Molecular-Scale Investigation with ESI-FT-ICR-MS ... - ACS Publications

Jan 27, 2016 - aromaticity index (AI) and H/C cutoffs according to Kellerman et al. ..... the abundance of the individual peak divided by the maximum...
0 downloads 0 Views 1MB Size
Subscriber access provided by NATIONAL UNIV OF SINGAPORE

Article

Molecular-Scale Investigation with ESI-FT-ICR-MS on Fractionation of Dissolved Organic Matter Induced by Adsorption on Iron Oxyhydroxides Jitao Lv, Shuzhen Zhang, Songshan Wang, Lei Luo, Dong Cao, and Peter Christie Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b04996 • Publication Date (Web): 27 Jan 2016 Downloaded from http://pubs.acs.org on January 28, 2016

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

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

Page 1 of 35

Environmental Science & Technology

1

Molecular-Scale Investigation with ESI-FT-ICR-MS on Fractionation of Dissolved

2

Organic Matter Induced by Adsorption on Iron Oxyhydroxides

3 4

Jitao Lva, Shuzhen Zhanga*, Songshan Wanga, Lei Luoa, Dong Caoa, Peter Christieb

5

a

6

Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing

7

100085, China

8

b

9

Belfast BT9 5PX, United Kingdom

State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research

Agri-Environment Branch, Agri-Food and Biosciences Institute, Newforge Lane,

10

ACS Paragon Plus Environment

Environmental Science & Technology

Page 2 of 35

11

ABSTRACT

12

Adsorption by minerals is a common geochemical process of dissolved organic matter

13

(DOM) which may induce fractionation of DOM at the mineral-water interface. Here,

14

we examine the molecular fractionation of DOM induced by adsorption onto three

15

common

16

Fourier-transform ion cyclotron resonance mass spectrometry (ESI-FT-ICR-MS).

17

Ferrihydrite exhibited higher affinity to DOM and induced more pronounced

18

molecular fractionation of DOM than did goethite or lepidocrocite. High molecular

19

weight (> 500 Da) compounds and compounds high in unsaturation or rich in oxygen

20

including polycyclic aromatics, polyphenols and carboxylic compounds had higher

21

affinity to iron oxyhydroxides and especially to ferrihydrite. Low molecular weight

22

compounds and compounds low in unsaturation or containing few oxygenated groups

23

(mainly alcohols and ethers) were preferentially maintained in solution. This study

24

confirms that the double bond equivalence and the number of oxygen atoms are

25

valuable parameters indicating the selective fractionation of DOM at mineral and

26

water interfaces. The results of this study provide important information for further

27

understanding the behavior of DOM in the natural environment.

28

Keywords: Dissolved organic matter (DOM), iron oxyhydroxides, fractionation,

29

ESI-FT-ICR-MS

iron

oxyhydroxides

using

electrospray

ionization

30

ACS Paragon Plus Environment

coupled

with

Page 3 of 35

Environmental Science & Technology

31

INTRODUCTION

32

Dissolved organic matter (DOM) is ubiquitous in the environment and many chemical

33

reactions such as complexation, redox reactions and photodegradation occur between

34

contaminants and DOM1-3. DOM therefore plays a role as a microreactor of

35

contaminants in the natural environment4, and its important influences on the

36

environmental behavior of contaminants have been widely reported. DOM has

37

extremely complicated constituents depending on various sources and surrounding

38

environments, and thus its molecular characterization is still elusive and interactions

39

between DOM and environmental matrices are poorly understood.

40

Interaction between natural organic matter (NOM) and minerals is widespread in

41

both aquatic and terrestrial environments and acts as a key factor in the long-term

42

storage of organic carbon, thus contributing to the global cycles of carbon5, 6. It also

43

modifies the mineral surface properties and reactivity which further influence the

44

transport and fate of many organic and inorganic contaminants and nutrients7-9.

45

According to the supramolecular association concept, DOM is composed of numerous

46

small molecular components which are self-assembled by hydrophobic interactions

47

and hydrogen bonds10, 11. We therefore hypothesize that such associations may be

48

destroyed when external forces from mineral surfaces to some small molecules are

49

greater than the intermolecular forces between the molecular components of DOM.

50

Molecules having higher affinity to mineral surfaces will therefore be sequestered

51

while other molecule associations remain in water, leading to the molecular

52

fractionation of DOM at the mineral-water interface. Previous studies have used

53

spectroscopic

54

Fourier-transform infrared (FTIR), nuclear magnetic resonance (NMR) and C K-edge

55

near-edge X-ray absorption fine structure (NEXAFS) spectroscopy to investigate the

methods

such

as

UV-visible

absorption

ACS Paragon Plus Environment

spectrum

(UV-vis),

Environmental Science & Technology

56

interactions between various DOM and minerals12-16. In general, hydrophobic,

57

aromatic, and carboxylic rich fractions are preferentially retained on mineral surfaces.

58

Spectroscopic methods provide insight into the features and relative abundance of

59

chemical groups such as carboxylic, aromatic and aliphatic moieties in DOM, but are

60

more limited in deriving information at the molecular level17. High-pressure size

61

exclusion chromatography (HPSEC) has been used to investigate size fractionation

62

resulting from the adsorption of DOM on minerals, and preferential adsorption of

63

high molecular weight (HMW) fractions has been reported12, 18. However, due to the

64

limitations of the method, the separated fractions are small clusters instead of single

65

molecules, and the molecular components therefore remain unknown. Reiller et al.19

66

made the first attempt to investigate fractionation of humic acid (HA) on hematite

67

induced by sorption using mass spectrometry. Although the exact molecular formulas

68

of HA were not given due to the poor mass resolution of electrospray ionization (ESI)

69

quadrupole time-of-flightmass spectra (QToF-MS), their work provides a new

70

approach to study the fractionation of DOM induced by sorption.

71

Recent advances have been made in high-resolution mass spectrum techniques,

72

particularly electrospray ionization coupled with Fourier-transform ion cyclotron

73

resonance mass spectrometry (ESI FT-ICR MS) which has been shown to be a very

74

powerful tool for the analysis of DOM at the molecular level. With the help of this

75

technique, thousands of molecular formulas in DOM can be obtained unambiguously

76

and this has greatly improved our understanding of the chemical nature of DOM at the

77

molecular level20-22. Very recently, Riedel et al.23 investigated the molecular

78

fractionation of peat leachate (PL), a type of DOM, coagulated by metal salts such as

79

Ca2+, Al3+ and Fe3+ using ESI-FT-ICR-MS. Results indicate that Fe3+ and Al3+

80

preferentially removed the most oxidized compounds (O/C ratio > 0.4) of DOM,

ACS Paragon Plus Environment

Page 4 of 35

Page 5 of 35

Environmental Science & Technology

81

while Ca2+ displayed no preference for distinct molecular fractions. Galindo et al.17,24

82

published two papers reporting the sorptive fractionation of aquatic fulvic acid (FA)

83

and terrestrial humic acid (HA) on aluminum oxide by ESI-FTMS using a linear ion

84

trap Orbitrap mass spectrometer. They demonstrated that molecule acidity and

85

hydrophobicity were the determining factors that controlled the adsorptive

86

fractionation of FA and HA on aluminum oxide. As far as we know, these are the only

87

two papers until now that describe the fractionation of DOM due to its adsorption on

88

minerals.

89

Fe oxyhydroxides are minerals that exhibit a great affinity for DOM9,

25, 26

.

