Sorption of Heterocyclic Organic Compounds to Multiwalled Carbon

Dec 19, 2017 - Centre for Water and Environmental Research (ZWU), University of Duisburg−Essen Universitätsstrasse 2, 45141 Essen, Germany...
1 downloads 0 Views 475KB Size
Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Article

Sorption of heterocyclic organic compounds to multi-walled carbon nanotubes Florian Metzelder, Matin Funck, and Torsten Claus Schmidt Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b05205 • Publication Date (Web): 19 Dec 2017 Downloaded from http://pubs.acs.org on December 19, 2017

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

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 27

Environmental Science & Technology

Table of Contents/Abstract Graphic 83x40mm (120 x 120 DPI)

ACS Paragon Plus Environment

Environmental Science & Technology

1

Sorption of heterocyclic organic compounds to

2

multi-walled carbon nanotubes

3

Florian Metzeldera; Matin Funcka; Torsten C. Schmidta,b,c*

4 5

a

Instrumental Analytical Chemistry, University of Duisburg-Essen, Universitätsstrasse 5, 45141 Essen, Germany

6 7

b

Centre for Water and Environmental Research (ZWU), University of Duisburg-Essen, Universitätsstrasse 2, 45141 Essen, Germany

8 9

c

IWW Water Centre, Moritzstraße 26, 45476 Mülheim an der Ruhr, Germany

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

*Corresponding author (current address):

25 26 27 28 29 30 31 32

Torsten C. Schmidt Instrumental Analytical Chemistry University of Duisburg-Essen Universitätsstrasse 5 45141 Essen, Germany phone: +49-201-1836774 fax: +49-201-1836773 e-mail: [email protected] ACS Paragon Plus Environment

Page 2 of 27

Page 3 of 27

Environmental Science & Technology

33

Abstract

34

Sorption is an important natural and technical process. Sorption coefficients are typically

35

determined in batch experiments but this may be challenging for weakly sorbing compounds.

36

An alternative method enabling analysis of those compounds is column chromatography. A

37

column packed with the sorbent is used and sorption data are determined by relating sorbate

38

retention to that of a non-retarded tracer. In this study, column chromatography was applied

39

for the first time to study sorption of previously hardly investigated heterocyclic organic

40

compounds to multi-walled carbon nanotubes (MWCNTs). Sorption data for these

41

compounds are very limited in literature and weak sorption is expected from predictions.

42

Deuterium oxide was used as non-retarded tracer. Sorption isotherms were well described by

43

the Freundlich model and data showed reasonable agreement with predicted values. Sorption

44

was exothermic and physisorption was observed. H-bonding may contribute to overall

45

sorption, which is supported by reduced sorption with increasing ionic strength due to

46

blocking of functional groups. Lowering pH reduced sorption of ionizable compounds, due to

47

electrostatic repulsion at pH 3 where sorbent as well as sorbates were positively charged.

48

Overall, column chromatography was successfully used to study sorption of heterocyclic

49

compounds to MWCNTs and could be applied for other carbon-based sorbents.

50

2 ACS Paragon Plus Environment

Environmental Science & Technology

51

1. Introduction

52

In the field of chemistry, research concerning carbon-based nanomaterials (CNMs)

53

and carbon nanotubes (CNTs) as one class of CNMs often takes advantage of their strong

54

sorption properties.1-5 Strong sorption of different compound classes led to the application of

55

for example multi-walled carbon nanotubes (MWCNTs) as solid phase extraction material for

56

enrichment of organic1, 4, 5 and inorganic compounds2, 3 prior to analysis.

57

To characterize these sorption properties and study the influence of environmental

58

conditions on sorption, batch experiments are typically conducted for organic6-17 and

59

inorganic compounds.18-22 Batch experiments are simple to perform, but are accompanied by

60

several drawbacks. Firstly, the long equilibration time from hours23 up to several weeks6, 7

61

applied in literature in case of organic compounds makes experiments time intensive. This

62

especially increases the time expenses to study the influence of environmental conditions on

63

sorption if preliminary tests are necessary to adjust sorbent and sorbate amount needed to

64

yield sufficient but not too strong / weak sorption. Secondly, the limited sorbent to solution

65

ratio in batch experiments may reveal difficulties in the determination of reliable sorption data

66

for highly soluble and weakly sorbing compounds (e.g., heterocyclic organic compounds like

67

pyridine).24 Thirdly, a high amount of sorbent is necessary in case of weak sorption to

68

investigate sorption, especially, when investigating the influence of environmental conditions

69

on sorption in batch experiments. As an example, for sorption of perchlorate to MWCNTs in

70

batch experiments, 37 mg sorbent were necessary when studying sorption at one pH value.18

71

Using ten different pH values in triplicates would require more than one gram of sorbent. In

72

case of more expensive sorbents like single-walled carbon nanotubes or functionalized CNTs,

73

one gram can cost up to 1000 euro. Finally, separation of CNMs from the water phase may be

74

a challenging step, especially in case of dispersed materials25, 26. In this case, typically used

75

techniques like filtration or sedimentation may be insufficient.6 Additionally, the

76

quantification of low concentrations of CNTs in matrices is challenging so that the efficiency

77

of phase separation processes can be hardly quantified.27 As consequence of these drawbacks,

78

sorption models like poly-parameter linear free-energy relationships (ppLFERs) were

79

developed for MWCNTs allowing the prediction of distribution coefficients (Kd) for organic

80

compounds,28,

81

previously unknown compounds. However, these models are only valid for one setup of

82

environmental conditions, which limits their applicability for varying environmental

29

without the necessity to perform additional sorption experiments for

3 ACS Paragon Plus Environment

Page 4 of 27

Page 5 of 27

Environmental Science & Technology

83

conditions. It has been previously shown that pH13, 15, ionic strength30, 31 and temperature32, 33

84

influence sorption properties of sorbates and CNMs.34

85

As alternative to batch experiments, column experiments can be applied to study

86

sorption, by determination of Kd values as well as complete isotherms using the retardation of

87

the sorbate in relation to a non-retarded tracer in a sorbent packed column. This experimental

88

approach has been used successfully for sorption studies to soot particles,35 soils24,

89

mineral surfaces,46, 47. Recently, sorption of inorganic anions to MWCNTs was also studied in

90

column experiments showing the general applicability of this approach for MWCNTs.48

91

Drawbacks of batch experiments can potentially be overcome by column experiments. The

92

analysis of weakly sorbing and highly soluble compounds is possible because of higher

93

sorbent to solution ratios without a simultaneous need of large amounts of sorbent material.

94

Furthermore, changing the composition of the mobile phase regarding ionic strength, type of

95

ions and pH or by changing the column temperature using a thermostat facilitates

96

investigations of the influence of environmental conditions on sorption. Additionally,

97

equilibration times can be significantly reduced. Additional equipment like centrifuges can be

98

avoided, as the sorbent remains fixed in the column.

