Elucidating Adsorptive Fractions of Natural Organic Matter on Carbon

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Elucidating Adsorptive Fractions of Natural Organic Matter on Carbon Nanotubes Mohamed Ateia, Onur G. Apul, Yuta Shimizu, Astri Muflihah, Chihiro Yoshimura, and Tanju Karanfil Environ. Sci. Technol., Just Accepted Manuscript • Publication Date (Web): 24 May 2017 Downloaded from http://pubs.acs.org on May 25, 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.

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 33

Environmental Science & Technology

1

2

Elucidating Adsorptive Fractions of Natural Organic Matter on

3

Carbon Nanotubes

4 5

Mohamed Ateia*,1, Onur G. Apul2,3, Yuta Shimizu1, Astri Muflihah1, Chihiro Yoshimura1, Tanju

6

Karanfil4

7 8

1

9

M1-4 Ookayama, Tokyo 152-8552, Japan

Department of Civil and Environmental Engineering, Tokyo Institute of Technology, 2-12-1-

10

2

11

Lowell, MA 01854, USA

12

3

13

AZ 85259, USA

14

4

15

SC 29625, USA

Department of Civil and Environmental Engineering, University of Massachusetts Lowell,

School of Sustainable Engineering and the Built Environment, Arizona State University, Tempe,

Department of Environmental Engineering and Earth Sciences, Clemson University, Anderson,

16 17

* Corresponding author: [email protected]; Phone: 0081-80-3430-9753

18

Submitted to: Environmental science & technology

19

Submission Date: March 10th, 2017

20

Revised and resubmitted: May 22nd, 2017 1 ACS Paragon Plus Environment


21

Abstract

22

Natural organic matter (NOM) is a heterogeneous mixture of organic compounds that is

23

omnipresent in natural waters. To date, our understanding of NOM components’ adsorption by

24

carbon nanotubes (CNTs) is limited due to limited number of comprehensive studies in the

25

literature examining the adsorption of NOM by CNTs. In this study, eleven standard NOMs from

26

various sources were characterized, and their adsorption on four different CNTs were examined

27

side-by-side using total organic carbon, fluorescence, UV-visible spectroscopy, and high

28

performance size exclusion chromatography (HPSEC) analysis. Adsorption was influenced by

29

chemical properties of NOM, including aromaticity, degree of oxidation and carboxylic acidities.

30

Fluorescence excitation-emission matrix analysis showed preferential adsorption of decomposed

31

and terrestrial-derived over freshly produced and microbial-derived NOM. HPSEC analysis

32

revealed preferential adsorption of fractions in the molecular weight (MW) range of 0.5−2 kDa of

33

humic acids, while in MW range of 1−3 kDa for all fulvic acids and reverse osmosis isolates.

34

However, the smallest characterized fraction (MW < 0.4 kDa) in all samples did not adsorb on

35

CNTs.

36 37 38 39 40

Keywords: natural organic matter, carbon nanotubes, adsorption, fluorescence index, molecular

41

weight

2 ACS Paragon Plus Environment

Page 2 of 33

Page 3 of 33

42


1. Introduction

43

Natural organic matter (NOM) is ubiquitous in fresh waters and found in concentrations that range

44

from 1−2 mg-C/L up to 40 mg-C/L depending on the source and climate.1, 2 The presence of NOM ,

45

which is a mixture of complex polyelectrolytes, during water treatment process possess a broad

46

range of problems, such as increase in chemical demands, formation of disinfection byproducts,

47

taste and odor problems, and fouling of activated carbons and membranes.3, 4 Therefore, several

48

processes have been employed to remove NOM from water, including coagulation, membrane

49

filtration, ion exchange, advanced oxidation processes and adsorption.5

50

Carbon nanotubes (CNTs), with their high surface areas, hydrophobicity, porosity, and rapid

51

sorption kinetics, have been explored as one of the next-generation adsorbents. 6, 7 Over the past

52

decade, extensive studies have investigated applications of CNTs to adsorb several organic and

53

inorganic water contaminants.8, 9 Furthermore, the interactions between CNTs and NOM have

54

been examined in some studies.1, 10-17 Adsorption mechanisms of NOM components by CNTs

55

include; hydrophobic interactions, π-π interactions, hydrogen bonding and electrostatic

56

interactions.10, 18 These interactions are influenced by: 1) characteristics of CNTs (e.g., surface

57

area, pore volume, and surface functionalities),10,

58

composition (e.g., molecular weight and size),13, 19 and 3) solution chemistry (e.g., pH, water

59

temperature, and ionic strength).1, 10

60

It was also reported that conformation and composition of NOM remaining in water will change

61

as a result of adsorption process due to NOM fractionation.15,

62

comprehensive studies in the literature examining the adsorption of NOM by CNTs. Today, limited

63

observations in previous studies are due to a number of factors: 1) different NOM used in these

14, 19

2) NOM molecular structure and


20

However, there is no


64

studies (i.e., different chemical and physical characteristics), 2) limitations in the number of NOM

65

tested under different experimental conditions, and 3) limited and different techniques employed

66

in NOM characterization (e.g., high performance size exclusion chromatography analysis

67

(HPSEC)10, 12 vs. E2/E3 (UV absorbance at 250 nm/365 nm) and E4/E6 (UV absorbance at 465

68

nm/665 nm)).11, 15

69

In this study, our objective was to systematically investigate the adsorption behavior of eleven

70

NOM from various sources performing isotherm and kinetic experiments and using a suite of

71

NOM characterization techniques (total organic carbon, fluorescence, UV-vis spectroscopy and

72

HPSEC). The results will give a more robust understanding for adsorption behavior of NOMs on

73

CNTs in natural waters or the potential applications for water treatment.

74

2. Materials and Methods

75

Materials

76

Four standard humic acids (Elliott Soil IV [ESHA], Pahokee Peat [PPHA], Leonardite [LHA], and

77

Suwannee River II [SRHA]), four standard fulvic acids (Suwannee River II [SRFA1], Suwannee

78

River I [SRFA2], Pahokee Peat II [PPFA], and Nordic Lake [NLFA]), and three RO isolates

79

(Nordic Reservoir [NNOM], Upper Mississippi River [MNOM], and Suwannee River [SNOM])

80

were purchased from the International Humic Substances Society (IHSS). The chemical properties

81

of the NOM, including elemental composition/ratio, carbon species, the concentration of acidic

82

functional groups, fluorescence indexes, and average molecular weights are listed in Table 1 and

83

Table S1.