90

Interactions between Fe oxyhydroxides and DOM have been widely studied27-30, but

91

research focusing on the fractionation of DOM on Fe oxyhydroxides at the molecular

92

level is still lacking. Fe oxyhydroxides have various crystal structures and often

93

display different thermodynamic stability and reaction activity31. We thus hypothesize

94

that adsorption of DOM on different types of Fe oxyhydroxides will induce distinct

95

molecular fractionation. Three common Fe oxyhydroxides, namely poorly crystalline

96

ferrihydrite and crystalline goethite and lepidocrocite, were therefore selected in the

97

present study to investigate the molecular fractionation of DOM at Fe oxyhydroxide

98

and water interfaces using ESI-FT-ICR-MS.

99 100

MATERIALS AND METHODS

101

Preparation and characterization of iron oxyhydroxides and dissolved organic

102

matter. The iron oxyhydroxides ferrihydrite, goethite and lepidocrocite were

103

synthesized according to methods described in the literature32,

104

provided in the Supporting Information (SI). Synthesized Fe oxyhydroxides were

105

repeatedly washed by centrifugation and resuspension in milliQ water and kept at

ACS Paragon Plus Environment

33

and details are

Environmental Science & Technology

106

4 °C. Transmission electron microscopy (TEM) and X-ray diffraction (XRD) patterns

107

of the three iron oxyhydroxides (Figure S1 in SI) indicate that the synthesized iron

108

oxyhydroxides were consistent with standard 2-line ferrihydrite, goethite and

109

lepidocrocite, respectively. The N2 Brunauer-Emmett-Teller (BET) specific surface

110

areas (SSA) and pore volumes of the ferrihydrite, goethite and lepidocrocite were

111

analyzed by nitrogen adsorption-desorption (Micromeritics ASAP 2460, USA). The

112

density of surface hydroxyl groups of iron oxyhydroxidesin water was measured

113

according to a saturated deprotonation method described in the literature34. The

114

surface characteristics of the iron oxyhydroxides are presented in Table S1.

115

DOM was extracted from Danish peat (supplied by Northeast China Normal

116

University). The composition of the original peat was 45.9 % total C, 0.68 % total N,

117

and 5.64 % total H as previously reported35. The extract was obtained by

118

end-over-end shaking 10 g of peat in 100 mL ultrapure water for 24 h. The

119

supernatant was obtained by centrifugation at 5000 rpm and filtered through 0.22-µm

120

Nylon filters. The pH of the final extract was 4.8 because of the acidic character of the

121

water-soluble compounds in peat. This DOM from PL was kept in the dark at 4 °C

122

until further use.

123

Sorption experiments. Batch experiments for the sorption of PL (0-100 mg L-1)

124

onto iron oxyhydroxides (0.82 g L-1ferrihydrite, 1.66 g L-1 goethite, 1.65 g L-1

125

lepidocrocite) without background electrolyte were conducted in glass centrifuge

126

tubes. The initial pH was adjusted to 7.0 and no further pH adjustment was made

127

during the experiment because there was only a slight change in pH value (less than

128

0.5 units) throughout the experiment. The suspensions were equilibrated on a

129

reciprocating shaker at 120 rpm at 25 °C for 24 h in the dark. Kinetic measurements

130

show that 24 h was a sufficient time period for PL to reach an equilibrium adsorption

ACS Paragon Plus Environment

Page 6 of 35

Page 7 of 35

Environmental Science & Technology

131

on all three iron oxyhydroxides (Figure S2A). After equilibration the samples were

132

centrifuged for 40 min at 5000 rpm and filtered through 0.22-µm Nylon filters.

133

UV-vis spectroscopy analysis of the filtered supernatants was conducted with a

134

Shimadzu UV-2600 UV-vis spectrometer. The concentration of organic carbon in the

135

filtered supernatants was determined with a TOC analyzer (Vario TOC, Elementar,

136

Hanau, Germany). The amount of PL retained by the iron oxyhydroxides was

137

calculated by the difference between concentrations in the initial and equilibrium

138

solutions. 50 mg L-1 PL solution without addition of iron oxyhydroxides was used as

139

the control. Experiments were carried out with three replicates. The ζ potentials of

140

iron oxyhydroxides before and after adsorption of PL were measured at different pH

141

values with a Malvern Nano ZS (Malvern Instruments, UK). FTIR spectra of selected

142

freeze-dried solid samples were recorded in the range 4000-600 cm−1 using a

143

NICOLET 10 MX micro-infrared spectrometer (Thermo Scientific, Waltham, MA).

144

ESI-FT-ICR-MS Analysis. Selected filtered supernatants of the sorption samples

145

(circles in Figure 1) and control samples were analyzed by ESI-FT-ICR-MS. Solid

146

phase extraction (SPE) was performed on all the samples to desalinate and

147

concentrate according to Dittmar et al.36 using Varian Bond Elute PPL cartridges (1 g

148

per 6 mL). Briefly, the cartridges were rinsed with 6 mL of methanol (MS grade) prior

149

to use. The samples were acidified to pH 2 with HCl (32%, ultrapure) to increase the

150

extraction efficiency for organic acids and phenols, and 20−50 mL (depending on

151

DOC concentration) were passed through the cartridges by gravity at a flow rate of

152

approximately 2 mL min-1. Cartridges were rinsed with three volumes of 0.01 M HCl

153

for removal of salts, dried with a stream of N2 and immediately extracted with two

154

volumes of methanol (MS grade). Eluted samples were blow-dried with N2 and kept

155

in the dark at -20 °C. These dried samples were dissolved in 1 mL of 50:50

ACS Paragon Plus Environment

Environmental Science & Technology

156

methanol/water (v/v) for ultrahigh resolution mass spectrometry analysis. The

157

extraction efficiencies of PL exceed 75% (TOC) using this method.

158

Ultrahigh resolution mass spectra were acquired using a Bruker Solari

159

XFT-ICR-MS equipped with a 15.0 T superconducting magnet and an ESI ion

160

source37. Samples for ESI FT-ICR-MS analysis were continuously infused into the

161

ESI unit by syringe infusion at a flow rate of 120 µL h-1. The ESI needle voltage was

162

set to -3.8 kV. All the samples were analyzed in negative ionization mode with

163

broadband detection. Ions accumulated in a hexapol ion trap for 0.2 s before being

164

introduced into the ICR cell. 4M words of data were recorded per broadband mass

165

scan. The lower mass limit was set to m/z = 200 Da and the upper mass limit to m/z =

166

1000 Da. 100 mass spectra were averaged per sample. The spectra were externally

167

calibrated with 10 mM of sodium formate solution in 50% isopropyl alcohol using a

168

linear calibration and then internally recalibrated using an in-house reference mass list.

169

After internal calibration the mass error was < 500 ppb over the entire mass range.

170

Peaks were identified with Bruker Data Analysis software.

171

Molecular formula assignment. All possible formulas were calculated with

172

Formula Calculator software based on the requirement that the mass error for a given

173

chemical formula between measured mass and calculated mass was < 0.5 ppm.

174

Molecular formulas were assigned to peaks with a signal-to-noise ratio (S/N) > 4,

175

according to stringent criteria with elemental combinations of C5–80H10–200O0–40N0-2.

176

Elements P and S were excluded because of the low levels found in PL (< 0.5%). The

177

elemental ratios of H/C < 2.0 and O/C < 1.2 were used as further restrictions for

178

formula calculation. The following parameters for data analysis were calculated

179

according to previous reports38-42: H/C ratio and O/C ratio, parameters for

180

constructing van Krevelen diagram39; Kendrick mass (KM) and Kendrick mass defect

ACS Paragon Plus Environment

Page 8 of 35

Page 9 of 35

Environmental Science & Technology

181

(KMD) in order to identify CH2-homologous molecular formulas series38; and double

182

bond equivalence (DBE), a measure of the number of double bonds and rings in a

183

molecule40. An aromaticity index was calculated from the formulas according to Koch

184

and Dittmar41 to estimate the fraction of aromatic and condensed aromatic structures.