36-45

or

99

However, column experiments also may have disadvantages. Column clogging and too

100

strong sorption can be serious problems, which lead to very high backpressure or cause long

101

retention times and deterioration of the signal by peak broadening, respectively. Concomitant,

102

backpressure and sorption can both be reduced by the dilution of the sorbent material with

103

inert material like quartz24,

104

with the sorbate. The dilution with inert material further reduces the sorbent demand and,

105

consequently, costs to purchase those materials. An important disadvantage, potentially

106

preventing the applicability of column experiments, is sorption hysteresis. Sorption hysteresis

107

describes the incomplete desorption of a sorbate from the sorbent. Using a Dirac input (i.e.,

108

injection of a sorbate pulse instead of a continuous sorbate input) desorption is a prerequisite

109

in column experiments because otherwise no signal can be recorded. If sorbate molecules

110

remain sorbed to the surface, changed interactions between the sorbent surface and sorbate

111

molecules will be the consequence in consecutive measurements and this will result in

112

potentially misleading sorption data. For the measurement of breakthrough curves, sorption

113

hysteresis is less problematic, but each column could only be used for one compound and one

114

introduced concentration. For further consecutive measurements the sorbent surface would

115

not be pristine as sorbate molecules remain sorbed on the surface. Sorption hysteresis has

40

or silicium carbide37, which should not show any interaction

4 ACS Paragon Plus Environment

Environmental Science & Technology

116

already been shown for sorption of some organic compounds to CNMs in batch

117

experiments49-51. However, for inorganic anions, sorption hysteresis to MWCNTs was not

118

observed in a previous study using column experiments.48

119

Thus, the aim of this study was to investigate sorption of weakly sorbing organic

120

compounds to MWCNTs using column chromatography and to identify the influence of

121

environmental conditions like temperature, ionic strength and pH on sorption. Heterocyclic

122

organic compounds were chosen as sorbates for this purpose because only very limited

123

sorption data are available in literature for this compound class and sorption data predicted

124

using ppLFERs from literature28 indicated only weak sorption that may preferably be

125

analyzed using column chromatography instead of classical batch experiments.

126 127

2. Materials and Methods

128

2.1 Chemicals and Packing Materials

129

Chemicals used as sorbates in this study were heterocyclic organic compounds (furan,

130

imidazole, oxazole, pyrazine, pyrazole, pyridazine, pyridine, pyrimidine, pyrrole, thiazole and

131

thiophene). Detailed information on manufacturer and purity are given in section SI.1 in the

132

supporting information. Additional chemicals used in this study were hydrochloric acid (HCl)

133

(37 %, analytical grade, Fisher Scientific), sodium hydroxide (NaOH) (1 mol L-1, Bernd

134

Kraft), calcium chloride (CaCl2 × 2 H2O) (≥ 99.5%, AppliChem Panreac) and deuterium

135

oxide (D2O) (≥ 99.9atom-%, Sigma-Aldrich). Eluents for high performance liquid

136

chromatography (HPLC) were prepared with ultrapure water (18.2 MΩ cm-1 resistance,

137

PURELAB Ultra, ELGA LabWater, Celle, Germany) and addition of appropriate amounts of

138

salts, acid or base to reach the desired concentrations or pH value. Individual stock solutions

139

of each sorbate and tracer were prepared in the respective eluent used for measurements and

140

stored at 4°C in the dark.

141

MWCNTs (C150HP) were purchased from Bayer Material Science (Leverkusen,

142

Germany). General properties of MWCNTs used are given elsewhere.1, 48 Amorphous carbon

143

was removed prior to use by heating MWCNTs to 350°C for 1 h.52 Removal of amorphous

144

carbon by this treatment was confirmed previously using transmission electron microscopy.52

145

Quartz, used as inert material, with a particle size < 63 µm was purchased from Fluka (purity

146

≥ 99.7%). 5 ACS Paragon Plus Environment

Page 6 of 27

Page 7 of 27

Environmental Science & Technology

147 148

2.2 Apparatus

149

Empty HPLC-columns (length: 14; diameter: 3 mm) and supplementary equipment

150

used for column packing (precolumns, connecting nuts, stainless steel sieves (3 µm pore size),

151

glass fiber filters and PTFE seals) were purchased from Bischoff Chromatography (Bischoff

152

Analysentechnik u. –geräte GmbH, Leonberg, Germany).

153

The HPLC-system (Knauer Wissenschaftliche Geräte GmbH, Berlin, Germany) used

154

for column packing and sorption studies consisted the following parts: HPLC-pump (K-1001),

155

degasser, autosampler used for sample injection (Triathlon Spark Holland), column oven

156

(Jetstream 2 Plus), diode-array-detector (DAD) with deuterium lamp (K-2700 and K-2701)

157

connected to a refractive index (RI) detector (K-2301) in serial connection. A more detailed

158

description and a sketch of the experimental setup is given elsewhere.48 Dirac input of sorbate

159

solutions was done by an injection-loop-valve using 5 µL injection volume.

160 161

2.3 Column chromatography

162

A procedure used in literature to pack columns with reference soils was adapted for

163

this study.53 A detailed description of the procedure used in this study was reported

164

previously.48 MWCNTs were diluted with quartz as inert material by a factor of 20 for all

165

sorbates except thiophene, i.e., a MWCNT content of 5% (wt/wt). Dilution with quartz

166

increased both the stability of the packing material inside the column and reduced the

167

backpressure. Even with this dilution, sorption was strong leading to measurement times of up

168

to 90 minutes per measurement run in case of pyridine. For thiophene a dilution by a factor of

169

200, i.e., a MWCNT content of 0.5% (wt/wt), was necessary as retention was too strong in

170

columns containing 5% MWCNTs. In 5% MWCNT columns, measurement times of more

171

than 150 minutes and a maximum absorption of only 60 mAU even for an almost saturated

172

solution of thiophene were recorded because of a pronounced tailing. Maximum intensity in

173

the chromatogram was much lower compared to the other sorbates which was typically

174

around 1500 mAU for the highest injected concentration. To reach a similar range in

175

absorption, dilution of MWCNTs had to be increased.

176

The conservative tracer used in this study was D2O. D2O was diluted 1:9 with the

177

respective eluent (vol.-%/vol.-%) and detected using the RI-detector. Other tracers typically 6 ACS Paragon Plus Environment

Environmental Science & Technology

178

used in column experiments detectable using a DAD like inorganic anions showed, depending

179

on the environmental conditions, strong sorption to MWCNTs.48 Retention times of sorbates

180

and tracer used for calculation of the retardation factor (R) (equation 1) and Kd values

181

(equation 2) were determined based on the half-mass point, i.e., the point at which the peak

182

area is divided in two equal parts.54 Peaks of sorbates in this study showed a pronounced

183

tailing (e.g. asymmetry factors of peaks of pyridine determined at 10% of the maximum

184

intensity were up to 7.9). Standard deviations of R and Kd, determined by error propagation,

185

were typically smallest for the half-mass method compared to other typically used approaches

186

like the apex point of the eluting peak or the first moment approach36, 37 (data not shown).

