Isoprene-Derived Organosulfates: Vibrational Mode Analysis by

Subscriber access provided by READING UNIV

Article

Isoprene-Derived Organosulfates: Vibrational Mode Analysis by Raman Spectroscopy, Acidity-Dependent Spectral Modes, and Observation in Individual Atmospheric Particles Amy Lynne Bondy, Rebecca Lynn Craig, Zhenfa Zhang, Avram Gold, Jason Douglas Surratt, and Andrew P Ault J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b10587 • Publication Date (Web): 08 Dec 2017 Downloaded from http://pubs.acs.org on December 8, 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 45 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

Isoprene-Derived Organosulfates: Vibrational Mode

2

Analysis by Raman Spectroscopy, Acidity-

3

Dependent Spectral Modes, and Observation in

4

Individual Atmospheric Particles

5

Amy L. Bondy, 1 Rebecca L. Craig, 1 Zhenfa Zhang,2 Avram Gold,2 Jason D. Surratt,2 Andrew P.

6

Ault 1,3*

7

1

Department of Chemistry, University of Michigan, Ann Arbor, Michigan, 48109

8

2

Department of Environmental Sciences and Engineering, Gillings School of Global Public

9

Health, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, 27599

10

3

11

48109

12 13 14

*corresponding author’s e-mail: [email protected]

15

Journal of Physical Chemistry A

Department of Environmental Health Sciences, University of Michigan, Ann Arbor, Michigan,

ACS Paragon Plus Environment

1

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

16

Page 2 of 45

Abstract

17

Isoprene, the most abundant biogenic volatile organic compound (BVOC) in the

18

atmosphere, and its low-volatility oxidation products lead to secondary organic aerosol (SOA)

19

formation. Isoprene-derived organosulfates formed from reactions of isoprene oxidation products

20

with sulfate in the particle phase are a significant component of SOA and can hydrolyze forming

21

polyols. Despite characterization by mass spectrometry, their basic structural and spectroscopic

22

properties remain poorly understood. Herein, Raman microspectroscopy and density functional

23

theory (DFT) calculations (CAM-B3LYP level of theory) were combined to analyze vibrational

24

modes of key organosulfates, 3-methyltetrol sulfate esters and 2-methylglyceric acid sulfate

25

esters (both racemic mixtures of two isomers), and hydrolysis products, 2-methyltetrols and 2-

26

methylglyceric acid. Two intense vibrational modes were identified, ν(RO-SO3) (846±4 cm–1)

27

and vs(SO3) (1065±2 cm–1), along with a lower intensity δ(SO3) mode (586±2 cm–1). For 2-

28

methylglyceric acid and its sulfate esters, deprotonation of the carboxylic acid at pH values

29

above the pKa decreased the carbonyl stretch frequency (1724 cm–1), while carboxylate modes

30

grew in for νs(COO–) and νa(COO–) at 1413 and 1594 cm–1, respectively. The ν(RO-SO3) and

31

vs(SO3) modes were observed in individual atmospheric particles and can be used in future

32

studies of complex SOA mixtures to distinguish organosulfates from inorganic sulfate or

33

hydrolysis products.

34


2

Page 3 of 45 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

35


Introduction

36

Isoprene is a biogenic volatile organic compound (BVOC) emitted by broadleaf trees and

37

is the largest global emission of all non-methane VOCs (~600 Tg y–1).1 Atmospheric oxidation

38

of isoprene leads to lower volatility products that partition to the particle phase, forming

39

secondary organic aerosol (SOA),2-5 which has been estimated to constitute 30% to 50% of the

40

global SOA budget.4 SOA-containing particles impact climate by scattering and absorbing solar

41

radiation or acting as cloud condensation nuclei (CCN).6 Recent work has also shown that

42

exposure of human lung cells to atmospherically relevant isoprene-derived SOA induces

43

oxidative stress.7-9 Furthermore, chronic obstructive pulmonary disease (COPD), which can

44

worsen due to oxidative stress in the human respiratory system,10-11 is higher in the southeastern

45

United States where isoprene-derived SOA contributes large mass fractions (up to 40%) of

46

submicron organic aerosol.12-14 Therefore, understanding the formation and evolution of

47

isoprene-derived SOA is important for understanding the overall impact of aerosols on climate

48

and health.