185

Furthermore, four compound groups were delineated by the aromaticity index (AI)

186

and H/C cutoffs according to Kellerman et al.42: combustion-derived condensed

187

polycyclic aromatics (AI>0.66), vascular plant-derived polyphenols (0.66≥AI>0.5),

188

highly unsaturated and phenolic compounds (AI≤0.50 and H/C<1.5), and aliphatic

189

compounds (2.0≥H/C≥1.5). The calculated formulas of relevant parameters are

190

provided in the SI.

191 192

RESULTS AND DISCUSSION

193

Adsorption Isotherms of PL on Iron Oxyhydroxides and UV Spectrum Analysis.

194

Adsorption isotherms for PL on ferrihydrite, goethite and lepidocrociteare shown in

195

Figure 1A. Typical adsorption isotherms of DOM on iron oxide surfaces showed an

196

initial steep slope and reached a plateau at elevated equilibrium C concentrations7.

197

However, aplateau of PL adsorbed on sorbent was not observed for all the three iron

198

oxyhydroxides, indicating that saturation adsorption of PL was not reached in the PL

199

concentration range employed (0-100 mg CL-1) (Figure 1). As shown in Figures 1,

200

goethite and lepidocrocite exhibited similar adsorption capacity to PL, but much

201

lower than that of ferrihydrite. This is consistent with the conclusion obtained from

202

previous studies that DOM preferentially interacts with poorly crystalline iron

203

oxyhydroxides in soils43, 44. Studies have demonstrated that the adsorption of DOM on

204

minerals is predominantly via ligand exchange and/or electrostatic interactions with

205

the surface hydroxylated groups18, 45. Ferrihydrite is the hydroxylation product of

ACS Paragon Plus Environment

Environmental Science & Technology

206

ferric ions with highly hydrated structure which has higher surface area and contains

207

more surface hydroxyl groups than goethite or lepidocrocite46 (data presented in Table

208

S1). The high density of surface hydroxylated groups of ferrihydrite therefore

209

contributed to the higher adsorption capacity of DOM compared with goethite or

210

lepidocrocite. Interaction with DOM also changes the surface charge of minerals. The

211

isoelectric points of goethite and lepidocrocite decreased from ~8.2 and ~9.2,

212

respectively, to about ~2.0 while only a slight decrease was found for ferrihydrite after

213

the adsorption of PL (from ~7.6 to ~7.0, as shown in Figure S3). This indicates that

214

the surface hydroxyl sites in goethite and lepidocrocite were almost completely

215

occupied by the negative DOM fractions, whereas many of those in ferrihydrite

216

remained unoccupied under the experimental conditions.

217

Specific UV absorbance at 254 nm (SUVA254, absorbance per mg-C) can be used as

218

an index of relative aromaticity of DOM14. Figure 1B shows that the SUVA254 of the

219

PL supernatant after adsorption on ferrihydrite decreased greatly for all the

220

PL-adsorbed samples and remained almost unchanged (about 0.1 L mg C-1m-1) as the

221

initial C concentration (C0) increased from 5 to 100 mg C L-1. This low SUVA254

222

suggests that a large amount of aromatic carbons in PL were removed, indicating that

223

adsorption of PL on ferrihydrite resulted in its fractionation between aromatic

224

moieties and other C containing moieties without SUVA254. The SUVA254 of the PL

225

supernatant after adsorption on goethite or lepidocrocite first decreased at low

226

concentration (C0<10 mg C L-1) and then continued to increase with increasing C

227

concentration, and was much higher than the corresponding value for the adsorption

228

of PL on ferrihydrite, indicating a similar but much poorer fractionation of aromatic

229

moieties in DOM after adsorption ongoethite and lepidocrocite. These results suggest

230

component fractionation of PL during its adsorption on the iron oxyhydroxides,

ACS Paragon Plus Environment

Page 10 of 35

Page 11 of 35

Environmental Science & Technology

231

especially on ferrihydrite. Detailed fractionations of PL were further investigated at

232

molecular scale using ESI FT-ICR-MS.

233

Molecular Characterization of PL by ESI FT-ICR-MS. The original PL and PL

234

supernatants extracted from the adsorption isotherm experiments (as shown in Figure

235

1A) for ferrihydrite (F1, F2), goethite (G1, G2) and lepidocrocite (L1, L2) were

236

determined by ESI FT-ICR-MS. As shown in Figure 2A, the ESI FT-ICR negative-ion

237

mass spectrum of the original PL contained 9472 peaks with S/N > 4 in the mass

238

range 200 to 800 Da and > 90% of the peaks were detected below 600 Da. Based on

239

the occurrence of carbon isotope peak twins with a distance of 1.0034/z, nearly all

240

ions were singly charged23. This molecular weight distribution is consistent with the

241

supramolecular association concept of humic substances in which many relatively

242

small and chemically diverse organic molecules form clusters linked by hydrogen

243

bonds and hydrophobic interactions10,

244

assigned to the peaks after quality assurance filtering, and 1923 molecular formulas

245

were identified in the original PL, among which only ~10% of assigned molecular

246

formulas contained one or two N atoms and the relative abundances of these peaks

247

were relatively low. This agrees well with previous studies of riverine DOM extracted

248

by C18 solid phase disk in which > 90% of the compound formulas obtained by

249

negative-mode ESI FT-ICR MS were assigned to CcHhOo47. Considering the low level

250

of N in PL, the following analysis will focus on compounds containing only C, H and

251

O. The CH2-based Kendrick mass analysis (Figure S4) reveals that PL was composed

252

of multitudinous -CH2 homologous series, suggesting that CH2 groups were the

253

building blocks of PL, which is consistent with the classic structure of natural DOM20.

254

Magnitude-weighted average composition of the original PL was C20.3H19.1O11.6 with

255

magnitude-averaged H/Cw, O/Cw and DBEw values of 0.94, 0.58 and 11.7,

11

. Molecular formulas CcHhNnOo were

ACS Paragon Plus Environment

Environmental Science & Technology

Page 12 of 35

256

respectively. More oxygen-rich and unsaturated molecules were present in the PL

257

compared to the riverine and offshore coastal DOM reported by Sleighter and

258

Hatcher47. Calculation formulas of Cw, Ow, Hw, H/Cw, O/Cw and DBEw according to

259

Sleighter and Hatcher47 are provided in the SI.

260

Molecular Weight Fractionation of PL Induced by Adsorption. Molecular weight

261

fractionation of PL was evidenced comparing the ESI-FT-ICR-MS spectrum between

262

the original PL and PL supernatant after adsorption on the iron oxyhydroxides with

263

the most pronounced for ferrihydrite (Figure 2B). The molecular weight (MW) above

264

500 decreased from 39.8% for the original PL to 5.7% (F1) and 13.7% (F2), while

265

MW below 400 increased from 24.2% (original PL) to 61.5% (F1) and 44.3% (F2). A

266

similar change was induced by adsorption on goethite and lepidocrocite, but to a

267

much lesser extent. The MW above 500 decreased from only 39.8% to 31.2% (G1)

268

and 28.8 % (L1), respectively, and no appreciable changes were observed for G2 and

269

L2. This is ascribed to the finite adsorptive capacities of goethite and lepidocrocite for

270

PL (as shown in Figure 1), making the differences between the original PL and PL

271

supernatant decrease with increasing equilibrium concentration. In general, molecules

272

with higher MW were preferentially adsorbed on the surfaces of iron oxyhydroxides,

273

and this trend is much more noticeable for ferrihydrite than for goethite or

274

lepidocrocite.