84

Three different multi-walled carbon nanotubes (MWCNT) and one single-walled carbon

85

nanotubes (SWCNT) were purchased from Wako pure chemicals, Japan and Chengdu Alpha Nano 4 ACS Paragon Plus Environment

Page 4 of 33

Page 5 of 33


86

Technology Co., Chinese Academy of Science, China. Characteristics of all carbon nanotubes used

87

in this study are listed in Table 2. The Brunauer–Emmett–Teller (BET) surface areas, pore

88

volumes and pore size distributions were measured from nitrogen physisorption data at 77 °K

89

obtained with ASAP 2020 analyzer (Micromeritics Instrument Corp. U.S.). The oxygen content of

90

adsorbents was analyzed using a Flash Elemental Analyzer 1112 series (Thermo Electron

91

Corporation).

92

Preparation of NOM solutions

93

NOM samples (25−50 mg) were dissolved in a 0.01 M NaOH solution to prepare a stock solution

94

of 10 g/L. The pH of each NOM solution was adjusted to 7.0 ± 0.2 with 0.01 M HCl or 0.01 M

95

NaOH. The volume of acid or alkali solution was recorded and was taken into account during

96

calculation of the final NOM concentration. All stock solution were stored in dark and refrigerated

97

(4 °C), when not in use.

98

Adsorption isotherms

99

Batch reactors were used to perform adsorption experiments. Each NOM stock solution was

100

diluted with 1 mM phosphate buffer solution then adjusted to pH of 7.0 ± 0.2. Based on preliminary

101

kinetic experiments, the equilibrium time was set at 24 h (Section S1 in SI). Adsorption isotherm

102

experiments were carried out under different conditions: 1) constant CNT dose and variable NOM

103

concentrations (2.5−60 mg/L), and 2) variable CNT dose (125−500 mg-CNT/L) and constant

104

NOM concentration. At the end of the equilibrium time, samples were collected, filtered (pre-

105

washed 0.45 µm PES filter, Membrane Solutions, Japan), and kept refrigerated until further

106

measurements. The adsorbed NOM amount (measured as dissolved organic carbon (DOC)) were

107

determined by subtracting the concentration of dissolved NOM remaining in the solution from the



108

initial concentration. Each isotherm point was performed in triplicates and variations were reported

109

as standard error. Blank experiments (with no adsorbents) were conducted in parallel, and the

110

change in NOM concentrations in blanks were found negligible.

111

Analysis

112

DOC concentrations were quantified using a total organic carbon (TOC) analyzer (Shimadzu V-

113

series, TOC-CHP, Japan). The UV absorbance at 254 nm was obtained with an UV−visible

114

spectrophotometer (Shimadzu UV-1800, Japan). The specific UV absorption (SUVA254) values

115

were calculated by dividing the UV absorbance at 254 nm (m-1) to the DOC concentration (mg/L).

116

Average molecular weight of all samples were determined using a high performance size exclusion

117

chromatography (HPSEC) unit (SHIMADZU Prominence, Japan), SEC column (Shodex column

118

[SB-803 HQ]) and mobile phase (75% phosphate buffer + 25% acetonitrile (Wako Pure Chemicals

119

ltd, Japan)) adjusted at pH 7. Detection of NOM was achieved with a UV detector set at 254 nm.

120

Polystyrene sulfonate (PSS) standards (1.0, 4.6, 18, and 67 kDa) were purchased from Polyscience

121

Inc. and were used as standard molecular weights. Molecular weights and polydispersity were

122

determined according to Chin, et al. 21 It should be noted that calculated average MW values by

123

HPSEC with UV-vis detectors are skewed by SUVA of the DOC. Variations in MW values

124

obtained from HPSEC in different studies are due to changes in mobile phases, electrostatic

125

repulsion, and column type; nevertheless, SEC is useful technique to compare changes in

126

molecular weights among different samples.22

127

EEM fluorescence was measured as described by Murphy, et al. 23 using a RF-5300 fluorescence

128

spectrometer (Shimadzu Co. Ltd., Japan) and a quartz cuvette (1 cm path length). Raw EEMs were

129

collected from 240 nm to 450 nm for excitation (Ex: at 5 nm intervals) and from 300 nm to 600

130

nm for emission (Em: at 2 nm intervals). EEM spectra were processed by spectral correction, inner 6 ACS Paragon Plus Environment

Page 6 of 33

Page 7 of 33


131

filter correction, Raman correction, and quinine sulfate calibration, and then analyzed using the

132

FDOMcorr toolbox.23 Excitation-emission pairs (peak picking)24 used in this study are described

133

in Table S2. Fluorescence intensities are reported in Raman units (RU). Fluorescence Index (FI)

134

was calculated by the ratio of the emission intensity at Em 450 nm relative to that at Em 500 nm

135

at Ex 370 nm. The freshness index (BIX) was calculated by taking the intensity ratio of Em 380

136

nm relative to that of Em 430 nm at Ex 310 nm.25

137

3. Results and Discussion

138

3.1. Characterization of NOM

139

Variations in the molecular composition (i.e., C%, O%, and H%) of NOM used in this study are

140

shown in Table S1. Generally, carbon content was higher in humic acids than fulvic acids and RO

141

isolates, and oxygen content was high for NOM from aquatic sources. The heterogeneity of natural

142

organic matter depends on its source (e.g., aquatic, soil, etc.), which also influences their

143

adsorption behavior.

144

(one humic acid), peat (one humic acid and one fulvic acid), and aquatic sources from rivers and

145

lakes (one humic acid, three fulvic acids, and three RO isolates).

146

Other major differences among NOM were SUVA254 and average molecular weight (MW) (Fig.

147

S1 and S2). Aromaticity, as indicated by SUVA254, was higher for humic acids [4.9−7.0 L/mg.m]

148

(LHA > PPHA > ESHA > SRHA) than fulvic acids [3.4−4.6 L/mg.m] (NLFA > SRFA2 > PPFA

149

> SRFA1) and RO isolates [2.6−3.6 L/mg.m] (SRNOM > NNOM > MNOM) (Table 1). This trend

150

is in agreement with previous studies by Hyung and Kim 10 and Rodríguez and Núñez 27. To the

151

best of our knowledge, among the total eleven NOM in this study, the average MW values of only

152

five NOM (i.e., ESHA, LHA, SRHA, SRFA1, and SRFA2) were previously reported and showed,

26

Sources of NOM in this study included soil (one humic acid), leonardite



Page 8 of 33

153

in general, good agreement with our measurements.2, 28-31 We also noted that some previous studies

154

reported incomparable average MW values for some of IHSS standard NOM, thus, they were

155

excluded from the comparison. For example, we found that reported average MW for NLFA by

156

Persson, et al. 32, ESHA by Heo, et al. 33, and SNOM by Wagoner, et al.