49

Organosulfates are a major class of compounds in SOA, estimated to contribute 5-10% of

50

the total organic aerosol mass over the continental US.15 Due to their high polarity and water

51

solubility, these hygroscopic compounds could enhance the CCN activity of organic aerosol.16

52

Isoprene-derived organosulfates formed from reactions of sulfate with isoprene oxidation

53

products in the particle phase are often reported as the most abundant organosulfates in ambient

54

aerosols,17-20 with mass concentrations of up to 8% of organic matter.21-24 The isoprene-derived

55

organosulfates detected in ambient aerosol formed under low-NOx conditions are the

56

diastereomeric methyltetrol sulfate ester racemates,17-23, 25-29 while under high-NOx conditions,

57

racemic 2-methylglyceric acid sulfate ester is also detected.17-18, 20-22, 25, 29 A simplified scheme of


3


Page 4 of 45

58

reactions leading to the formation of isoprene-derived organosulfates under low- and high-NOx

59

conditions is shown in Figure 1. The sulfate esters form via a nucleophilic oxirane ring-opening

60

reaction by sulfate. Sulfate attacks δ-isoprene epoxydiol (δ-IEPOX), an isomer of IEPOX

61

formed in significant proportion,30 under low-NOx conditions at the primary carbon (C1) to yield

62

racemic 3-methylerythritol sulfate ester ((2S,3R)/(2R,3S)-2,3,4-trihydroxy-3-methylbutyl sulfate)

63

and racemic 3-methylthreitol sulfate ester ((2R,3R)/(2S,3S)-2,3,4-trihydroxy-3-methylbutyl

64

sulfate).31-33 Two high NOx isoprene-oxidation products, methacrylic acid epoxide (MAE; 2-

65

methyoxirane-2-carboxylic acid) and hydroxymethyl-methyl-α-lactone (HMML), yield racemic

66

2-methylglyceric acid sulfate ester (2-carboxyl-2-hydroxylpropyl sulfate). Methyltetrol sulfate

67

esters have been observed in more than 65% of particles in the southeastern US and are among

68

the most abundant individual organic compounds in atmospheric aerosol.17, 19 Although previous

69

studies in China and the southeastern United States showed that 2-methylglyceric acid sulfate

70

ester is not as abundant in ambient aerosol as the methyltetrol sulfate esters, it nevertheless

71

accounted for ~0.5% of organic matter by mass.21-22 Isoprene-derived organosulfates have long

72

atmospheric lifetimes and thus are substantial contributors to ambient organic aerosol.31 The

73

sulfate esters can undergo hydrolysis within particles (hydrolysis lifetime, 60 to 460 h,

74

depending on particle acidity),31, 34 leading to the formation of the diastereomeric 2-methyltetrol

75

racemates (erythritol and threitol) and racemic 2-methylglyceric acid (2,3-dihydroxy-2-

76

methylpropanoic acid).22, 25, 31-32, 34-36 The prevalence of these organosulfates and their hydrolysis

77

products as well as their continuing chemistry in the aerosol phase, has recently elicited

78

considerable research interest.5, 20, 27, 29, 35, 37-39


4

Page 5 of 45 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


79 80 81 82 83 84 85

Figure 1. Scheme leading to the formation of isoprene-derived SOA compounds: 3-methyltetrol sulfate esters, 2-methyltetrols, 2-methylglyceric acid sulfate ester, and 2-methylglyceric acid. Both the high- and low-NOx pathways are shown.22, 32, 36 For simplicity, only one isomer of each respective compound is shown. 2-Methylglyceric acid, 2-methylglyceric acid sulfate ester, 2methyltetrols, and the 3-methyltetrol sulfate esters are present in the particle phase, while the epoxides are in the gas phase.