275

chromatography with UV detection (SEC-UV) that higher MW or intermediate MW

276

fractions of DOM are preferentially adsorbed on minerals. For example, Davis and

277

Gloor48 found that the DOM fraction with MW 1000–3000 Da was the most strongly

278

adsorbed fraction among their three fractions of <500, 4000–500, and >4000 Da.

279

Zhou et al.12 observed preferential adsorption of an intermediate MW fulvic acid

280

fraction (MW 1250–3750 Da) on goethite. Similar results were reported on the

Some

previous

studies

have

observed

ACS Paragon Plus Environment

by

size-exclusion

Page 13 of 35

Environmental Science & Technology

281

adsorption of Suwannee River fulvic acid (mean MW = 2180 ± 14 Da) on kaolinite

282

and hematite18. It is necessary to note that the measured MW of DOM fractions by

283

SEC are much larger than that obtained by ESI FT-ICR MS in the present study. The

284

fractions separated by SEC are more likely molecular clusters but not individual

285

molecules. In addition, only fractions with UV absorbance such as aromatic groups

286

can be detected. Therefore, the results obtained by SEC-UV are more likely size

287

fractionation of aromatic associates in DOM but not MW fractionation. To our

288

knowledge, this is the first study to reveal the detailed MW fractionation of DOM

289

induced by adsorption on iron oxyhydroxides using ESI FT-ICR MS.

290

Molecular Component Fractionation of PL Induced by Adsorption. The van

291

Krevelen diagram is a useful tool which can provide a visual graphic display of

292

compound distribution39. Compounds such as aromatics, polyphenols and aliphatics

293

are located in different regions of the van Krevelen diagram. Aromaticity Index (AI)

294

has been widely used to evaluate the aromaticity of a given molecular formula41. By

295

using AI and element ratios of H/C, all the detected molecules displayed in the van

296

Krevelen diagram were divided into four groups according to their molecular

297

components as described in the method (as shown in Figure 3A-E). For the original

298

PL the majority (~ 85%) of the compounds were polyphenols (18%, group 2) and

299

highly unsaturated organic acids and phenolics (67%, group 3), typical of soil-derived

300

DOM in rivers49. The proportions of condensed polycyclic aromatics (group 1) and

301

aliphatic compounds (group 4) were 4.9% and 9.7%, respectively. Ferrihydrite

302

exhibited more pronounced difference in affinity to different groups of the compounds

303

and therefore resulted in more pronounced group fractionation at its interface than

304

goethite or lepidocrocite. After adsorption on ferrihydrite, the chemical constituents of

305

PL supernatant changed, and about 62 and 43% of the total compounds in the original

ACS Paragon Plus Environment

Environmental Science & Technology

306

PL did not further exist in the PL supernatant (F1 and F2). Specifically, about 100 and

307

96% of the compounds in group 1, 88 and 77 % of the compounds in group 2, 57 and

308

35% of the compounds in group 3, and 33 and 8% of the compounds in group 4 in the

309

original PL disappeared in the PL supernatant (F1 and F2), indicating that the

310

affinities of the compounds in these groups to ferrihydrite were in the order: group 1

311

> group 2 > group 3 > group 4 (as shown in Figure 3F), whereas the differences

312

in compounds between the original PL and PL supernatant in each group decreased in

313

the inverse order. This further indicates the preferential adsorption of aromatics and

314

then polyphenols, while aliphatic compounds were preferentially retained in the

315

aqueous phase. Moreover, the percentages of the compounds in groups 1 and 2 in the

316

PL supernatant increased and the percentages of the compounds in groups 3 and 4

317

decreased and this may explain the reduction in SUAV254 of PL after adsorption on

318

ferrihydrite.

319

Compared with ferrihydrite, the differences in fractionations between groups due to

320

adsorption were much smaller for goethite and lepidocrocite. Furthermore, the

321

fractionation behavior of PL due to adsorption on lepidocrocite was very similar to

322

that for goethite, and therefore will not be described here. After adsorption on goethite,

323

about 32 and 10% of the total compounds of the original PL were undetectable in the

324

PL supernatant (G1 and G2), in which 55 and 24% of the compounds in group 1, 34

325

and 11% in group 2, 27 and 6% in group 3, and 50 and 28% in group 4 were

326

sequestered by goethite. In addition to polycyclic aromatic compounds, aliphatic

327

compounds were preferentially retained on the surfaces of goethite and this was

328

confirmed by the presence of a strong infrared absorption band at ~2920 cm-1 assigned

329

to C–H stretching of –CH3 and –CH2 groups13 (as shown in Figure S5). This may be

330

attributed to hydrophobic interactions between aliphatic compounds in PL supernatant

ACS Paragon Plus Environment

Page 14 of 35

Page 15 of 35

Environmental Science & Technology

331

and hydrophobic fractions adsorbed on goethite surfaces. Although the major sorption

332

contribution would be ligand exchange45, 50, studies have demonstrated the potential

333

importance of hydrophobic interactions between sorbed organic molecules and

334

organic molecules in solution via association of the hydrophobic portions of the

335

molecules18,

336

interaction would increase with increasing occupation of surface hydroxyl sites on

337

goethite by DOM components due to the decrease in available surface hydroxylsites.

338

As analyzed by adsorption isotherms and surface charge, the surface coverage of PL

339

fractions on goethite was much higher than that of ferrihydrite under the experimental

340

conditions. Thus, the contribution of hydrophobic interactions to the adsorption of PL

341

was greater on goethite than on ferrihydrite. This hydrophobic interaction was less

342

specific and reduced the adsorptive fractionation effect of PL.

50

. We assume that the relative contribution of this hydrophobic

343

Molecules present in the original PL but absent from the supernatant after

344

adsorption were considered to be completely sequestered by iron oxyhydroxides.

345

These molecules were extracted by subtracting the supernatant molecules from the

346

original molecules, and the quantitative distributions of O/C, H/C, and DBE values

347

for the molecules in the original PL, PL supernatant and sequestered PL are shown in

348

Figure 4. This mode of data presentation is advantageous for displaying the

349

abundance of compounds and provides a visualized picture for molecular

350

fractionation of PL at the mineral-water interface. It is evident that PL supernatant

351

was rich in compounds with low O/C, low DBE and high H/C, while the majority of

352

PL sequestered by ferrihydrite comprised compounds with high O/C, high DBE and

353

low H/C. With increasing equilibrium concentration, the number of molecules

354

remaining in PL supernatant increased but their relative distribution remained almost

355

unchanged. The H/C and DBE values are directly related to the saturation level of a

ACS Paragon Plus Environment

Environmental Science & Technology

356

compound. This indicates that compounds with lower unsaturation level were likely

357

rich in aliphatic groups and preferentially retained in the aqueous phase, while

358

oxidized compounds with higher unsaturation level were preferentially sequestered by

359

ferrihydrite. In contrast, slight differences in the relative distributions of these

360

parameters were observed between the molecules in the PL supernatant and

361

sequestered molecules after adsorption on goethite, although the number of molecules

362

decreased markedly, further indicating the less selective adsorption of PL on goethite

363

than on ferrihydrite. In order to further explain this difference, the adsorption contents

364

of DOM normalized to the reactive surface hydroxyls were calculated (Figure S2B).