157

400%, and 1200% higher than average values, respectively. However, specific reasons for the

158

overestimated values in the studies mentioned above were not obvious due to the fact that: 1) some

159

studies did not report details of samples’ preparation (reported as N.A. in Table S3), and 2) the

160

limited number of samples in each study. Detailed comparison of the results and experimental

161

conditions between this and previous studies was summarized in Table S3.

162

EEM is being used frequently for NOM characterization in water owing to its high sensitivity,

163

good selectivity and preservation of samples.35 Therefore, two EEM indices, FI and BIX, were

164

used to categorize the source and age of NOM. FI tracks the source of NOM as microbial-derived

165

(e.g., organic matter from aquatic microorganisms) and terrestrially-derived.

166

values higher than 1.8 and less than 1.2 are classified as microbial-derived NOM and terrestrial-

167

derived NOM, respectively.25 FI values between 1.2 and 1.8 indicate that NOM is a mixture of

168

microbial and terrestrial aquatic substances. In the present study, FI values were higher for fulvic

169

acids (FA) [1.16−1.29] and RO isolates [1.33−1.49] than those of humic acids (HA) [0.84−1.0]

170

(Table 1). Higher FI for FA than HA was previously reported in Rodríguez, et al. 37, and, to the

171

best of our knowledge, FI values for standard IHSS RO isolates were not previously reported in

172

the literature. BIX is called index of recent autochthonous contribution and can be described by

173

the ratio between the recently-derived NOM and the decomposed (i.e., older) NOM.38 According

174

to BIX classification, all NOM in the present study appeared to be decomposed with BIX values

175

less than 0.6, and none was freshly produced (BIX > 1) (Table 1). BIX values for IHSS samples


34

were around 300%,

36

NOM with FI

Page 9 of 33


176

fall in the same range, between 0.2 and 0.4, in previous studies.39 NOM were further characterized

177

based on EEM peaks (Table S2) and showed variations in component compositions (P1−P5) as

178

illustrated in Fig. S3. Detailed characterizations of EEM peaks in this study are described in section

179

S2 in SI.

180

3.2. Effect of NOM molecular properties on the adsorption affinity

181

The results of adsorption isotherms and corresponding Freundlich model (Eq. 1) for eleven NOM

182

on four CNTs are presented in Table S4.

183

Qe  K F Ce1/ n

(1)

184

where; KF [(mg-C/g)/(mg-C/L)1/n] is Freundlich constant and 1/n is an exponent.

185

Fig. 1 shows the results of adsorption isotherm experiments for MWCNT C1 (Table 2). Freundlich

186

model showed good fit with the isotherms data as indicated by the coefficient of determinations

187

(R2) ranging between 0.87−0.98 (Table S4), which agrees with previous reports of NOM

188

adsorption on activated carbon2, carbon nanotubes10, and graphene nanomaterials 3.

189

Based on the NOM sources, adsorption capacities follow the order of: leonardite > peat > soil >

190

aquatic. According to NOM type, the adsorption affinity of humic acids was generally higher than

191

fulvic acids (Fig. 1 and Table S4). This was attributed to higher aromaticity of humic, as indicated

192

by their SUVA254 values (Table 1). Similar trends were previously reported for adsorption of

193

NOM by graphene and MWCNTs3, 10. Among NOM characteristics, SUVA254, an indicator for

194

aromaticity, showed a strong linear relationship with KF regardless of the source and the type of

195

NOM (Fig. 2A and S4) 10. The adsorption affinity was inversely proportional to the atomic ratio

196

of H/C and the polarity index (i.e., the atomic ratio (O+N)/C) (see Fig. 2C, 2D, S5). These findings

197

show that hydrophobic repulsion from water and π–π attractions are the main driving forces of 9 ACS Paragon Plus Environment


Page 10 of 33

198

adsorption between NOM and CNTs. In addition, carboxylic acidities of NOM showed a negative

199

correlation with the adsorption affinity (Fig. S6). A similar trend was previously reported for

200

adsorption of NOM with different carboxylic acidities on granular activated carbon.2 Lastly, NOM

201

with higher average molecular weight exhibited higher adsorption affinity (Fig. 2B), however, the

202

detailed effect of molecular weight will be further discussed in Section 3.4.

203

All adsorption results were also processed for the adsorption affinity (Qe) at three different

204

equilibrium concentrations (Ce = 5±1, 11±1, and 22±1 mg-C/L). According to our analysis, the

205

examined adsorption capacities showed similar trends to those of KF for the tested parameters (See

206

Fig. S4, S5, and S6).

207

Oxygen containing groups on CNTs’ surface give them more hydrophilic character and can result

208

in reducing the adsorption capacity of organic compounds

209

for the CNTs in this study fall in narrow range (i.e., 0.21−0.30 wt.% for MWCNTs and 1.3 wt.%

210

for SWCNTs) (Table 2). Thus, the adsorption dependency on the physical characteristics of CNTs

211

(i.e., surface area and pores size) was examined and discussed in section S3 in SI. BET results

212

showed that CNTs have different surface areas (98−380 m2/g) and pore volumes (0.4−0.9 cm3/g)

213

(Table 2). Due to the fact that micropores (< 2nm) are not likely accessible to all NOM molecules12,

214

only meso- (2 nm > PV> 50 nm) and macropores (> 50 nm) were taken into account. Basically,

215

the sorption coefficient (Kd = Q/Ce) was normalized to the surface area and pores volume of four

216

different CNTs (i.e., one SWCNTs and three MWCNTs) to adsorb SRHA, SRFA1, and SNOM.

217

Wang, et al.12 reported, in addition to the effect pores size, that normalized adsorption affinity

218

increased when they used CNTs with large outer diameter (OD). This can be linked to the enhanced

219

dispersion of CNTs with large OD compared to that with small OD41 and/or less curvatures of

220

CNT-surface allowing better alignment of adsorbate’s molecules.42

40

. However, surface oxygen content


Page 11 of 33

221


3.3. Preferential adsorption related to the source and the age of NOM

222

Lavonen, et al. 43 reported FI to be a good indicator to check the changes in NOM during several

223

different treatment processes in water treatment plants (e.g., coagulation, filtration, disinfection,

224

etc.). Similarly, Rodríguez, et al.37 applied FI to monitor the compositional changes in humic

225

substances after ozonation treatments. In another study by Kothawala, et al.