187

Furthermore, the half-mass point has previously shown its applicability in sorption studies on

188

soil or minerals24, 39, 40, 46 but also for MWCNTs.48

189

R = (t – t*) / (t0 – t´)

(1)

190

R = 1 + ρb / θ x Kd

(2)

191

t is the retention time of the sorbate while t0 is the retention time of the tracer. t* is the

192

traveling time through injector and capillaries between autosampler and DAD without

193

column, while t´ is the traveling time from autosampler to the RI-detector without column.

194

Both detectors were connected in serial connection, which required different corrections for

195

the retention time of sorbate and tracer. t* and t´ were determined by repeated injections of

196

sorbate and tracer without column installed in the system (n = 10). For measurements without

197

column, capillaries were connected using a low dead volume stainless steel union (0.25mm

198

bore, VICI Jour.). θ [-] and ρb [kg L-1] are the porosity and the bulk density of the column

199

used, respectively. Both were calculated from repeated measurements of D2O with and

200

without column installed in the system (n = 10). Determined values for both characteristics of

201

columns used in this study are given in the supporting information (section SI.2). Recovery of

202

D2O was determined from peak areas with and without column and was found to be

203

quantitative (90 to 105%). Regular measurements of D2O during sorption studies showed no

204

significant shift of the retention time of D2O for more than 200 injections (data not shown).

205

Wavelengths used to record chromatograms of the different sorbates are given in

206

section SI.1 in the supporting information. Interaction of the different sorbates with the inert

207

material was checked in a column packed with exclusively quartz and was negligible (Kd

208

calculated from the quartz column < 0.1% of that calculated from the MWCNT column). Kd

209

values for pyrazole, which was used as reference compound for each MWCNT-column, 7 ACS Paragon Plus Environment

Page 8 of 27

Page 9 of 27

Environmental Science & Technology

210

varied with a relative standard deviation of 6.2% (ten columns), when injecting the same

211

concentration in each column. This indicates an acceptable reproducibility for self-packed

212

columns. Mass recovery of sorbates was calculated by comparison of peak areas of each

213

sorbate from MWCNT-columns with those from measurements without column or a quartz

214

column. Recoveries ranged between 80 and 120% for all sorbates with exceptions for very

215

low injected concentrations. In these cases, the measured absorption was very low and

216

baseline noise became an issue because of peak broadening. Conclusively, sorption hysteresis

217

was not observed in this study as recovery was quantitative. Error bars depicted in figures

218

indicate the maximum error, which was derived from the individual standard deviations of

219

retention times of sorbate and tracer with and without column via error propagation.

220

Columns were equilibrated for at least 12 hours by flushing at 0.1 mL min-1 with the

221

respective eluent. Preliminary measurements were performed for each sorbate at 10 mM

222

CaCl2, pH 6 and 25 °C for different flow rates / pore water velocities to test the degree of non-

223

equilibrium in the experiments. One concentration of each sorbate was separately injected

224

(concentrations are mentioned in section SI.3 in the supporting information). Sorption was at

225

or very close to equilibrium, because determined R values varied only slightly with maximum

226

8% at different flow rates / velocities ranging from 0.015 – 0.1 mL min-1 or 0.33 – 2.2 cm

227

min-1, respectively (shown for selected compounds in section SI.3 in the supporting

228

information). Because of the small influence of the flow rate, 0.1 mL min-1 was used as flow

229

rate for all experiments.

230 231

2.4 Isotherm determination

232

Isotherms of sorbates were determined using an eluent with 10 mM CaCl2 at pH 6 and

233

25 °C controlled by the column thermostat, as these conditions were also applied in previous

234

studies using the same type of MWCNTs.28, 48 Sorbates were injected separately in MWCNT

235

columns with increasing concentration and three replicates for each concentration with

236

exception of pyridine, where two replicates were used due to long measurement times. At

237

least nine different concentrations were injected and used for isotherm generation of each

238

sorbate. The injected concentrations of each sorbate are given in section SI.4 in the supporting

239

information.

240

Different approaches have been used in literature to construct sorption isotherms from

241

column studies.35-37 In this study, the approach used by Schenzel et al.37 was adapted. The 8 ACS Paragon Plus Environment

Environmental Science & Technology

242

eluted peak of the respective sorbate was used to derive the representative dissolved

243

concentration at equilibrium (cw,eq). Using an external calibration on a spectrophotometer

244

(Shimadzu UV-1650PC, Shimadzu Europe GmbH, Duisburg, Germany), the absorption in the

245

chromatogram at the half-mass point was converted into cw,eq. Calibrations used for isotherm

246

calculation are given in section SI.5 in the supporting information. cw,eq was multiplied with

247

the corresponding Kd value to finally yield the corresponding sorbed concentration at

248

equilibrium (cs,eq). This approach is later referred as the “classical approach”.

249

As alternative method to derive sorption isotherms, the approach by Schenzel et al.37

250

was adapted and combined with the approach by Bürgisser et al..55 For this method, the whole

251

tailing of the peak in the chromatogram was used to construct the sorption isotherm from one

252

injected concentration of the sorbate instead of the injection of several concentrations.

253

Bürgisser et al.55 derived the sorption isotherm directly from the chromatogram. The highest

254

concentration injected of each sorbate was used for this purpose. The absorption in the

255

chromatogram was transferred into the dissolved concentration at equilibrium (cw,eq) using the

256

external calibration starting at the determined half-mass point in the chromatogram. For each

257

data point in the chromatogram following the half-mass point, a Kd value was calculated using

258

equation 1 and 2, which was multiplied with the corresponding cw,eq in order to yield cs,eq.

259

This approach is later referred as the “alternative approach”. Both approaches are explained

260

stepwise in the supporting information (section SI.4) The similarity between both approaches

261

was evaluated because applying the alternative approach reduces the time necessary for

262

determination of an isotherm almost by a factor of ten compared to the classical approach in

263

case of nine different concentrations used for isotherm construction.

264

2.5 Influence of environmental conditions on sorption

265

To reduce the time expense in studying the influence of environmental conditions on

266

sorption, a selection of heterocyclic compounds was chosen and only one concentration but

267

still three replicates per sorbate were injected. Evaluation was done based on the “alternative

268

approach”. Compound selection was done according to the sorption strength / retention inside

269

the column represented by the Freundlich-coefficient (Kf) of isotherms previously determined

270

(see chapter 2.4), water-air distribution by the Kwa-value and the pKa-value of the

271

corresponding acid of the compound (section SI.7 in the supporting information). A broad

272

compound range was represented by the selected compounds, as weakly to strongly sorbing

273

compounds, compounds with low to high Kwa-values were present and compounds of which

274

speciation is changed or remains the same by changing the pH value were used. Mean values 9 ACS Paragon Plus Environment

Page 10 of 27

Page 11 of 27

Environmental Science & Technology

275

of the three replicates of Kd as well as Kf and the Freundlich exponent (n) derived from

276

chromatograms at varied environmental conditions are reported. Injected concentrations for

277

the different conditions are given in the supporting information (section SI.7).