365

As shown in Figure S2B, more DOM was adsorbed on the unit surface hydroxyl in

366

ferrihydrite compared with goethite, especially when the equilibrium concentration of

367

DOM was above 10 mgL-1. This indicates that reactive surface hydroxyl sites in

368

ferrihydrite and goethite displayed different affinity to DOM. Therefore, we compared

369

the types of the surface hydroxyl groups in ferrihydrite and goethite from structural

370

chemistry as reported by previous studies51-53. Goethite is needle-shaped and consists

371

of about 80–95% 110/100 like faces, whereas spherical ferrihydrite was dominated by

372

001/021-like faces. On the 110/100 like faces, triply-coordinated (≡Fe3OH) surface

373

hydroxyl makes the main contribution to reactive sites, while the reactive sites of

374

001/021-like faces are dominated by singly coordinated ( ≡ FeOH) surface

375

hydroxyl51-53. It is reasonable to expect that singly-coordinated (≡FeOH) groups

376

exhibit a higher affinity to organic substances with more selectivity than

377

triply-coordinated ( ≡ Fe3OH) surface hydroxyls, leading to different molecular

378

fractionation of DOM on ferrihydrite and goethite.

379

In addition, it appears that oxygen atom content and unsaturation level of

380

compounds are important parameters governing their adsorption behavior. Graphics of

ACS Paragon Plus Environment

Page 16 of 35

Page 17 of 35

Environmental Science & Technology

381

the number of O atoms versus DBE values in the molecular formulas of PL before and

382

after adsorption on ferrihydrite and goethite were therefore investigated in detail. As

383

shown in Figure 5, most of the oxygenated compounds in the original PL contained 3

384

to 19 oxygen atoms with DBE values in the range 4 to 23. Almost all the compounds

385

containing over 14 O atoms or with DBE values above 14 were sequestered by

386

ferrihydrite (Figures 5A and B). In the case of goethite, only compounds with O atoms

387

over 19 or DBE value above 21 disappeared in G1 (Figure 5C). In general, one epoxy,

388

acyl, carbanyl or carboxyl can induce a unit increase in DBE value per compound

389

while ether and alcohol make no contribution to the DBE values. The overall DBE

390

distribution can be described in terms of the magnitude-weighted average DBE

391

(DBEw)54, calculated as

392

DBEw =(∑ I × (DBE  ⁄∑ I

393

where Ii and (DBE)i are the relative abundance and DBE value of peak i, respectively.

394

The relative abundance is calculated as the abundance of the individual peak divided

395

by the maximum abundance in a given spectrum.

(1)

396

DBEw is presented in Figure 5E as a function of the number of oxygen atoms

397

obtained from molecular formulas of PL before and after adsorption on ferrihydrite

398

and goethite. Slopes, y-intercepts, and R2 values were obtained by linear regression

399

analysis. Analysis indicates a strongly significant correlation between the number of

400

oxygen atoms and the DBE values of DOM molecules (R2>0.97). Although the

401

calculated equation of DBE (c-h/2+1 for CcHhOo) does not contain the number of

402

oxygen atoms in a molecule, two oxygen atoms in the carboxyl group can contribute

403

to one DBE value. The contributions of epoxy and acyl groups were neglected here

404

because of their rare low contents even if present in natural organic substances55, 56.

405

The slope of the line between DBE and number of O atoms for original PL was

ACS Paragon Plus Environment

Environmental Science & Technology

406

0.72, and slight decreases were found in the slopes for G1 (0.68) and G2 (0.70). A

407

slope approaching 0.7 indicates that the addition of two oxygen atoms resulted in an

408

increase of ~1.4 DBEw. As explained above, two oxygen atoms in a carboxyl group

409

contributed to only one DBE value. The extra 0.4 DBE can be explained by the

410

presence of cyclic structures or unsaturated chemical bonds if all oxygen atoms were

411

included in carboxyl groups. Considering the presence of ether and alcohol groups in

412

PL, the contributions of cyclic structures or unsaturated chemical bonds to the DBE

413

values should be greater than estimated. After adsorption on ferrihydrite, the slopes

414

for F1 and F2 decreased to 0.39 and 0.40, respectively, indicating that the addition of

415

two oxygen atoms resulted in an increase of only ~0.8 DBEw (below one). This

416

suggests that the contribution of oxygen atoms to DBE decreased, and their actual

417

contribution should be smaller because of the presence of cyclic structures and

418

unsaturated chemical bonds in compounds. This indicates that the content of carboxyl

419

groups decreased while those of ethers and alcohols in oxygenated molecules

420

increased in the PL supernatant after adsorption on ferrihydrite. Taking account of the

421

low SUVA254 of the PL supernatant after adsorption by ferrihydrite and the MW and

422

group fractionation stated above, the PL supernatants contained mainly low MW

423

compounds rich in aliphatic and alkyl groups substituted with ethers and alcohols. In

424

other words, compounds sequestered by ferrihydrite were mainly high MW

425

compounds rich in carboxyl groups and aromatic rings as demonstrated in previous

426

studies by spectroscopic techniques such as FT-IR and nuclear magnetic resonance

427

(NMR)15, 16

428

It is noteworthy that the slopes of the original PL and PL supernatant after

429

adsorption on goethite were close to the inland riverine DOM (mainly as terrigenous

430

DOM) and they were close to the offshore coastal DOM after adsorptive fractionation

ACS Paragon Plus Environment

Page 18 of 35

Page 19 of 35

Environmental Science & Technology

431

by ferrihydrite54. This may explain the field observation by Riedel et al.57 that redox

432

interfaces containing iron in sediment acted as selective intermediate barriers and

433

limited the flux of land-derived DOM to oceanic waters. The present study highlights

434

the role of poorly crystalline iron oxyhydroxides in the selective fractionation of

435

DOM at soil/sediment and water interfaces.

436

Geochemical and Environmental Implications. DOM is a labile organic carbon

437

pool. Previous studies have demonstrated the importance of minerals, especially iron

438

oxyhydroxides, for stabilization and preservation of labile organic matter in soils and

439

sediments5,6. However, very few studies have focused on the changes in molecular

440

composition of DOM at the mineral-water interface. Using ESI FT-ICR MS, this

441

study addresses molecular scale fractionation of DOM induced by adsorption on iron

442

oxyhydroxides. The results indicate that poorly crystalline iron oxyhydroxides with

443

surface dominated by singly-coordinated Fe-OH sites play a more important role in

444

molecular fractionation of DOM at mineral and water interfaces. DOM compounds

445

with low molecular weight and low unsaturation or compounds containing few

446

oxygenated groups (mainly alcohols and ethers) are preferentially maintained in

447

solution, while adverse compounds are preferentially sequestered by iron

448

oxyhydroxides. Similar molecular fractionation of FA and HA has also been observed

449

at the surface of aluminum oxides17, 24. It is interesting to note that NOM in natural

450

waters and DOM in soil solutions largely contains hydrophilic fractions including

451

short chain aliphatic carboxylic acids, hydrocarbons and amino acids but few

452

high-molecular-weight, aromatic compounds and phenolic components58. This is

453

highly consistent with the DOM fractions in the supernatant after molecular

454

fractionation by iron oxyhydroxides that we observed in the present study and in

455

aluminum oxides reported by Galindo et al.17, 24. Therefore, it is to be expected that

ACS Paragon Plus Environment

Environmental Science & Technology

456

the adsorptive fractionation of DOM at hydroxylated mineral surfaces to a great

457

extent determines the chemical components of DOM in labile phases such as surface

458

water and pore water, and a large scale field investigation is needed using ultrahigh

459

resolution mass spectrometry to reveal the molecular composition of labile organic

460

matter in the aquatic and terrestrial environments. In addition, from an environmental

461

perspective, such fractionation led to changes in chemical components of DOM as

462

well as the surface characteristics of the minerals, which will further impact the

463

behaviors of contaminants in the environment such as sorption, transportation and

464

transformation processes. Future studies are also necessary to investigate the detailed

465

mechanisms by which molecular fractionation of DOM affects the behaviors of

466

contaminants in the environment.