226

evaluate the characteristics of NOM after sorption to soil particles. In the present study, as listed

227

in Table 1, FI values for the bulk samples of NOM (i.e., before adsorption) indicated that all HAs

228

were more likely to be terrestrial-derived NOM (FI < 1.2). However, FAs and RO isolates share

229

the characteristics of mixtures of both terrestrial and microbial aquatic NOM (1.2 < FI < 1.8).25, 45

230

Fig. 3A shows that, FI values for all NOM types were increased after adsorption (FI > 1.2). This

231

is an indication of a preferential adsorption for terrestrial-derived compounds. This is consistent

232

with notion that NOM with low FI are rich with aromatic moieties (Fig. S7A), as previously

233

reported by Ateia, et al.46 and Shimabuku, et al.20 High negative correlation (R2 = 0.89) was also

234

found between FI and average MW of all NOM (Fig. S7B). This confirm the adsorption results in

235

the previous section, where the adsorption affinity of NOM with high average MW (i.e., HAs) was

236

higher than NOM with smaller average MW (i.e., FAs and RO isolates).

237

Unlike FI, the increasing trend of BIX was observed only for HAs, while BIX values were slightly

238

changed for FAs and RO isolates after adsorption on CNTs. BIX increment was more pronounced

239

for ESHA and LHA followed by PPHA and SRHA (Fig. 3B), however, the overall trend illustrates

240

the preferential adsorption for decomposed NOM fractions over freshly produced NOM. One

241

possible reason is, according to Hunt and Ohno 47 and Yu-Lai, et al. 48, that decomposed fractions

242

(i.e., low BIX) have higher C/N ratio and higher aromaticity than freshly produced NOM. Further,

243

Fig. S8 illustrates the gradual changes in FI and BIX with time in kinetic adsorption experiments 11 ACS Paragon Plus Environment

44

, FI was used to


Page 12 of 33

244

with six NOM (i.e., ESHA, SRHA, SRFA, NLFA, NNOM, and SNOM). From these results, FI

245

was more sensitive and suitable to describe and track the changes of NOM compositions.

246

NOM is a heterogeneous mixture of complex organic compounds. Therefore, preferable adsorptive

247

portions of NOM dominate the adsorption behavior at low concentration conditions and the

248

adsorption process would be limited by the available adsorption sites (i.e., surface area) when

249

elevating NOM concentrations.10 In this regard, further analysis showed that the variations in FI

250

and BIX were also concentration dependent. As shown in Fig. S9 and S10, changes in FI and BIX

251

were pronounced at equilibrium concentration (Ce) up to 5 mg-C/L, however, at high

252

concentrations they were similar to original values before adsorption. Thus, fluorescence indices

253

can serve as an easy and practical tool for tracking NOM fractionations during the adsorption

254

process on CNTs.

255

Although, EEM was applied successfully to track the changes of NOM compositions after

256

biological43 and physical37 treatment processes, very few studies utilized EEM as a tracking tool

257

for NOM preferential adsorption on graphene nanomaterials, e.g. Peng, et al.17 and Lee, et al.26. In

258

this study, five EEM peaks with different pairs reflected several fractions with variations in the

259

source and the size (Fig. S3 and Table S2).36, 49 Fig. 4 shows the normalized removal of each

260

fraction for six NOM as a function of time. Terrestrial and hydrophobic components showed the

261

highest preferential adsorption. This observation was reported for adsorption of landfill leachate

262

on activated carbon 50, and soil-derived HA on graphene oxide.26 It can also reinforce the argument

263

of the preferential adsorption of terrestrial fractions over microbial NOM similar to FI values.

264

Interestingly, the intensity of P5 [Excitation/Emission: 270/492] (Fig. 4E) did not decrease after

265

adsorption of all NOM tested in this study on CNTs. This peak represents aromatic amino acids


Page 13 of 33


266

(protein-like or tryptophan-like component).25,

267

checking molecular weight of the residual fractions, as discussed in details in the following section.

268

26

This observation was further examined by

3.4. Preferential adsorption based on the average molecular weight of NOM

269

NOM contain mixtures of compounds that vary in the aromatic content and adsorption capacities.

270

Among NOM used in this study, all humic acids, with wider molecular weight range (0.2−90 kDa)

271

and higher polydispersity (ρ = 2.7−5.8), showed similar preferential adsorption in the MW range

272

between 0.5 to 2 kDa. Meanwhile, the adsorption in MW range more than 1 kDa was more

273

preferable for all fulvic acids and RO isolates which have narrower molecular weight (0.2−10 kDa)

274

and lower polydispersity (ρ = 1.5−1.9) (Fig. S11). Thus, the molecular weight distribution was

275

changed after adsorption which was clearly reflected on the avg. MW of the residual fractions (Fig.

276

5). According to Gotovac, et al. 51, higher number of benzene rings in the aromatic structure of a

277

specific fraction increases its adsorption affinity on CNTs. This selectivity was examined and

278

tracked as function of time as visualized in Fig. S12. Further, the selective behavior was confirmed

279

by two ways; 1) increasing the CNT dose (increasing the SSA) (Fig. 6A), and 2) adsorption on

280

different types of CNTs (Fig. 6B). Previous studies on NOM adsorptive fractions on CNTs or

281

graphene oxides only reported that high MW fraction are more preferable for adsorption. However,

282

they solely drew their conclusion based on aquatic NOM

283

changes in NOM molecular compositions were not reported (e.g., avg. MW, MW distribution, and

284

polydispersity).

285

Another observation in this study was the ‘selective’ adsorption on CNTs for the fraction with

286

MW of 1−3 kDa from the bulk samples of FA and RO isolates (Fig. 6, Fig. S11 and Fig. S12).

287

According to Lv, et al.

288

comparison results between HPSEC and Fourier transform ion cyclotron resonance mass

52

3, 10

or soil FA 11, and details about the

, this fraction is likely to be aromatic molecular groups, based on the



Page 14 of 33

289

spectrometry (FT-ICR-MS). In another study by Li, et al.19 tested CNTs to adsorb styrene sulfonate

290

(SS) and polystyrene sulfonate (PSS) with different molecular weight values (0.2, 4.3, 6.8, 10.0,

291

17.0, and 70.0 kDa) and the highest adsorption capacity was for PSSs with avg. MW of 4.7 kDa.