278

1) Temperature: Column CNT52 and CNT53; flow rate of 0.1 mL min-1; 10 mM CaCl2

279

in the mobile phase; pH 6; 25, 40 and 55 °C. CNT52 was used for the high

280

concentrations of each compound, while CNT53 was used for the low concentrations.

281 282

2) Ionic strength: Column CNT54; flow rate of 0.1 mL min-1; 25°C; pH 6; 1, 10 or 100 mM CaCl2 in the mobile phase.

283

3) pH value: Column CNT55; flow rate of 0.1 mL min-1; 25°C; pH 3, 6 or 9 using 10

284

mM CaCl2. Ultrapure water with 10 mM CaCl2 had the desired pH value of 6. NaOH

285

was used to change the pH to 9. At pH 3, a concentration of 1 mM HCl was added to

286

reach the desired pH value. For pH 3 and 9 the ionic strength was chanced by less than

287

5% compared to pH 6 and was assumed to be constant.

288

Variation of ionic strength at pH 3: Column CNT56; flow rate of 0.1 mL min-1; 25°C;

289

pH 3; 1 mM CaCl2 + 1 mM HCl, 10 mM CaCl2 + 1 mM HCl or 100 mM CaCl2 + 1

290

mM HCl in the mobile phase.

291

3. Results and discussion

292

3.1 Determination of sorption isotherms

293

Different heterocyclic organic compounds were analyzed to determine sorption

294

isotherms. With increasing injected concentration, the apex point of the peak increases in

295

intensity and decreases in retention time. Selected chromatograms of pyrazole for different

296

injected concentrations are shown in figure S4 in the supporting information (section SI.4).

297

The same behavior is observed for the determined half-mass point and for all compounds

298

investigated in this study.

299

Consequently, increasing injected concentration resulted in decreasing Kd values. In

300

case of pyrazole, Kd decreased by more than a factor of five (111.8 L kg-1 for 15 mg L-1

301

injected concentration and 19.5 L kg-1 for 1000 mg L-1). For the other compounds, the

302

decrease was similar relative to the respective values. Decrease in Kd was much stronger

303

compared to a previous study dealing with sorption of inorganic anions to MWCNTs in

304

column experiments.48 For inorganic anions, like iodide, the decrease was at maximum about

305

50% but in most cases only about 20% or less, when increasing the injected concentration

306

from 0.6 to 63 mg L-1. 10 ACS Paragon Plus Environment

Environmental Science & Technology

Page 12 of 27

307

Strongly decreasing Kd values with increasing sorbate concentrations are a strong

308

indicator for non-linear sorption. Sorption isotherms derived from chromatograms were

309

calculated according to both methods mentioned in section 2.4 and fitted using the Freundlich

310

model (cs,eq = Kf x cw,eqn), which is often used in batch experiments on sorption to CNTs.10, 13,

311

15, 16, 28, 33, 56-58

312

information. values for Kf, n and correlation coefficients (R2) of all sorbates are summarized

313

in table 1.

314 315 316 317 318 319

Table 1: Determined Freundlich coefficients (Kf) [(mg kg-1) (mg L-1)-n], Freundlich exponents (n) [-] and correlation coefficients (R2) [-] of isotherms of sorbates calculated based on the classical and the alternative approach. Uncertainties given in the table are the standard errors of the regression with Kf as y-intercept and n as slope of the regression for the classical approach and the standard deviation of the three replicates for the alternative approach.

Selected isotherms are also shown in an diagram in figure S5 in the supporting

Classical approach sorbate furan imidazole oxazole pyrazine pyrazole pyridazine pyridine pyrimidine pyrrole thiazole thiophenea

Kf

n

132.2 ± 2.3 39.3 ± 0.3 36.5 ± 0.8 222.2 ± 5.2 57.6 ± 2.9 137.9 ± 0.4 243.1 ± 3.0 154.0 ± 4.9 65.2 ± 2.2 158.4 ± 3.3 127.5 ± 0.8

0.60 ± 0.01 0.69 ± 0.01 0.77 ± 0.01 0.58 ± 0.01 0.71 ± 0.02 0.64 ± 0.01 0.53 ± 0.01 0.67 ±0.01 0.71 ± 0.02 0.59 ± 0.01 0.74 ± 0.01

Alternative approach

R2 0.9995 0.9997 0.9978 0.9967 0.9920 0.9999 0.9985 0.9961 0.9963 0.9978 0.9999

Kf

n

R2

154.8 ± 0.7 48.8 ± 0.6 47.9 ± 0.2 231.7 ± 3.8 73.3 ± 0.2 144.9 ± 0.1 261.1 ± 2.9 177.0 ± 2.9 82.2 ± 0.1 187.9 ± 0.3 234.2 ± 1.3

0.54 ± 0.01 0.60 ± 0.01 0.68 ± 0.01 0.57 ± 0.01 0.65 ± 0.01 0.62 ± 0.01 0.50 ± 0.01 0.62 ± 0.01 0.62 ± 0.01 0.53 ± 0.01 0.61 ± 0.01

0.9998 ± 0.0001 0.9998 ± 0.0001 0.9994 ± 0.0001 0.9987 ± 0.0007 0.9965 ± 0.0001 0.9999 ± 0.0001 0.9994 ± 0.0001 0.9996 ± 0.0001 0.9997 ± 0.0001 0.9997 ± 0.0001 0.9991 ± 0.0001

320

a: determined with a column containing only 0.5% MWCNTs, which may result in slightly smaller values

321

compared to the other compounds.