467 468

ASSOCIATED CONTENT

469

Supporting Information

470

This information is available free of charge via the Internet at http://pubs.acs.org/.

471

AUTHOR INFORMATION

472

Corresponding Author

473

*E-mail: [email protected]

474

Notes

475

The authors declare no competing financial interest.

476

ACKNOWLEDGMENTS

477

This work was funded by the 973 Program of China (Grant 2014CB441102), the

478

Strategic Priority Research Program of the Chinese Academy of Sciences (Grant

479

XDB14020202), and the National Natural Science Foundation of China (Projects

480

21407161 and 21537005). We thank Dr. Yunzhu Zhao at Michigan Technological

ACS Paragon Plus Environment

Page 20 of 35

Page 21 of 35

Environmental Science & Technology

481

University for sharing invaluable experience in FT-MS data analysis. We also thank

482

for three anonymous reviewers for their constructive comments which have greatly

483

improved the presentation of this work.

484 485

Literature Cited

486

(1) Maurer, F.; Christl, I.; Kretzschmar, R., Reduction and reoxidation of humic acid:

487

Influence on spectroscopic properties and proton binding. Environ Sci Technol

488

2010, 44 (15), 5787-5792.

489

(2) Gaberell, M.; Chin, Y. P.; Hug, S. J.; Sulzberger, B., Role of dissolved organic

490

matter composition on the photoreduction of Cr(VI) to Cr(III) in the presence of

491

iron. Environ Sci Technol 2003, 37 (19), 4403-4409.

492

(3) Huang, W. L.; Ping, P. A.; Yu, Z. Q.; Fu, H. M., Effects of organic matter

493

heterogeneity on sorption and desorption of organic contaminants by soils and

494

sediments. Appl Geochem 2003, 18 (7), 955-972.

495 496

(4) Hassett, J. P., Chemistry-Dissolved natural organic matter as a microreactor. Science 2006, 311 (5768), 1723-1724.

497

(5) Keil, R. G.; Montlucon, D. B.; Prahl, F. G.; Hedges, J. I., Sorptive preservation

498

of labile organic-matter in marine-sediments. Nature 1994, 370 (6490), 549-552.

499

(6) Torn, M. S.; Trumbore, S. E.; Chadwick, O. A.; Vitousek, P. M.; Hendricks, D.

500

M., Mineral control of soil organic carbon storage and turnover. Nature 1997,

501

389 (6647), 170-173.

502

(7) Murphy, E. M.; Zachara, J. M.; Smith, S. C.; Phillips, J. L., The sorption of

503

humic acids to mineral surfaces and their role in contaminant binding. Sci Total

504

Environ 1992, 118, 413-423.

505

(8) Cygan, R. T.; Tazaki, K., Interactions of kaolin minerals in the environment.

ACS Paragon Plus Environment

Environmental Science & Technology

506 507 508 509 510 511 512

Elements 2014, 10 (3), 195-200. (9) Polubesova, T.; Chefetz, B., DOM-affected transformation of contaminants on mineral surfaces: A review. Crit Rev Env Sci Tec 2014, 44 (3), 223-254. (10) Piccolo, A., The supramolecular structure of humic substances. Soil Sci 2001, 166 (11), 810-832. (11) Sutton, R.; Sposito, G., Molecular structure in soil humic substances: The new view. Environ Sci Technol 2005, 39 (23), 9009-9015.

513

(12) Zhou, Q. H.; Maurice, P. A.; Cabaniss, S. E., Size fractionation upon adsorption

514

of fulvic acid on goethite: Equilibrium and kinetic studies. Geochim Cosmochim

515

Ac 2001, 65 (5), 803-812.

516

(13) Fu, H. B.; Quan, X., Complexes of fulvic acid on the surface of hematite,

517

goethite, and akaganeite: FTIR observation. Chemosphere 2006, 63 (3), 403-410.

518

(14) Claret, F.; Schafer, T.; Brevet, J.; Reiller, P. E., Fractionation of Suwannee River

519

fulvic acid and Aldrich humic acid on alpha-Al2O3: Spectroscopic evidence.

520

Environ Sci Technol 2008, 42 (23), 8809-8815.

521

(15) Eusterhues, K.; Rennert, T.; Knicker, H.; Kogel-Knabner, I.; Totsche, K. U.;

522

Schwertmann, U., Fractionation of organic matter due to reaction with

523

ferrihydrite: Coprecipitation versus adsorption. Environ Sci Technol 2011, 45 (2),

524

527-533.

525

(16) Chen, C.; Dynes, J. J.; Wang, J.; Sparks, D. L., Properties of Fe-organic matter

526

associations via coprecipitation versus adsorption. Environ Sci Technol 2014, 48

527

(23), 13751-13759.

528

(17) Galindo, C.; Del Nero, M., Molecular level description of the sorptive

529

fractionation of a fulvic acid on aluminum oxide using electrospray ionization

530

Fourier transform mass spectrometry. Environ Sci Technol 2014, 48 (13),

ACS Paragon Plus Environment

Page 22 of 35

Page 23 of 35

Environmental Science & Technology

531 532 533

7401-7408. (18) Hur, J.; Schlautman, M. A., Molecular weight fractionation of humic substances by adsorption onto minerals. J Colloid Interf Sci 2003, 264 (2), 313-321.

534

(19) Reiller, P.; Amekraz, B.; Moulin, C., Sorption of Aldrich humic acid onto

535

hematite: Insights into fractionation phenomena by electrospray ionization with

536

quadrupole time-of-flight mass spectrometry. Environ Sci Technol 2006, 40 (7),

537

2235-2241.

538

(20) Hertkorn, N.; Frommberger, M.; Witt, M.; Koch, B. P.; Schmitt-Kopplin, P.;

539

Perdue, E. M., Natural organic matter and the event horizon of mass

540

spectrometry. Anal Chem 2008, 80 (23), 8908-8919.

541

(21) Minor, E. C.; Swenson, M. M.; Mattson, B. M.; Oyler, A. R., Structural

542

characterization of dissolved organic matter: A review of current techniques for

543

isolation and analysis. Environ Sci-Proc Imp 2014, 16 (9), 2064-2079.

544

(22) Reemtsma, T., Determination of molecular formulas of natural organic matter

545

molecules by (ultra-) high-resolution mass spectrometry status and needs. J

546

Chromatogr A 2009, 1216 (18), 3687-3701.

547 548

(23) Riedel, T.; Biester, H.; Dittmar, T., Molecular fractionation of dissolved organic matter with metal salts. Environ Sci Technol 2012, 46 (8), 4419-4426.

549

(24) Galindo, C.; Del Nero, M., Chemical fractionation of a terrestrial humic acid

550

upon sorption on alumina by high resolution mass spectrometry. Rsc Adv 2015, 5

551

(89), 73058-73067.

552 553

(25) Lalonde, K.; Mucci, A.; Ouellet, A.; Gelinas, Y., Preservation of organic matter in sediments promoted by iron. Nature 2012, 483 (7388), 198-200.

554

(26) Melton, E. D.; Swanner, E. D.; Behrens, S.; Schmidt, C.; Kappler, A., The

555

interplay of microbially mediated and abiotic reactions in the biogeochemical Fe

ACS Paragon Plus Environment

Environmental Science & Technology

556

cycle. Nat Rev Microbiol 2014, 12 (12), 797-808.