292

It is important to highlight that, based on HPSEC, the small fraction (MW< 0.4 kDa) did not adsorb

293

on CNTs (Fig. 6, S11 and S12). This supports EEM results and can be linked to protein-like

294

compounds (Fig. 4E) which were not removed by CNTs as well. Cai, et al. 3 reported the same

295

‘no uptake’ behavior of small fraction for adsorption of NOM on graphene oxides. Therefore, we

296

interpret that NOM fractions with MW < 0.4 kDa are more likely to be hydrophilic compounds

297

that can’t adsorb on CNTs.22

298

In this study, the comprehensive characterization of NOM before and remaining after adsorption

299

along with the characterization of CNTs provided important insights to the adsorption of NOM by

300

CNTs. The results indicate that FI can serve an easy, and practical tool to describe and track the

301

fractionation of NOM during the adsorption process. The selective adsorption for the MW range

302

of 0.5−2 kDa for humic acids, and MW range of 1−3 kDa for all fulvic acids and RO isolates was

303

also revealed. Given the wide variety of NOMs in aquatic and soil environments, findings from

304

this study further contribute to clarify and predict the adsorption behavior of NOMs on CNTs in

305

natural waters and/or the potential applications for water treatment. However, further studies to

306

examine and identify the adsorptive fractions of NOM by CNTs using sensitive techniques (e.g.,

307

mass spectroscopies) are still needed to elucidate the factors controlling adsorption behavior of

308

this complex material.

309

Associated Content

310

Supporting Information. Adsorption kinetics (section S1, Table S5, Fig. S13). EEM peaks

311

description (section S2, Table S2, Fig. S3). Effect of CNT surface area and pore volume on 14 ACS Paragon Plus Environment

Page 15 of 33


312

adsorption (section S3, Fig. S14, S15). NOM characteristics (Table S1). HPSEC results (Table S3,

313

Fig. S1, S11, S12). Freundlich isotherm results (Table S4). SUVA254 vs. Avg. MW (Fig. S2). KF

314

vs. NOM molecular properties (Fig. S4, S5, S6). FI vs. SUVA254 and Avg. MW (Fig. S7). FI and

315

BIX at different times (Fig. S8) and with elevating concentrations (Fig. S9, S10).

316

Acknowledgments

317

This work was supported in part by Program for Leading Graduate School "Academy for Co-

318

creative Education of Environment and Energy Science", MEXT, Tokyo Institute of Technology

319

and the JSPS Core-to-Core Program. We would like to thank Mahmut Ersan, Department of

320

Environmental Engineering and Earth Science, Clemson University for providing BET results.

321

References

322 323

(1) Engel, M.; Chefetz, B., Adsorption and desorption of dissolved organic matter by carbon nanotubes: Effects of solution chemistry. Environmental Pollution 2016, 213, 90-98.

324

(2) Karanfil, T.; Kilduff, J. E.; Schlautman, M. A.; Weber, W. J., Adsorption of organic

325

macromolecules by granular activated carbon. 1. Influence of molecular properties under anoxic

326

solution conditions. Environmental science & technology 1996, 30, (7), 2187-2194.

327 328

(3) Cai, N.; Peak, D.; Larese-Casanova, P., Factors influencing natural organic matter sorption onto commercial graphene oxides. Chemical Engineering Journal 2015, 273, 568-579.

329

(4) Karanfil, T.; Kilduff, J. E.; Schlautman, M. A.; Weber Jr, W. J., The oxygen sensitivity of

330

organic macromolecule sorption by activated carbon: effects of solution chemistry. Water

331

Research 1998, 32, (1), 154-164.

332 333

(5) Bhatnagar, A.; Sillanpää, M., Removal of natural organic matter (NOM) and its constituents from water by adsorption–A review. Chemosphere 2017, 166, 497-510. 15 ACS Paragon Plus Environment


Page 16 of 33

334

(6) Santhosh, C.; Velmurugan, V.; Jacob, G.; Jeong, S. K.; Grace, A. N.; Bhatnagar, A., Role

335

of nanomaterials in water treatment applications: A review. Chemical Engineering Journal 2016,

336

306, 1116-1137.

337 338 339 340

(7) Ahmadi, M.; Elmongy, H.; Madrakian, T.; Abdel-Rehim, M., Nanomaterials as sorbents for sample preparation in bioanalysis: A review. Analytica Chimica Acta 2016, 958, 1-21. (8) Apul, O. G.; Karanfil, T., Adsorption of synthetic organic contaminants by carbon nanotubes: A critical review. Water research 2015, 68, 34-55.

341

(9) Smith, S. C.; Rodrigues, D. F., Carbon-based nanomaterials for removal of chemical and

342

biological contaminants from water: a review of mechanisms and applications. Carbon 2015, 91,

343

122-143.

344

(10) Hyung, H.; Kim, J.-H., Natural organic matter (NOM) adsorption to multi-walled carbon

345

nanotubes: effect of NOM characteristics and water quality parameters. Environmental science &

346

technology 2008, 42, (12), 4416-4421.

347 348

(11) Yang, K.; Xing, B., Adsorption of fulvic acid by carbon nanotubes from water. Environmental pollution 2009, 157, (4), 1095-1100.

349

(12) Wang, X.; Shu, L.; Wang, Y.; Xu, B.; Bai, Y.; Tao, S.; Xing, B., Sorption of peat humic

350

acids to multi-walled carbon nanotubes. Environmental science & technology 2011, 45, (21),

351

9276-9283.

352

(13) Lin, D.; Li, T.; Yang, K.; Wu, F., The relationship between humic acid (HA) adsorption

353

on and stabilizing multiwalled carbon nanotubes (MWNTs) in water: effects of HA, MWNT and

354

solution properties. Journal of hazardous materials 2012, 241, 404-410.

355 356

(14) Wang, F.; Yao, J.; Chen, H.; Yi, Z.; Xing, B., Sorption of humic acid to functionalized multi-walled carbon nanotubes. Environmental pollution 2013, 180, 1-6.


Page 17 of 33

357 358


(15) Engel, M.; Chefetz, B., Adsorptive fractionation of dissolved organic matter (DOM) by carbon nanotubes. Environmental Pollution 2015, 197, 287-294.

359

(16) Zhang, D.; Pan, B.; Cook, R. L.; Xing, B., Multi-walled carbon nanotube dispersion by the

360

adsorbed humic acids with different chemical structures. Environmental Pollution 2015, 196, 292-

361

299.