11 ACS Paragon Plus Environment

Page 13 of 27

Environmental Science & Technology

322 323

For both evaluation approaches, data were well fit by the Freundlich model as

324

indicated by the high R2 (R2 > 0.992). Determined Kf values differed strongly between the

325

different sorbates. The smallest value was determined for oxazole while an about six times

326

higher value was determined for pyridine. The alternative approach gave generally higher

327

values for Kf, but results were in all cases very similar to the classical approach with less than

328

30% higher values for the alternative approach with exception of thiophene (84% higher). The

329

MWCNT content inside the column was lower for thiophene compared to the other sorbates,

330

which led to a stronger increase in signal intensity with increasing injected concentration and

331

a generally lower retention inside the column compared to the other sorbates (figure S6 and

332

S7 in the supporting information). This results in a larger deviation between both approaches

333

as the absorption used to calculate cw,eq in case of the alternative approach is much higher

334

compared to the absorption used for the classical approach. Determined n values indicate that

335

sorption of all sorbates was non-linear (0.50 – 0.75). Non-linearity was in all cases larger for

336

the alternative approach, but deviation was less than 20% for all compounds, which again

337

shows the similarity of results obtained by both evaluation approaches. To further compare

338

sorption isotherms, Kd values calculated based on the classical and alternative approach for

339

1%, 0.1% and 0.01% of the maximum solubility in water of the respective compound (figures

340

S5 – S6 in the supporting information (section SI.6)) agree well with each other with a

341

deviation of less than 0.3 log units. Consequently, the alternative approach proved its

342

suitability as Kf, n and Kd values are similar to those of the classical approach but requires

343

significantly reduced time. Non-linear sorption to CNTs is commonly found in literature for organic10, 13, 15, 16, 28,

344 345

33, 56-58

and inorganic compounds18,

20, 21

346

inorganic anions in a previous study48 or in a very low concentration range for PAHs6,

347

sorption to MWCNTs was found to be linear. When extending the concentration range also

348

for PAHs6 the sorption isotherms became non-linear. The non-linearity is attributed to the

349

heterogeneous energy distribution of sorption sites becoming relevant when reaching higher

350

concentrations. At low concentrations only high energy sorption sites like defects or

351

functional groups are occupied, while at higher concentrations also sorption sites exhibiting a

352

lower energy become relevant and sorption affinity decreases.34

in batch experiments. Only in rare cases, like for

353

Only for pyridine a sorption isotherm using MWCNTs as sorbent was found in

354

literature, but experimental conditions were different.59 Sorption experiments were done in 12 ACS Paragon Plus Environment

Environmental Science & Technology

355

batch experiments at pH 7 and 10, i.e. higher than the pH of 6 used in this study, which

356

influences the speciation of pyridine. The fraction of positively charged pyridine drops from

357

15% at pH 6 to 1.5% at pH 7, which may also result in changing interactions between sorbent

358

and sorbate due to charged sorbate molecules and sorbent surface. Sorption was found to be

359

almost linear as n was with 0.94 close to unity.59 Reason may be the small concentration

360

range investigated. In a narrow concentration range, linearity may always approach one

361

compared to a wider concentration range as already observed for PAHs.6 Kf was found to be

362

almost one order of magnitude higher in literature (2.5 x 103 (mg kg-1) (mg L-1)-n (changed

363

from unit (mmol kg-1) (mmol L-1)-n)).59 These observations show substantial differences

364

between both isotherms, which may be related to differences in the investigated concentration

365

range, environmental conditions (pH, ionic strength), type of MWCNTs and sorbent to

366

solution ratio limiting the comparability and resulting in large deviations. As further

367

experimental sorption data for the other sorbates are missing, Kd values for all compounds

368

were calculated from the respective isotherms and had to be compared to Kd values predicted

369

using ppLFERs.28

370

ppLFERs were derived from batch isotherms using the exact same environmental

371

conditions and sorbent material so that the experimental conditions are well comparable.28

372

Solute descriptors are mentioned in section SI.6 in the supporting information. The

373

comparison of Kd values is shown for 1%, 0.1% and 0.01% of the maximum solubility in

374

water in figures S8 – S10 in the supporting information (section SI.6). For oxazole,

375

pyrimidine, pyrrole and thiazole agreement of data calculated based on the isotherms of the

376

classical and alternative approach with ppLFER data was good. Deviation was less than 0.3

377

log-units at all three concentrations levels, which is typically an acceptable deviation for

378

predicted values. Higher deviations were observed for the other compounds of more than 0.5

379

log-units with exception of furan and pyrazole at 0.01%. Maximum deviation was 0.9 log-

380

units. For those compounds no general tendency could be observed, as predicted values were

381

larger than measured values for furan, pyridine and thiophene, while for imidazole, pyrazole

382

and pyridazine the opposite behavior was found. Regarding the relationship between structure

383

and sorption of heterocyclic compounds it could be observed that sulfur atoms may enhance

384

sorption. Thiazole has the heteroatoms in the same positions than imidazole and oxazole, but

385

has a sulfur atom instead of oxygen or nitrogen. Thiazole showed much stronger sorption at

386

the same relative concentration with regard to the calculated Kd values. Equally, thiophene

387

containing sulfur had higher Kd values compared to furan containing oxygen and pyrrole

388

containing nitrogen. For the ppLFER predictions it should be noted here, that predicted values 13 ACS Paragon Plus Environment

Page 14 of 27

Page 15 of 27

Environmental Science & Technology

389

are also connected with an uncertainty. Firstly, solute descriptors used to set up the ppLFERs

390

and predict Kd values are connected with an uncertainty. Secondly, mainly non-heterocyclic

391

compounds were used to set up the ppLFERs with the exception of larger heterocyclic

392

compounds indole and benzofuran, which may result in deviations for the compounds

393

analyzed in this study. Thirdly, only non-ionic species were used to set up the ppLFERs.28

394

Thus, pyridine and imidazole, which were partially charged at the environmental conditions

395

used in this study, are compounds with limited suitability of the ppLFERs. This may also

396

explain to some extent the larger deviations to predicted values for these two compounds.

397

In summary, column experiments proved suitable as fast technique to determine

398

sorption isotherms for heterocyclic organic compounds. Additionally, much smaller amounts

399

of approximately 8 mg sorbent material per sorbate and column were necessary because of the

400

dilution with quartz and the possibility to use columns for several consecutive measurements

401

applying different concentrations. Using batch experiments, the sorbent demand would have

402

been much higher. More than one gram of the sorbent per isotherm of a sorbate would be

403

necessary assuming triplicates and the sorbent demand used in case of sorption of

404

trihalomethanes to MWCNTs.33 The alternative approach in isotherm evaluation showed its

405

applicability and reduces the time demand for isotherm measurement by almost a factor of

406

ten. Consequently, for further experiments to evaluate the influence of environmental

407

conditions on sorption the alternative approach was used. Thus, besides Kd values also

408

isotherms could be derived using only one injected concentration measured in triplicates.

409 410

3.2 Influence of temperature

411

The influence of temperature on sorption is shown in figure 1. Sorption coefficients

412

decreased with increasing temperature for all analytes. The sorption enthalpy (∆H) was

413

calculated using the van´t Hoff equation from the slope of the regression line in figure 1

414

according to equation 3.

415

ln Kd = ∆S/R - ∆H/R x 1/T

416 417

(3)

Where ∆H is the sorption enthalpy [kJ mol-1], ∆S the entropy change [kJ mol-1 K-1], R the gas constant [8.3145 × 10-3 kJ mol-1 K-1] and T the temperature [K].