86

Currently, identification and quantitation of isoprene-derived organosulfates in individual

87

aerosol particles has been limited due to the complexity of SOA particles (thousands of species


5


Page 6 of 45

88

are often present in attoliter volumes), instrumental difficulties in differentiating organosulfates

89

from inorganic sulfate, and a lack of authentic standards. Most methods currently used to detect

90

organosulfate species have been offline mass spectrometry-focused, relying on filter extractions

91

followed by ultra-performance liquid chromatography/electrospray ionization high-resolution

92

quadrupole time-of-flight mass spectrometry (UPLC/ESI-HR-QTOFMS).20-22,

93

Online analysis has been attempted by Farmer et al., who used an online aerosol mass

94

spectrometer (AMS) to analyze an isoprene organosulfate surrogate standard, though due to

95

extensive fragmentation, this compound could not be differentiated from inorganic sulfate in

96

ambient aerosol.45 Some spectroscopic analysis has been done on organosulfates using Fourier

97

transform infrared spectroscopy,46-50 but identification has focused on a single lower frequency

98

vibration (C-S of methane sulfonic acid at 876 cm–1), rather than the entire fingerprint region,

99

making interpretation challenging. Although bulk methods are typically used to detect and

100

quantify organosulfates, they provide information limited to average aerosol composition and do

101

not provide information concerning the abundance of organosulfate-containing particles.

102

24-29, 32, 36, 40-44

Single particle mass spectrometry methods have been used to provide information

103

regarding the mixing state of organosulfates.

Hatch et al. detected isoprene-derived

104

organosulfates using aerosol time-of-flight mass spectrometry (ATOFMS), and observed high

105

fractions during two field studies in Atlanta, Georgia.17-18 Froyd et al. detected significant levels

106

of isoprene-derived organosulfates in the free troposphere by particle ablation laser mass

107

spectrometry (PALMS).19 These studies revealed that isoprene-derived organosulfates not only

108

account for a sizeable fraction of organic aerosol mass, but also are ubiquitous, present in more

109

than 70% of aerosols over the continental US.17, 19 However, issues such as fragmentation and


6

Page 7 of 45 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


110

shot-to-shot variability for laser desorption/ionization in both the ATOFMS and PALMS17, 19, 23

111

make quantifying organosulfates particularly challenging.

112

In contrast to bulk analysis and single particle mass spectrometry methods, Raman

113

microspectroscopy is a nondestructive technique that can be used to analyze the vibrational

114

modes of functional groups within individual particles.51-66 In addition to identifying vibrational

115

modes, Raman microspectroscopy is sensitive to changes in bonding and molecular environment,

116

such as aqueous versus solid particles67 or acidity-dependent protonation states.55,

117

Raman scattering, unlike mass spectrometry, is not dependent on particle ionization and

118

fragmentation patterns,19,

119

such as inorganic sulfate and organosulfates, since each class exhibits unique vibrational modes.

120

Previous Raman studies of ambient SOA identified nitrates, sulfates, carbonates, alcohols/water,

121

silicates and aluminosilicates, soot, hydrocarbons, and humic like substances within individual

122

particles.53, 56-57, 59-60 However, despite the rich vibrational spectra that Raman can provide, the

123

complexity of ambient SOA makes unambiguous identification of specific constituents

124

challenging. In order to identify isoprene-derived organosulfates within ambient SOA particles

125

using Raman spectroscopy, a thorough analysis of their vibrational modes is therefore necessary.

126

Herein,

the

45

58

Since

it can be used to differentiate classes of compounds within SOA,

Raman

vibrational

spectra

of

the

atmospherically-relevant

3-

127

methylerythritol/3-methylthreitol sulfate ester mixture derived from δ-IEPOX and 2-

128

methylglyceric acid sulfate ester derived from MAE and HMML are recorded and observed band

129

frequencies are compared to calculated frequencies. Raman spectra of the organosulfate

130

hydrolysis products, 2-methyltetrols and 2-methylglyceric acid, were examined to assist in the

131

identification of the organosulfate-related modes. Density functional theory (DFT), in

132

conjunction with published frequencies of small-molecule organosulfates, was used to predict


7


Page 8 of 45

133

optimized structures of the isoprene-derived organosulfates and to assign vibrational modes in

134

experimental Raman spectra. Acidity-dependent shifts of key modes were identified.

135

Furthermore, key organosulfate-related Raman modes were observed in ambient atmospheric

136

aerosol particles collected from the southeast United States. This study identifies signature

137

Raman vibrational modes for isoprene-derived organosulfates that can be used to identify these

138

compounds in chamber studies and ambient particles.