557

(27) Filius, J. D.; Lumsdon, D. G.; Meeussen, J. C. L.; Hiemstra, T.; Van Riemsdijk,

558

W. H., Adsorption of fulvic acid on goethite. Geochim Cosmochim Ac 2000, 64

559

(1), 51-60.

560

(28) van Schaik, J. W. J.; Persson, I.; Kleja, D. B.; Gustafsson, J. P., EXAFS study on

561

the reactions between iron and fulvic acid in acid aqueous solutions. Environ Sci

562

Technol 2008, 42 (7), 2367-2373.

563 564

(29) Adhikari, D.; Yang, Y., Selective stabilization of aliphatic organic carbon by iron oxide. Sci Rep 2015, 5, 11214.

565

(30) O'Loughlin, E. J.; Gorski, C. A.; Scherer, M. M.; Boyanov, M. I.; Kemner, K. M.,

566

Effects of oxyanions, natural organic matter, and bacterial cell numbers on the

567

bioreduction of lepidocrocite (gamma-FeOOH) and the formation of secondary

568

mineralization products. Environ Sci Technol 2010, 44 (12), 4570-4576.

569 570 571 572

(31) Vodyanitskii, Y. N., Iron hydroxides in soils: A review of publications. Eurasian Soil Sci 2010, 43 (11), 1244-1254. (32) Matta, R.; Hanna, K.; Chiron, S., Fenton-like oxidation of 2,4,6-trinitrotoluene using different iron minerals. Sci Total Environ 2007, 385 (1-3), 242-251.

573

(33) Ona-Nguema, G.; Morin, G.; Juillot, F.; Calas, G.; Brown, G. E., EXAFS

574

analysis of arsenite adsorption onto two-line ferrihydrite, hematite, goethite, and

575

lepidocrocite. Environ Sci Technol 2005, 39 (23), 9147-9155.

576

(34) Zhang, T.; Li, C. J.; Ma, J.; Tian, H.; Qiang, Z. M., Surface hydroxyl groups of

577

synthetic α-FeOOH in promoting OH generation from aqueous ozone: Property

578

and activity relationship. Applied Catalysis B: Environmental 2008, 82 (1-2),

579

131-137.

580

(35) Qin, F.; Wen, B.; Shan, X. Q.; Xie, Y. N.; Liu, T.; Zhang, S. Z.; Khan, S. U.,

ACS Paragon Plus Environment

Page 24 of 35

Page 25 of 35

Environmental Science & Technology

581

Mechanisms of competitive adsorption of Pb, Cu, and Cd on peat. Environ Pollut

582

2006, 144 (2), 669-80.

583

(36) Dittmar, T.; Koch, B.; Hertkorn, N.; Kattner, G., A simple and efficient method

584

for the solid-phase extraction of dissolved organic matter (SPE-DOM) from

585

seawater. Limnol Oceanogr-Meth 2008, 6, 230-235.

586

(37) Cao, D.; Huang, H.; Hu, M.; Cui, L.; Geng, F.; Rao, Z.; Niu, H.; Cai, Y.; Kang,

587

Y., Comprehensive characterization of natural organic matter by MALDI- and

588

ESI-Fourier transform ion cyclotron resonance mass spectrometry. Anal Chim

589

Acta 2015, 866, 48-58.

590

(38) Hughey, C. A.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G.; Qian, K. N.,

591

Kendrick

mass

defect

spectrum:

A

compact

visual

analysis

for

592

ultrahigh-resolution broadband mass spectra. Anal Chem 2001, 73 (19),

593

4676-4681.

594

(39) Kim, S.; Kramer, R. W.; Hatcher, P. G., Graphical method for analysis of

595

ultrahigh-resolution broadband mass spectra of natural organic matter, the van

596

Krevelen diagram. Anal Chem 2003, 75 (20), 5336-5344.

597

(40) Stenson, A. C.; Marshall, A. G.; Cooper, W. T., Exact masses and chemical

598

formulas of individual Suwannee River fulvic acids from ultrahigh resolution

599

electrospray ionization Fourier transform ion cyclotron resonance mass spectra.

600

Anal Chem 2003, 75 (6), 1275-1284.

601

(41) Koch, B. P.; Dittmar, T., From mass to structure: An aromaticity index for

602

high-resolution mass data of natural organic matter. Rapid Commun Mass Sp

603

2006, 20 (5), 926-932.

604

(42) Kellerman, A. M.; Dittmar, T.; Kothawala, D. N.; Tranvik, L. J., Chemodiversity

605

of dissolved organic matter in lakes driven by climate and hydrology. Nat

ACS Paragon Plus Environment

Environmental Science & Technology

606

Commun 2014, 5, 3804-3911.

607

(43) Pronk, G. J.; Heister, K.; Kogel-Knabner, I., Iron oxides as major available

608

interface component in loamy arable topsoils. Soil Sci Soc Am J 2011, 75 (6),

609

2158-2168.

610

(44) Kleber, M.; Mikutta, R.; Torn, M. S.; Jahn, R., Poorly crystalline mineral phases

611

protect organic matter in acid subsoil horizons. Eur J Soil Sci 2005, 56 (6),

612

717-725.

613

(45) Gu, B. H.; Schmitt, J.; Chen, Z.; Liang, L. Y.; Mccarthy, J. F., Adsorption and

614

desorption of different organic-matter fractions on iron-oxide. Geochim

615

Cosmochim Ac 1995, 59 (2), 219-229.

616 617

(46) Jolivet, J. P.; Chaneac, C.; Tronc, E., Iron oxide chemistry. From molecular clusters to extended solid networks. Chem Commun 2004, (5), 481-487.

618

(47) Sleighter, R. L.; Hatcher, P. G., Molecular characterization of dissolved organic

619

matter (DOM) along a river to ocean transect of the lower Chesapeake Bay by

620

ultrahigh resolution electrospray ionization Fourier transform ion cyclotron

621

resonance mass spectrometry. Mar Chem 2008, 110 (3-4), 140-152.

622

(48) Davis, J. A.; Gloor, R., Adsorption of dissolved organics in lake water by

623

aluminum oxide. Effect of molecular weight. Environ Sci Technol 1981, 15 (10),

624

1223-9.

625

(49) Santl-Temkiv, T.; Finster, K.; Dittmar, T.; Hansen, B. M.; Thyrhaug, R.; Nielsen,

626

N. W.; Karlson, U. G., Hailstones: A window into the microbial and chemical

627

inventory of a storm cloud. Plos One 2013, 8 (1), e53550.

628 629 630

(50) Evanko, C. R.; Dzombak, D. A., Surface complexation modeling of organic acid sorption to goethite. J Colloid Interface Sci 1999, 214 (2), 189-206. (51) Boily, J. F.; Lutzenkirchen, J.; Balmes, O.; Beattie, J.; Sjoberg, S., Modeling

ACS Paragon Plus Environment

Page 26 of 35

Page 27 of 35

Environmental Science & Technology

631

proton binding at the goethite (alpha-FeOOH)-water interface. Colloid Surface A

632

2001, 179 (1), 11-27.

633

(52) Venema, P.; Hiemstra, T.; Weidler, P. G.; van Riemsdijk, W. H., Intrinsic proton

634

affinity of reactive surface groups of metal (hydr)oxides: Application to iron

635

(hydr)oxides. J Colloid Interf Sci 1998, 198 (2), 282-295.

636

(53) Hiemstra, T.; Van Riemsdijk, W. H., A surface structural model for ferrihydrite I:

637

Sites related to primary charge, molar mass, and mass density. Geochim

638

Cosmochim Ac 2009, 73 (15), 4423-4436.