362

(17) Peng, M.; Li, H.; Li, D.; Du, E.; Li, Z., Characterization of DOM adsorption of CNTs by

363

using

364

Environmental Technology 2016, 38, 1351–1361.

365 366

excitation–emission

matrix

fluorescence

spectroscopy

and

multiway

analysis.

(18) Chen, W.; Duan, L.; Zhu, D., Adsorption of polar and nonpolar organic chemicals to carbon nanotubes. Environmental science & technology 2007, 41, (24), 8295-8300.

367

(19) Li, T.; Lin, D.; Li, L.; Wang, Z.; Wu, F., The kinetic and thermodynamic sorption and

368

stabilization of multiwalled carbon nanotubes in natural organic matter surrogate solutions: the

369

effect of surrogate molecular weight. Environmental Pollution 2014, 186, 43-49.

370

(20) Shimabuku, K. K.; Kennedy, A. M.; Mulhern, R. E.; Summers, R. S., Evaluating Activated

371

Carbon Adsorption of Dissolved Organic Matter and Micropollutants Using Fluorescence

372

Spectroscopy. Environmental Science & Technology 2017, 51, (5), 2676-2684.

373

(21) Chin, Y.-P.; Aiken, G.; O'Loughlin, E., Molecular weight, polydispersity, and

374

spectroscopic properties of aquatic humic substances. Environmental Science & Technology 1994,

375

28, (11), 1853-1858.

376

(22) Sandron, S.; Rojas, A.; Wilson, R.; Davies, N. W.; Haddad, P. R.; Shellie, R. A.;

377

Nesterenko, P. N.; Kelleher, B. P.; Paull, B., Chromatographic methods for the isolation,

378

separation and characterisation of dissolved organic matter. Environmental Science: Processes &

379

Impacts 2015, 17, (9), 1531-1567.



380

(23) Murphy, K. R.; Butler, K. D.; Spencer, R. G.; Stedmon, C. A.; Boehme, J. R.; Aiken, G.

381

R., Measurement of dissolved organic matter fluorescence in aquatic environments: an

382

interlaboratory comparison. Environmental science & technology 2010, 44, (24), 9405-9412.

383 384 385 386 387 388 389 390

Page 18 of 33

(24) Coble, P. G., Characterization of marine and terrestrial DOM in seawater using excitationemission matrix spectroscopy. Marine chemistry 1996, 51, (4), 325-346. (25) Coble, P. G.; Lead, J.; Baker, A.; Reynolds, D. M.; Spencer, R. G., Aquatic organic matter fluorescence. Cambridge University Press: 2014. (26) Lee, B.-M.; Seo, Y.-S.; Hur, J., Investigation of adsorptive fractionation of humic acid on graphene oxide using fluorescence EEM-PARAFAC. Water research 2015, 73, 242-251. (27) Rodríguez, F. J.; Núñez, L. A., Characterization of aquatic humic substances. Water and Environment Journal 2011, 25, (2), 163-170.

391

(28) Zhou, Q.; Cabaniss, S. E.; Maurice, P. A., Considerations in the use of high-pressure size

392

exclusion chromatography (HPSEC) for determining molecular weights of aquatic humic

393

substances. Water Research 2000, 34, (14), 3505-3514.

394

(29) Chin, Y.-P.; Aiken, G. R.; Danielsen, K. M., Binding of pyrene to aquatic and commercial

395

humic substances: the role of molecular weight and aromaticity. Environmental Science &

396

Technology 1997, 31, (6), 1630-1635.

397

(30) Hur, J.; Schlautman, M. A., Using selected operational descriptors to examine the

398

heterogeneity within a bulk humic substance. Environmental science & technology 2003, 37, (5),

399

880-887.

400

(31) McKay, G.; Kleinman, J. L.; Johnston, K. M.; Dong, M. M.; Rosario-Ortiz, F. L.; Mezyk,

401

S. P., Kinetics of the reaction between the hydroxyl radical and organic matter standards from the

402

International Humic Substance Society. Journal of soils and sediments 2014, 14, (2), 298-304.


Page 19 of 33


403

(32) Persson, L.; Alsberg, T.; Ledin, A.; Odham, G., Transformations of dissolved organic

404

matter in a landfill leachate—a size exclusion chromatography/mass spectrometric approach.

405

Chemosphere 2006, 64, (7), 1093-1099.

406

(33) Heo, J.; Yoon, Y.; Kim, D.-H.; Lee, H.; Lee, D.; Her, N., A new fluorescence index with a

407

fluorescence excitation-emission matrix for dissolved organic matter (DOM) characterization.

408

Desalination and Water Treatment 2016, 57, (43), 20270-20282.

409

(34) Wagoner, D. B.; Christman, R. F.; Cauchon, G.; Paulson, R., Molar mass and size of

410

Suwannee River natural organic matter using multi-angle laser light scattering. Environmental

411

science & technology 1997, 31, (3), 937-941.

412

(35) Li, W.-T.; Chen, S.-Y.; Xu, Z.-X.; Li, Y.; Shuang, C.-D.; Li, A.-M., Characterization of

413

dissolved organic matter in municipal wastewater using fluorescence PARAFAC analysis and

414

chromatography multi-excitation/emission scan: a comparative study. Environmental science &

415

technology 2014, 48, (5), 2603-2609.

416

(36) McKnight, D. M.; Boyer, E. W.; Westerhoff, P. K.; Doran, P. T.; Kulbe, T.; Andersen, D.

417

T., Spectrofluorometric characterization of dissolved organic matter for indication of precursor

418

organic material and aromaticity. Limnology and Oceanography 2001, 46, (1), 38-48.

419

(37) Rodríguez, F. J.; Schlenger, P.; García-Valverde, M., A comprehensive structural

420

evaluation of humic substances using several fluorescence techniques before and after ozonation.

421

Part I: Structural characterization of humic substances. Science of The Total Environment 2014,

422

476, 718-730.

423

(38) Huguet, A.; Vacher, L.; Relexans, S.; Saubusse, S.; Froidefond, J.-M.; Parlanti, E.,

424

Properties of fluorescent dissolved organic matter in the Gironde Estuary. Organic Geochemistry

425

2009, 40, (6), 706-719.



426

(39) Birdwell, J. E.; Engel, A. S., Characterization of dissolved organic matter in cave and

427

spring waters using UV–Vis absorbance and fluorescence spectroscopy. Organic Geochemistry

428

2010, 41, (3), 270-280.