139

Experimental Methods

140

Reagents. 2-Methylglyceric acid, racemic 2-methylglyceric acid sulfate ester potassium

141

salt (81%), a diastereomeric mixture of racemic 2-methyltetrols (racemic 2-methylerythritol and

142

racemic 2-methylthreitol), and sodium salts of the 3-methyltetrol sulfate esters (86%) (racemic 3-

143

methylerythritol sulfate ester and racemic 3-methylthreitol sulfate ester) were synthesized by the

144

Surratt group and used without further purification.24,

145

methyltetrol sulfate esters and 2-methyltetrols are given in the Supporting Information (SI). 2-

146

Methylglyceric acid was synthesized by a published method.69 2-Methylglyceric acid sulfate

147

ester potassium salt was synthesized with a slight modification of the published method41 using

148

methylglyceric acid as the starting material. Trace quantities of inorganic sulfate are present in

149

the 3-methyltetrol and 2-methylglyceric acid sulfate esters. Target structures were verified by

150

proton nuclear magnetic resonance spectroscopy (1H NMR), Fourier transform infrared

151

spectroscopy (FTIR) and energy dispersive X-ray spectroscopy (EDX) (SI, Figures S1-S8).

41, 68-69

Synthetic details for the 3-

152

Raman Microspectroscopy. Aqueous 0.05 M solutions of 2-methylglyceric acid, 2-

153

methylglyceric acid sulfate ester, 2-methyltetrols, and 3-methyltetrol sulfate esters were prepared

154

by dissolution in 18.3 MΩ Milli-Q water. For Raman analysis, 2 µL droplets (~1 mm diameter)


8

Page 9 of 45 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


155

were deposited onto quartz substrates (Ted Pella). The relatively dilute solutions with low ionic

156

strengths used in this work most likely rules out non‐ideal solution effects on the Raman spectra.

157

Additionally, aerosol particles were generated by nebulizing aqueous 0.05 M solutions of 2-

158

methylglyceric acid sulfate ester and 3-methyltetrol sulfate esters onto quartz using a concentric

159

glass nebulizer (TR-30-A1, Meinhard). Particles from aerosolized compounds, with projected

160

area diameters ranging from ~3.5 to 9 µm, produced identical, but lower intensity Raman spectra

161

as large droplets of the aqueous solutions, therefore only Raman spectra of the organosulfate

162

solutions are shown in the subsequent analysis. See SI for Raman spectra of aerosolized

163

compounds and pure crystals/liquids (Figures S9-S12).

164

The standards were probed using a Raman microspectrometer (LabRAM HR Evolution,

165

HORIBA, Ltc.) at ambient temperature and relative humidity. The Raman spectrometer was

166

coupled with a confocal optical microscope (100x long working distance Olympus objective, 0.9

167

numerical aperture) and equipped with a Nd:YAG laser source (50 mW, 532 nm) operated with a

168

neutral density (ND) filter at 100% and CCD detector. The 1800 g/mm diffraction grating

169

yielded a spectral resolution of ~0.7 cm–1. The instrument was calibrated daily using a silicon

170

wafer standard against the Stokes Raman signal of pure Si at 520 cm–1. Spectra were collected

171

with three accumulations at 60-second acquisition times for the following spectral ranges: 500-

172

4000 cm–1 (2-methylglyceric acid sulfate ester and 2-methyltetrols), 500-2200 cm–1 and 2200-

173

4000 cm–1 (2-methylglyceric acid), 500-1800 cm–1 and 2500-4000 (3-methyltetrol sulfate esters).

174

For 2-methylglyceric acid and the 3-methyltetrol sulfate esters, the lower frequency ranges (500-

175

2200 cm–1 and 500-1800 cm–1, respectively) and the higher frequency range (2500-4000 cm–1)

176

were collected separately using the same number of accumulations and acquisition time

177

described previously to minimize water evaporation.


9


Page 10 of 45

178

To explore the nature of the carboxylic acid vibrational mode and identify the cause of

179

spectral variation between 2-methylglyceric acid sulfate ester and 2-methylglyceric acid, the

180

effect of pH was studied. Assuming a similar pKa for 2-methylglyceric acid compared to glyceric

181

acid (pKa glyceric acid ~3.5), the pH of an aqueous solution of 2-methylglyceric acid was

182

adjusted from 1.3 to approximately 1.8, 3, and 10 through the addition of NaOH, measured using

183

pH paper. Raman spectra were collected of the aqueous solutions of 2-methylglyceric acid at

184

each pH from 500 to 4000 cm–1 using three accumulations, 10-second acquisition times, and a

185

600 g/mm diffraction grating with a spectral resolution of ~1.7 cm–1.