639

(54) Bae, E.; Yeo, I. J.; Jeong, B.; Shin, Y.; Shin, K. H.; Kim, S., Study of double

640

bond equivalents and the numbers of carbon and oxygen atom distribution of

641

dissolved organic matter with negative-mode FT-ICR MS. Anal Chem 2011, 83

642

(11), 4193-4199.

643

(55) Solomon, D.; Lehmann, J.; Kinyangi, J.; Liang, B. Q.; Schafer, T., Carbon

644

K-edge NEXAFS and FTIR-ATR spectroscopic investigation of organic carbon

645

speciation in soils. Soil Sci Soc Am J 2005, 69 (1), 107-119.

646 647

(56) Schaumann, G. E.; Thiele-Bruhn, S., Molecular modeling of soil organic matter: Squaring the circle? Geoderma 2011, 166 (1), 1-14.

648

(57) Riedel, T.; Zak, D.; Biester, H.; Dittmar, T., Iron traps terrestrially derived

649

dissolved organic matter at redox interfaces. P Natl Acad Sci USA 2013, 110 (25),

650

10101-10105.

651

(58) Karavanova, E. I., Dissolved organic matter: Fractional composition and

652

sorbability by the soil solid phase (Review of literature). Eurasian Soil Sci 2013,

653

46 (8), 833-844.

654 655

ACS Paragon Plus Environment

Environmental Science & Technology

656

FIGURE LEGENDS

657

Figure 1 Adsorption of PL on ferrihydrite (Fh.), goethite (Gth.), lepidocrocite (Lep.)

658

(A); SUVA254 values (B) of residual PL suspension after adsorption on Fh., Gth., and

659

Lep., dash circles F1, F2, G1, G2 and L1, L2 in (A) are samples selected to analysis

660

by ESI-FTICR-MS.

661

Figure 2 Broadband ESI-FTICR-MS spectrum (A) and molecular weight distribution

662

(B) of original PL and residual PL suspension after adsorption on ferrihydrite,

663

goethite, and lepidocrocite, F1, F2, G1, G2 and L1, L2 are samples labeled in Figure

664

1.

665

Figure 3 The molecular composition of PL before and after adsorption on iron

666

oxyhydroxides. A, B, C, D and E are van Krevelen diagrams of PL, F1, F2, G1 and

667

G2, respectively; color bar represents the relative intensity of peak with % unit. Four

668

groups in van Krevelen diagram are delineated by the aromaticity index (AI) and H/C

669

cutoffs: group 1 represents combustion-derived condensed polycyclic aromatics (AI>

670

0.66), group 2 represents vascular plant-derived polyphenols (0.66≥AI>0.5), group 3

671

represents highly unsaturated and phenolic compounds (AI≤0.50 and H/C<1.5), and

672

group 4 represents aliphatic compounds (2.0≥H/C≥1.5); 1,2,3,4 in A represent groups

673

1, 2, 3, and 4, respectively. F is the relative percentage of total compounds and

674

compounds in each group in F1, F2, G1, G2 compared to the original PL.

675

Figure 4 Quantitative distributions of O/C (A, D), H/C (B, E), and DBE (C, F) of

676

molecules in PL before and after adsorption on ferrihydrite and goethite; subscript

677

suspension

denotes molecules of PL remaining in solution after adsorption; subscript

678

sequestered

denotes molecules of PL sequestered by iron oxyhydroxides.

679

Figure 5 Plot of DBE versus the number of O atoms (A, B, C, D) and the DBEw as a

680

function of the number of O atoms (E) in molecular formulas of PL before and after

ACS Paragon Plus Environment

Page 28 of 35

Page 29 of 35

Environmental Science & Technology

681

adsorption on ferrihydrite and goethite.

682

ACS Paragon Plus Environment

Environmental Science & Technology

500

A

F2

-2

PL adsorbed (µg C m )

400 300 G2, L2

F1

200 G1, L1

100

Fh. Gth. Lep.

0 0

20 40 60 -1 Solution equilibrium concentration (mg C L )

0.4

B

-1

-1

SUVA254 (L mgC m )

0.3

0.2 Fh. Gth. Lep.

0.1

0.0 0

683 684

20 40 60 -1 Solution equilibrium concentration (mg C L )

Figure 1

ACS Paragon Plus Environment

Page 30 of 35

Page 31 of 35

Environmental Science & Technology

90

A

60 30

L1

Relative-abundance

0 90 60 30

G1

0 90 60 30

F1

0 90 60 30

PL

0 200

300

400

500

600

700

m/z

25

PL G1 F2

B

Percentage( %)

20

F1 L1 G2 L2

15 10 5 0 200 -250

685 686

250 -300

300 350 -350 -400

400 450 500 550 -450 -500 -550 -600

600 -650

650 -700

m/z distribution

Figure 2

687

ACS Paragon Plus Environment

Environmental Science & Technology

100

1.2

3 50

1.0 0.8

H/C

H/C

1.4

2

0.4

0.6 O/C

0.8

1.0

50

100

D

1.6

1.4

1.4

1.2

50

1.0 0.8

0.4

0.6 O/C

0.8

0.0 0.2

0.4

0.6 O/C

0.8

1.0

0.4

0.0 0.2

0.4

0.6 O/C

1.0

100

E

0.8

1.0

F

total group1 group2 group3 group4

0.8

1.2

50

1.0

0.6 0.4 0.2

0.6

0.4

50

1.0

1.0

0.8

0.6

1.2

0.6

0.0

1.8

1.6

100

C

0.8

0.2

H/C

H/C

1.4

1.2

0.4

0.0

1.8

689

1.4

0.6

1

1.8 1.6

0.8

0.2

688

100

B

1.6

1.0

0.6 0.4

1.8

H/C

A

4

1.6

Relative frequency

1.8

Page 32 of 35

0.4

0.0 0.2

0.4

0.6 O/C

0.8

0.0

1.0

Figure 3

690 691

ACS Paragon Plus Environment

F1

F2

G1

G2

Page 33 of 35

Environmental Science & Technology

400

F2sequestered

200

100

26 -2 8

22 -2 4 24 -2 6

16 -1 8 18 -2 0 20 -2 2

12 -1 4 14 -1 6

PL G1suspension

F

200

100

G1sequestered

Number of molecules

G2sequestered

0

G2suspension

200

G2sequestered

100

H/C

Figure 4

694

ACS Paragon Plus Environment

DBE

26 -2 8

24 -2 6

22 -2 4

18 -2 0 20 -2 2

16 -1 8

12 -1 4 14 -1 6

10 -1 2

68

46

810

02

24

>1 .9

1. 71.9

-1 .7 1.5

1.3 -1 .5

1.1 -1 .3

0.9 -1 .1

-0 .7

0.7 -0 .9

0 1

0.9 -1

0.8 -0 .9

0.6 -0 .7

0.7 -0 .8

0.4 -0 .5

0. 50.6

0.2 -0 .3

0

G2suspension

0.5

100

300

-0 .5

200

300

G1sequestered

0.3

G2sequestered

0.3 -0 .4

02

PL G1suspension

E Number of molecules

G2suspension

1 .9

DBE

G1sequestered

300

0.1 -0 .2

1. 71.9

1.3 -1 .5

1.5 -1 .7

0.9 -1 .1

1.1 -1 .3

0.5 -0 .7

0.7 -0 .9

1

0.9

0.8 -0 .9

0.6 -0 .7

0.7 -0 .8

0.4 -0 .5

0. 50.6

0.2 -0 .3

0.3 -0 .4