14 ACS Paragon Plus Environment

Environmental Science & Technology

ln Kd [L/kg]

148.4

54.6

Page 16 of 27

furan imidazole oxazole pyrazole pyridine pyrrole

20.1

7.4

0.0030

0.0031

0.0032

0.0033

0.0034

1/T [1/K]

418 419 420 421 422

Figure 1: Determined Kd values of the analytes plotted against the reciprocal temperature at 10mM CaCl2 in the eluent at pH 6. Error bars indicate the maximum error estimated from standard deviation of retention times of sorbate and tracer via error propagation (n=3).

423

The determined values for ∆H ranged between -20.6 and -27.1 kJ mol-1 (all values are

424

shown in section SI.8 in the supporting information). Consequently, sorption of these organic

425

compounds is an exothermic process. In literature, sorption of different organic compounds

426

like

427

ciprofloxacin50 or toluene, m-xylene, ethylbenzene23 to different CNTs was also found to be

428

exothermic. Conversely, sorption of 1,2-dichlorobenzene,13 sulfamethazine,64 or dyes14 to

429

different CNTs was found to be endothermic or showed in case of dissolved organic matter to

430

SWCNTs65 no temperature dependency. Consequently, there is no coherent thermodynamic

431

basis for all compound classes reported in literature.

trihalomethanes,33

atrazine,32,

60

nitroaromatics,61

phenols,15,

62

oxaquindox,63

432

Covalent bonding like in case of chemisorption is absent in this study as determined

433

values for ∆H were smaller than typically determined values in the range from -60 to -80 kJ

434

mol-1 for organic compounds.66 Consequently, physisorption seems to be the underlying

435

mechanism. This was also concluded in other studies dealing with sorption to CNTs.50, 61, 63

436

Additionally, this is supported by quantitative recovery and, consequently, fully reversible

437

sorption, which is characteristic for physisorption. H-bonding (-10 – -40 kJ mol-1)67 with

438

functional groups located on the MWCNTs surface may be responsible for interactions

439

between the analytes and sorbent. Functional groups offering this kind of interactions like -

440

COOH, C=O and C-OH were present on the surface of MWCNTs as already shown

441

previously.48 Additionally, also van-der-Waals (vdW) forces between sorbent and sorbates

442

may be contributing due to the molecular structure of the sorbates, although typical ∆H-values 15 ACS Paragon Plus Environment

Page 17 of 27

Environmental Science & Technology

443

are much smaller (vdW: -1 – -2 kJ mol-1)68 when exclusively these forces contribute to

444

sorption.

445

When the injected concentration of each analyte was reduced, the temperature effect

446

on sorption was more pronounced for all analytes with exception of oxazole. Values for ∆H

447

ranged at the lower injected concentration between -18.2 and -33.8 kJ mol-1 (all values are

448

shown in section SI.8 in the supporting information). Although the change in ∆H due to the

449

lower concentration was relatively small, it was higher than the measurement uncertainty. A

450

similar trend regarding ∆H with increasing concentration was also observed for sorption of

451

ciprofloxacin to SWCNTs50 or various heterocyclic compounds including pyridine to

452

reference soils.40 The more pronounced temperature dependency is related to a reduced

453

loading of the surface of MWCNTs with the consequence that energy distribution of sorption

454

sites is less spread compared to the higher injected concentration. At the high concentration,

455

energetically favorable and non-favorable sites are occupied, while at the lower concentration

456

a similar number of favorable sites and less non-favorable sites are occupied.34 This leads to

457

more negative values for ∆H. This was also concluded in case of sorption to reference soils.39

458

Sorption isotherms derived from measurements at various temperatures showed that Kf

459

decreased with increasing temperature while n only slightly increased (section SI.8 in the

460

supporting information). Decreasing Kf could be expected as sorption was generally reduced

461

with increasing temperature. Sorption linearity expressed by n was not influenced by

462

temperature. Similar observations regarding Kf and n were reported by Yan et al.60.

463 464

3.3 Influence of ionic strength

465

The influence of ionic strength on sorption of the selected compounds is shown in

466

figure 2. Determined Kd values reduced steadily but in total only slightly with increasing ionic

467

strength of the eluent for all compounds by 6 – 42% with exception of imidazole and pyridine.

468

For imidazole and pyridine, an overall tendency of reduced sorption with increasing ionic

469

strength was also observed, but Kd at 10 mM CaCl2 was even lower than those at 1 and

470

100 mM CaCl2. Imidazole and pyridine are the only compounds that are partially positively

471

charged (imidazole: ≈ 91%; pyridine: 15%) at the tested environmental conditions (pH 6),

472

while the sorbent was negatively charged at these conditions.48 Therefore, increased ionic

473

strength would constantly and more pronounced reduce sorption if electrostatic attraction

474

between charged sorbate molecules and the oppositely charged sorbent surface was 16 ACS Paragon Plus Environment

Environmental Science & Technology

475

significantly contributing to overall sorption. Consequently, electrostatic attraction seems to

476

be less important in this case and is contributing at the most slightly to sorption of these two

477

compounds, which may be related to the competition of positively charged sorbate molecules

478

with Ca2+ present in solution. Ca2+ may interact preferably with the negatively charged

479

sorbent surface shielding the negative charge of the surface visible to positively charged

480

sorbate molecules so that electrostatic attraction of those is prevented. 120

1 mM 10 mM 100 mM

100

Kd [L/kg]

80 60 40 20

py rr ol e

in e py rid

py ra zo le

id az ol e

ox az ol e

481

im

fu ra n

0

482 483 484 485

Figure 2: Determined Kd values of the different analytes for three different concentrations of CaCl2 in the eluent (1, 10 and 100 mM). Error bars indicate the maximum error estimated from standard deviation of retention times of sorbate and tracer via error propagation (n=3).

486

Sorption isotherms derived from measurements showed that Kf and n remained almost

487

constant or only slightly changed (section SI.9 in the supporting information). This could be

488

expected as Kd values were also only slightly influenced. Kf of pyridine reduced stronger but

489

this was a consequence of a spectral interference with the increasing chloride concentration,

490

which is further discussed in the supporting information (section SI.9).

491

Reduced sorption to CNTs of chlorophenols with increasing concentration of sodium

492

chloride (NaCl) was reported in literature, while in the same study increasing concentration of

493

potassium chloride (KCl) showed no or only minimal reduction.31 The authors assigned the

494

difference to deviations in pi-cation interactions of Na+ and K+ between sorbents and sorbate.

495

Also in case of roxarsone30 increased ionic strength reduced sorption. Decreasing sorption

496

was assigned to changes in CNT-aggregates reducing the available surface area and cations

497

that interact with functional groups and therefore inhibit interactions between functional

498

groups and sorbates. Also for inorganic anions48 or cations69 sorption decreased with

499

increasing ionic strength of the eluent. Similar effects are expected in this study as increased 17 ACS Paragon Plus Environment

Page 18 of 27

Page 19 of 27

Environmental Science & Technology

500

concentrations of ions in solution may interfere H-bonding interactions of polar heterocyclic

501

organic compounds with the sorbent as surface functional groups like COOH, C=O and C-OH

502

may be blocked by ions from the liquid matrix accumulating at these locations.