186

Density Functional Theory Calculations. Geometry optimization, Raman shift, and

187

Raman scattering activity calculations were performed using Gaussian 09W.70 Initial geometry

188

optimizations were performed using DFT with the CAM-B3LYP functional and 3-21G basis set.

189

An additional geometry optimization was performed and Raman vibrational mode frequencies

190

were calculated at the DFT CAM-B3LYP level of theory with the 6-311 ++ G(2d,p) basis set,

191

using an ultrafine pruned grid and tight optimization criteria. All calculations are unscaled and

192

run using water as a solvent. Calculated Raman frequencies and activity were used, in

193

conjunction with literature,71-77 to assign vibrational modes to the experimental Raman spectra

194

collected for each compound. The DFT calculations for 2-methylglyceric acid sulfate ester, 3-

195

methylerythritol sulfate ester, and 3-methylthreitol sulfate ester were performed with a -1 charge

196

on the compounds. Since the DFT calculated frequencies for the sulfate functional group-related

197

modes deviate from literature frequencies, multiple functionals and basis sets were tested for

198

methyl sulfate, a model compound (Table S1), and CAM-B3LYP was selected for the following

199

analysis as a compromise between accuracy and calculation expense.


10

Page 11 of 45 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


200

Ambient Aerosol Collection. Ambient particles from the Southern Oxidant and Aerosol

201

Study (SOAS) during the summer of 2013 were analyzed for isoprene-derived organosulfate

202

Raman signatures. Description of the site and particle collection is provided elsewhere.78-80

203

Briefly, particles were impacted on quartz substrates (Ted Pella Inc.) using a micro-orifice

204

uniform deposit impactor (MOUDI, Model 110, MSP Corp.) in Centreville, AL, a rural, forested

205

region near Talladega National Forest with high isoprene emissions. After collection, all

206

substrates were sealed and stored at -22 °C prior to analysis. Individual particles were analyzed

207

using computer controlled-Raman microspectroscopy (CC-Raman), according to the method

208

described previously by Craig et al.54

209

Results and Discussion

210

To identify Raman signatures for isoprene-derived organosulfates, differences between

211

the experimental spectra of the organosulfates and their hydrolysis products were noted. Figure 2

212

shows experimental Raman spectra of the organosulfates and their hydrolysis products overlaid

213

with intensities normalized to the modes at 1460 cm–1 and 2950 cm–1 (the most intense modes in

214

the low- and high-frequency regions, respectively, excluding possible sulfate modes). Distinct

215

modes present in the spectra of both organosulfates at ~1065 cm–1 and ~850 cm–1, assigned to the

216

organosulfate functional group, are highlighted. Aside from the sulfate modes, key spectral

217

differences include a very intense inorganic sulfate mode (from synthesis or hydrolysis) present

218

in the spectrum of the 3-methyltetrol sulfate ester mixture at 982 cm–1, and differences in the

219

carbonyl stretching region (1400-1750 cm–1) for 2-methylglyceric acid sulfate ester and 2-

220

methylglyceric acid which will be explored in more detail below.


11


Page 12 of 45

221 222 223

Figure 2. Experimental Raman spectra of a) the 3-methyltetrol sulfate esters overlaid with 2methyltetrols, and b) 2-methylglyceric acid sulfate ester overlaid with 2-methylglyceric acid.

224

To correlate the spectral differences observed in Figure 2 with the presence of the

225

organosulfate functional group, quantum chemical calculations were performed for the 3-

226

methyltetrol sulfate esters, 2-methyltetrols, 2-methylglyceric acid sulfate ester, and 2-

227

methylglyceric acid. Figure 3a shows experimental Raman modes within the fingerprint region

228

(500-1500 cm–1) and at higher frequencies (2700-3500 cm–1) for the mixture of diastereomeric 3-

229

methyltetrol sulfate esters in an aqueous solution. Using DFT, the Raman activity (Figure 3b)

230

was calculated for the minimum-energy geometry of each diastereomer (Figure 3c and 3d). The

231

optimal geometries, with the terminal (C4) hydroxyl group twisted toward the sulfate group,

232

were used in the subsequent analysis, although the energies were not substantially different for

233

other conformers (difference

Isoprene-Derived Organosulfates: Vibrational Mode Analysis by

Recommend Documents