429

(40) Apul, O. G.; Wang, Q.; Zhou, Y.; Karanfil, T., Adsorption of aromatic organic

430

contaminants by graphene nanosheets: Comparison with carbon nanotubes and activated carbon.

431

Water research 2013, 47, (4), 1648-1654.

432 433 434 435

Page 20 of 33

(41) Lin, D.; Xing, B., Tannic acid adsorption and its role for stabilizing carbon nanotube suspensions. Environmental science & technology 2008, 42, (16), 5917-5923. (42) Apul, O. G.; Shao, T.; Zhang, S.; Karanfil, T., Impact of carbon nanotube morphology on phenanthrene adsorption. Environmental Toxicology and Chemistry 2012, 31, (1), 73-78.

436

(43) Lavonen, E.; Kothawala, D.; Tranvik, L.; Gonsior, M.; Schmitt-Kopplin, P.; Köhler, S.,

437

Tracking changes in the optical properties and molecular composition of dissolved organic matter

438

during drinking water production. Water research 2015, 85, 286-294.

439 440

(44) Kothawala, D.; Roehm, C.; Blodau, C.; Moore, T., Selective adsorption of dissolved organic matter to mineral soils. Geoderma 2012, 189, 334-342.

441

(45) Cory, R. M.; McKnight, D. M.; Chin, Y. P.; Miller, P.; Jaros, C. L., Chemical

442

characteristics of fulvic acids from Arctic surface waters: Microbial contributions and

443

photochemical transformations. Journal of Geophysical Research: Biogeosciences 2007, 112, 1-

444

14.

445

(46) Ateia, M.; Ran, J.; Fujii, M.; Yoshimura, C., The relationship between molecular

446

composition and fluorescence properties of humic substances. International Journal of

447

Environmental Science and Technology 2017, 14, 867-880.


Page 21 of 33


448

(47) Hunt, J. F.; Ohno, T., Characterization of fresh and decomposed dissolved organic matter

449

using excitation-emission matrix fluorescence spectroscopy and multiway analysis. Journal of

450

agricultural and food chemistry 2007, 55, (6), 2121-2128.

451

(48) Yu-Lai, W.; Chang-Ming, Y.; Li-Min, Z.; Heng-Zhao, C., Spatial distribution and

452

fluorescence properties of soil dissolved organic carbon across a riparian buffer wetland in

453

Chongming Island, China. Pedosphere 2015, 25, (2), 220-229.

454

(49) Hertkorn, N.; Benner, R.; Frommberger, M.; Schmitt-Kopplin, P.; Witt, M.; Kaiser, K.;

455

Kettrup, A.; Hedges, J. I., Characterization of a major refractory component of marine dissolved

456

organic matter. Geochimica et Cosmochimica Acta 2006, 70, (12), 2990-3010.

457

(50) Lee, S.; Hur, J., Heterogeneous adsorption behavior of landfill leachate on granular

458

activated carbon revealed by fluorescence excitation emission matrix (EEM)-parallel factor

459

analysis (PARAFAC). Chemosphere 2016, 149, 41-48.

460

(51) Gotovac, S.; Honda, H.; Hattori, Y.; Takahashi, K.; Kanoh, H.; Kaneko, K., Effect of

461

nanoscale curvature of single-walled carbon nanotubes on adsorption of polycyclic aromatic

462

hydrocarbons. Nano Letters 2007, 7, (3), 583-587.

463

(52) Lv, J.; Zhang, S.; Wang, S.; Luo, L.; Cao, D.; Christie, P., Molecular-Scale Investigation

464

with ESI-FT-ICR-MS on Fractionation of Dissolved Organic Matter Induced by Adsorption on

465

Iron Oxyhydroxides. Environmental science & technology 2016, 50, (5), 2328-2336.

466 467 468 469 470 21 ACS Paragon Plus Environment


471

List of Tables:

472

Table 1. The types, codes and selected physicochemical and fluorescence properties of natural

473

organic matters used in this study.

474

Table 2. Characteristics of all carbon nanotubes used in this study.


Page 22 of 33

Page 23 of 33

475 476


Table 1. The types, codes and selected physicochemical and fluorescence properties of natural organic matters used in this study. Type

Humic Acids

Fulvic Acids

RO Isolates 477 478 479 480 481

Code

Cat. No.

SUVA254a (L/mg.m)

MW (kDa)b

Mn (kDa)b

ρ(-)c

ESHA

4S102H

6.2

5.8

1.0

PPHA

1S103H

6.9

6.2

LHA

1S104H

7.0

SRHA

2S101H

SRFA1

EEM Indices FI d

BIX e

5.8

0.84

0.25

1.5

4.2

0.86

0.24

4.8

1.0

4.8

0.94

0.28

4.9

3.9

1.4

2.7

1.00

0.25

2S101F

3.5

1.8

1.0

1.8

1.19

0.27

SRFA2

1S101F

4.0

1.8

1.2

1.5

1.16

0.24

PPFA

2S103F

3.9

1.4

0.9

1.7

1.29

0.30

NLFA

1R105F

4.6

1.8

1.0

1.8

1.20

0.28

NNOM

1R108N

3.6

1.3

0.6

1.9

1.33

0.31

MNOM

1S110N

2.6

1.3

0.7

1.7

1.49

0.39

SNOM

1R101N

3.3

1.5

0.8

1.8

1.34

a

0.30 (m-1)

The specific UV absorption (SUVA) values were calculated by dividing the UV absorbance at 254 nm divided by the DOC concentration (mg/L). b weight average MW and number average MW were measured using HPSEC. c Polydispersity (ρ) was calculated as MW/Mn, d Fluorescence Index (FI) was calculated by the ratio of the emission intensity at Em 450 nm relative to that at Em 500 nm at Ex 370 nm. e The freshness index (BIX) was calculated by taking the intensity ratio of Em 380 nm relative to that of Em 430 nm at Ex 310 nm.

482 483 484 485 486 487 488 489



490

491 492 493 494

Page 24 of 33

Table 2. Characteristics of all carbon nanotubes used in this study.