503 504

3.4 Influence of pH

505

Finally, the influence of the pH value on sorption is shown in figure 3. Determined Kd

506

values were almost constant for pyrazole, pyrrole, furan and oxazole, while Kd decreased

507

strongly with decreasing pH for imidazole (54%) and pyridine (79%). In case of pyridine,

508

decrease in Kd is sharp between pH 6 and 3. Speciation of the sorbates imidazole and pyridine

509

is changed due to the pH value, while for the other compounds speciation remained

510

unchanged (oxazole, pyrrole, furan) or was only changed very slightly (pyrazole). While at

511

pH 3 more than 99% of imidazole and pyridine are protonated and positively charged, only

512

15% in case of pyridine and 92% in case of imidazole are positively charged at pH 6. At pH 9

513

only 1% or less of both compounds is present with a positive charge. Besides the speciation of

514

the sorbates also the speciation of the sorbent is changed. At pH 3 the surface is positively

515

charged, while at pH 6 and 9 the surface is negatively charged.48 Consequently, sorption of

516

imidazole and pyridine may be reduced at pH 3 due to electrostatic repulsion occurring

517

between the positively charged sorbent surface and the equally charged sorbate molecules.

518

Contribution of electrostatic repulsion could be confirmed due to measurements at increased

519

ionic strength at pH 3. At 100 mM CaCl2 sorption of pyridine and imidazole increased

520

compared to 10 mM CaCl2 by 90% and 50%, respectively (figure S12 in the supporting

521

information). Increased ionic strength results in reduction of the electrostatic repulsion of

522

equal charges by shielding those with oppositely charged ions from the liquid matrix. 100

pH 3 pH 6 pH 9

Kd [L/kg]

80 60 40 20

e py rr ol

in e py rid

le py ra zo

le

le id az o

ox az o

523

im

fu r

an

0

18 ACS Paragon Plus Environment

Environmental Science & Technology

524 525 526

Figure 3: Determined Kd values of the different analytes for three different pH values of the eluent (3, 6 and 9). Error bars indicate the maximum error estimated from standard deviation of retention times of sorbate and tracer via error propagation (n=3).

527

Sorption isotherms derived from measurements showed that Kf of pyridine decreased

528

strongly with decreasing pH as it was observed for Kd (section SI.10 in the supporting

529

information). In case of imidazole the change was less pronounced but still visible. For the

530

other compounds, no clear trend could be observed regarding Kf and sorption linearity

531

expressed by n as it was also found for Kd.

532

Decreasing sorption to CNTs as consequence of electrostatic repulsion was also

533

described in literature. In case of different anilines sorption decreased with decreasing pH due

534

to electrostatic repulsion.17 As an example, the pKa-value of 4-chloroaniline is 4.15 indicating

535

an increasing amount of the positively charged species with decreasing pH. Sorption reduced

536

strongly when pH decreased below 6 when the fraction of the positively charged species

537

started to increase. Similar observations were made for different phenol derivates10, 16, 17, 31, 57,

538

62, 70

539

sorbates were negatively charged, electrostatic repulsion occurred. No influence of pH on

540

sorption for compounds of which speciation is not changed with changing pH was also

541

reported in literature for naphthalene or 1,2-dichlorobenzene,10 phenanthrene,16 atrazine58, 60

542

and m-xylene, toluene or ethylbenzene.23 This fits well to the results presented in this study.

, when pH was increased above the pKa of the respective compound. As sorbents and

543

Overall, column chromatography was used successfully for the first time to study

544

sorption of heterocyclic organic compounds to MWCNTs and the influence of varying

545

environmental conditions on their sorption. A fortitude of the column method is the fast and

546

simple investigation of the influence of environmental conditions on sorption while

547

concomitantly reducing the sorbent demand strongly. Trends and observations made with

548

classical batch experiments regarding the influence of temperature, pH and ionic strength on

549

sorption could also be observed using the column method presented here so that this method

550

represents a promising alternative also for sorption studies regarding CNMs with the aim to

551

investigate the influence of environmental conditions on sorption. This method is applicable

552

for all UV-active organic compounds. Coupling to other detection techniques (e.g. mass

553

spectrometry (MS)) would also allow the analysis of compounds with insufficient UV

554

absorption. Depending on the type of compound, dilution of MWCNTs may have to be

555

increased or decreased to be able to get sufficient but also not too strong retention inside the

556

column. The applicability of the method for other types of carbon-based sorbents is currently 19 ACS Paragon Plus Environment

Page 20 of 27

Page 21 of 27

Environmental Science & Technology

557

investigated to compare sorption properties of these materials under similar conditions using

558

column experiments.

559 560

Supporting Information

561

Additional information about details of chemicals used in this study, column characteristics,

562

equilibrium and isotherm concentrations, UV-calibrations, solute descriptors for ppLFER and

563

details on isotherms derived at different environmental conditions are added.

564 565 566

Acknowledgements

567

This project was supported by the German Research Foundation (SCHM 1372/10-2) and the

568

Professor Werdelmann Foundation.

569 570

References

571

1.

572

walled carbon nanotubes as sorptive material for solventless in-tube microextraction (ITEX2)

573

- a factorial design study. Anal. Bioanal. Chem. 2013, 405, (26), 8387-8395.

574

2.

575

multiwalled carbon nanotubes as solid phase extraction adsorbent. J. Hazard. Mater. 2009,

576

169, (1-3), 466-471.

577

3.

578

Application of multi-walled carbon nanotubes as substrate for the on-line preconcentration,

579

speciation and determination of vanadium by ETAAS. J. Anal. At. Spectrom. 2007, 22, (10),

580

1290-1295.

581

4.

582

in environmental water samples by solid-phase extraction using multi-walled carbon

583

nanotubes as adsorbent coupled with high-performance liquid chromatography. J.

584

Chromatogr. Sci. 2015, 53, (2), 380-384.

Hüffer, T.; Osorio, X. L.; Jochmann, M. A.; Schilling, B.; Schmidt, T. C. Multi-

Duran, A.; Tuzen, M.; Soylak, M. Preconcentration of some trace elements via using

Gil, R. A.; Goyanes, S. N.; Polla, G.; Smichowski, P.; Olsina, R. A.; Martinez, L. D.

Ma, J. P.; Lu, X.; Xia, Y.; Yan, F. L. Determination of pyrazole and pyrrole pesticides

20 ACS Paragon Plus Environment

Environmental Science & Technology

585

5.

Li, Q. L.; Yuan, D. X.; Lin, Q. M. Evaluation of multi-walled carbon nanotubes as an

586

adsorbent for trapping volatile organic compounds from environmental samples. J.

