm)

1

3 4

κ=

0.2 tail length

0.

0.5 0.4 0.3

-3

-3

C12

10

10

C10

1 0.05 2 0.04 3 0.03 4 0.02 5 6 7 0.01 8 50 9

sc (%) The Journal of Physical C 1 Chemistry

100 200 300 Dd (nm)

Apparent κ (−)

B 10-1 A The Journal of Physical Chemistry Page 40 of 43 C14

8 5 4

C14 4

-2 3 1 10 2 2 2 3 3 = tail length 3 -3 4 10 5 6 tail length ACS Paragon Plus Environment 3 4 5 7 9 7 -4 0 1 1 2 8 10 −C(=O)OH −C(OOH)−O− 9

O

2

O

O

O

1 10-2 1 2 3 4 -3 5 10 2 6 3 7 8 10-4 ACS Paragon Plus Environment 9 1 2 10 16 18 19 11 Carbons (−)

O

Apparent κ (−)

O

O

O

Carbonyls

10-141 Page The of Journal 43 of Physical Chemistry

3

Modeled κ (−)

1 A 10

B

9

12

-2

1 10-4 2 3 10-6 4 5 10-8 6 7 low 8 9

C The Journal of56Physical Chemistry 7 8

10

8 11

10

1

6

4 3

9

11

1 2 3 4

10

5 7 61 8 12

9 11

0.1

D

Page 425 of 43 12

6

7

2 43

1

8

-2

10

2

10-3

5 2 12

724

10-8 10-6 10-4 10-2 Observed κ (−)

1

ACS Paragon Plus Environment -8 10 10-6 10-4 10-2 1 10-8 10-6 10-4 10-2 1 Observed κ (−) Observed κ (−)

3 10 9

11

4

10-3 10-2 0.1 Observed κ (−)

New hygroscopicity measurements ➛➛ New interaction Page 43 The of Journal 43 of Physical Chemistry parameters

0.1

κ (-)

0

C10 12

-2

Before

log10 UNIFAC κ (−)

14 C18 1 16 0.01 2 C16 C10 12 14 16 3 After Carbonyls 4 -3 Hydroperoxyacids 10 Dihydroperoxides 5 ACS Paragon Plus Environment 6 C19 107-4 -4 -6 -8

24

0

-2 -4 -6 -8

low

-8

-6

-4

-2

0

log10 Measured κ (−)