Code

Type

SA (m2/g)a

C1

MWCNT

136

C2

SWCNT

380

C3

MWCNT

114

C4

MWCNT

96

PVTotal (cm3/g)a

PVMicro (cm3/g) [%]a

PVMeso (cm3/g) [%]a

PVMacro (cm3/g) [%]a

0.45 [100] 0.9 [100] 0.47 [100]

0.02 [4] 0.05 [5] 0.02 [3]

0.34 [68] 0.81 [91] 0.36 [65]

0.14 [28] 0.04 [4] 0.18 [32]

0.4 [100]

0.01 [2]

0.34 [76]

0.10 [22]

OD (nm)b

Oxygen Content (wt.%)c

Purity (%)b

Supplier

3-20

0.30

> 95

Wako Chemicals, Japan

1-2

1.32

> 90

Alpha Nano, Chinese Academy of Science, China

5-10

0.21

> 98


10-20

0.25

> 98


a

The BET surface areas, pore volumes and pore size distributions were measured from nitrogen physisorption data at 77 °K obtained with ASAP 2020 analyzer (Micromeritics Instrument Corp. U.S.). b Information provided by the suppliers. SA is surface area, PVtotal is the total pore volume, PVMicro is the volume of micropores (i.e., PV < 2nm), PV Meso is the volume of mesopores (i.e., 2nm 50nm). b Information provided by the suppliers. c The oxygen contents were analyzed by using a Flash Elemental Analyzer 1112 series (Thermo Electron Corporation).


Page 25 of 33


495

List of Figures

496

Fig. 1. Isotherm results for [A] humic acids, [B] fulvic acids, [C] RO isolates used in this study.

497

The initial NOM concentrations was increased with fixing the same CNT-dose (C1: SSA = 134

498

m2/g) at 250 mg-CNT/L and at the same equilibrium time of 24 h. Experimental data were fitted

499

to Freundlich model (dashed line).

500

Fig. 2. Relationship between Freundlich constant KF with [A] SUVA254, [B] average molecular

501

weight, [C] H/C elemental ratio, and [D] O/C elemental ratio for all NOM used in this study

502

(ESHA: Gray solid circle, PPHA: Solid square, LHA: Solid diamond, SRHA: Gray solid Triangle,

503

SRFA1: Open circle, SRFA2: Open square, PPFA: Open diamond, NLFA: Open triangle, NNOM:

504

star, MNOM: cross, SNOM: ×).

505

Fig. 3. Changes in values of [A] florescence index (FI) and [B] freshness index (BIX) for all NOM

506

in this study. Ce is 2−3 mg-C/L and equilibrium time is 24 h, at pH 7±0.2. T: terrestrial derived

507

DOM and M: microbial derived DOM.

508

Fig. 4. Kinetics of EEM peaks [P1-P5] for NNOM, SNOM, SRFA1, NLFA, ESHA, and SRHA.

509

Initial NOM concentration was 5 mg/L and CNT concentration was 250 mg/L at pH 7 ± 0.2.

510

Fig. 5. Changes in the average MW for all NOM used in this study. Ce is 2−3 mg-C/L and

511

equilibrium time is 24 h, at pH 7±0.2.

512

Fig. 6. HPSEC graphs for adsorption experiments with Suwanee River different types of natural

513

organic matter (i.e., SRHA, SRFA, and SNOM) for [A] experiments at the same initial NOM

514

concentration (10 mg/L) with five different CNT-doses of one type of CNTs (C1: 134 m2/g), and

515

[B] experiments at the same initial NOM concentration (10 mg/L) with the same CNT dose (250

516

mg-CNT/L) of four different CNTs (C2: SWCNT, and C1,C3, C4: MWCNT). Green color 25 ACS Paragon Plus Environment


Page 26 of 33

517

highlights the MW-range for the preferential adsorptive fraction. Red color highlights the MW-

518

range for the non-preferable fraction.

519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 26 ACS Paragon Plus Environment

Page 27 of 33


542 543

Fig. 1. Isotherm results for [A] humic acids, [B] fulvic acids, [C] RO isolates used in this study.

544

The initial NOM concentrations was increased with fixing the same CNT-dose (C1: SSA = 134

545

m2/g) at 250 mg-CNT/L and at the same equilibrium time of 24 h. Experimental data were fitted

546

to Freundlich model (dashed line).



Page 28 of 33

547 548

Fig. 2. Relationship between Freundlich constant KF with [A] SUVA254, [B] average molecular

549

weight, [C] H/C atomic ratio, and [D] Polarity index (the atomic ratio (O+N)/C) ratio for all NOM

550

used in this study (ESHA: Gray solid circle, PPHA: Solid square, LHA: Solid diamond, SRHA:

551

Gray solid Triangle, SRFA1: Open circle, SRFA2: Open square, PPFA: Open diamond, NLFA:

552

Open triangle, NNOM: star, MNOM: cross, SNOM: ×).

553 554 555 556 557 28 ACS Paragon Plus Environment

Page 29 of 33


558 559

Fig. 3. Changes in values of [A] florescence index (FI) and [B] freshness index (BIX) for all NOM

560

in this study. Ce is 2−3 mg-C/L and equilibrium time is 24 h, at pH 7±0.2. T: terrestrial derived

561

DOM and M: microbial derived DOM.

562 563 564 565 566 567 29 ACS Paragon Plus Environment


568 569

Fig. 4. Kinetics of EEM peaks [P1-P5] for NNOM, SNOM, SRFA1, NLFA, ESHA, and SRHA.

570

Initial NOM concentration was 5 mg/L and CNT concentration was 250 mg/L at pH 7 ± 0.2.

571 572 573 574 575 576 577 578 579 580 581 582 583 30 ACS Paragon Plus Environment

Page 30 of 33

Page 31 of 33


584 585

Fig. 5. Changes in the average molecular weight for natural organic matters used in this study. Ce

586

is 2−3 mg-C/L and equilibrium time is 24 h, at pH 7±0.2.

587 588 589 590 591 592 593 594 595 596 597



Page 32 of 33

598 599 600 601 602 603 604 605

Fig. 6. Molecular weight distribution for adsorption experiments with Suwanee River humic acid, fulvic acid, and RO isolates for [A] experiments at the same initial NOM concentration (10 mg/L) with five different CNT-doses of one type of CNTs (C1: 134 m2/g), and [B] experiments with four different CNTs (C2: SWCNT, and C1,C3, C4: MWCNT) at the same initial NOM concentration (10 mg/L) with the same CNT dose (250 mg-CNT/L). Green color highlights the MW-range for the preferential adsorptive fraction. Red color highlights the MW-range for the non-preferable fraction.

606 32 ACS Paragon Plus Environment

Page 33 of 33

607


Graphical abstract

608 609 610


Elucidating Adsorptive Fractions of Natural Organic Matter on Carbon

Recommend Documents