587

Chromatogr. A 2004, 1026, (1-2), 283-288.

588

6.

589

adsorption of PAHs to carbon nanotubes over a six order of magnitude wide concentration

590

range. Environ. Sci. Technol. 2011, 45, (14), 6011-6017.

591

7.

592

to carbon nanotubes. Environ. Sci. Technol. 2007, 41, (24), 8295-8300.

593

8.

594

ionizable organic contaminants on multi-walled carbon nanotubes with different oxygen

595

contents. J. Hazard. Mater. 2011, 186, (1), 407-415.

596

9.

597

single-walled carbon nanotubes: Effects of aqueous solution chemistry. Environ. Sci. Technol.

598

2008, 42, (19), 7225-7230.

599

10.

600

substituted aromatics to carbon nanotubes. Environ. Sci. Technol. 2008, 42, (18), 6862-6868.

601

11.

602

Geissler, E.; Laszlo, K. Interaction of phenol and dopamine with commercial MWCNTs. J.

603

Colloid Interface Sci. 2011, 364, (2), 469-475.

604

12.

605

and 3-chlorophenol on purified MWCNTs. J. Colloid Interface Sci. 2012, 387, 244-249.

606

13.

607

Adsorption of 1,2-dichlorobenzene from water to carbon nanotubes. Chem. Phys. Lett. 2003,

608

376, (1-2), 154-158.

609

14.

610

carbon nanotubes: Determination of equilibrium, kinetics and thermodynamics parameters. J.

611

Colloid Interface Sci. 2008, 327, (2), 308-315.

612

15.

Sheng, G. D.; Shao, D. D.; Ren, X. M.; Wang, X. Q.; Li, J. X.; Chen, Y. X.; Wang, X.

613

K.

Kinetics and thermodynamics of adsorption of ionizable aromatic compounds from

614

aqueous solutions by as-prepared and oxidized multiwalled carbon nanotubes. J. Hazard.

615

Mater. 2010, 178, (1-3), 505-516.

616

16.

617

organic chemicals by carbon nanotubes: Effects of background solution chemistry. Water Res.

618

2010, 44, (6), 2067-2074.

Kah, M.; Zhang, X. R.; Jonker, M. T. O.; Hofmann, T. Measuring and modelling

Chen, W.; Duan, L.; Zhu, D. Q. Adsorption of polar and nonpolar organic chemicals

Li, X. N.; Zhao, H. M.; Quan, X.; Chen, S. O.; Zhang, Y. B.; Yu, H. T. Adsorption of

Chen, J. Y.; Chen, W.; Zhu, D. Adsorption of nonionic aromatic compounds to

Chen, W.; Duan, L.; Wang, L. L.; Zhu, D. Q. Adsorption of hydroxyl- and amino-

Tóth, A.; Torocsik, A.; Tombacz, E.; Olah, E.; Heggen, M.; Li, C. L.; Klumpp, E.;

Tóth, A.; Torocsik, A.; Tombacz, E.; Laszlo, K. Competitive adsorption of phenol

Peng, X. J.; Li, Y. H.; Luan, Z. K.; Di, Z. C.; Wang, H. Y.; Tian, B. H.; Jia, Z. P.

Kuo, C. Y.; Wu, C. H.; Wu, J. Y. Adsorption of direct dyes from aqueous solutions by

Zhang, S. J.; Shao, T.; Bekaroglu, S. S. K.; Karanfil, T. Adsorption of synthetic

21 ACS Paragon Plus Environment

Page 22 of 27

Page 23 of 27

Environmental Science & Technology

619

17.

Yang, K.; Wu, W. H.; Jing, Q. F.; Zhu, L. Z. Aqueous adsorption of aniline, phenol,

620

and their substitutes by multi-walled carbon nanotubes. Environ. Sci. Technol. 2008, 42, (21),

621

7931-7936.

622

18.

623

nanotubes in aqueous solution. Carbon 2012, 50, (6), 2209-2219.

624

19.

625

nano-structures: copper oxide and carbon tubes. Clean Technol. Environ. Policy 2016, 18, (3),

626

817-827.

627

20.

628

Adsorption of phosphate by zirconia functionalized multi-walled carbon nanotubes. Chin. J.

629

Inorg. Chem. 2013, 29, (5), 965-972.

630

21.

631

using multi-walled carbon nanotubes. J. Fluorine Chem. 2011, 132, (8), 516-520.

632

22.

633

mercury from water by modified multi-walled carbon nanotubes: adsorption behaviour and

634

interference resistance by coexisting anions. Environ. Technol. 2014, 35, (15), 1935-1944.

635

23.

636

multi-walled carbon nanotubes with different oxygen contents from aqueous solutions. J.

637

Hazard. Mater. 2011, 192, (3), 1370-1379.

638

24.

639

to reference soils: Column studies for process identification. Environ. Sci. Technol. 2006, 40,

640

(19), 5962-5970.

641

25.

642

carbon nanotubes: Effect of NOM characteristics and water quality parameters. Environ. Sci.

643

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

644

26.

645

multiwalled carbon nanotubes in fresh surface water samples. Environ. Pollut. 2010, 158, (5),

646

1270-1274.

647

27.

648

Gogos, A.; Hanna, S.; Kagi, R.; Mansfield, E.; Bustos, A. R. M.; Plata, D. L.; Reipa, V.;

649

Westerhoff, P.; Winchester, M. R. Quantification of Carbon Nanotubes in Environmental

650

Matrices: Current Capabilities, Case Studies, and Future Prospects. Environ. Sci. Technol.

651

2016, 50, (9), 4587-4605.

Fang, Q. L.; Chen, B. L. Adsorption of perchlorate onto raw and oxidized carbon

Mahdavi, S.; Akhzari, D. The removal of phosphate from aqueous solutions using two

Zong, E. M.; Wei, D.; Huan, Z. K.; Peng, D.; Wan, H. Q.; Zheng, S. R.; Xu, Z. Y.

Ansari, M.; Kazemipour, M.; Dehghani, M. The defluoridation of drinking water

Chen, P. H.; Hsu, C. F.; Tsai, D. D. W.; Lu, Y. M.; Huang, W. J. Adsorption of

Yu, F.; Ma, J.; Wu, Y. Q. Adsorption of toluene, ethylbenzene and m-xylene on

Bi, E.; Schmidt, T. C.; Haderlein, S. B. Sorption of heterocyclic organic compounds

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

Lin, D. H.; Liu, N.; Yang, K.; Xing, B. S.; Wu, F. C.

Different stabilities of

Petersen, E. J.; Flores-Cervantes, D. X.; Bucheli, T. D.; Elliott, L. C. C.; Fagan, J. A.;

22 ACS Paragon Plus Environment

Environmental Science & Technology

652

28.

Hüffer, T.; Endo, S.; Metzelder, F.; Schroth, S.; Schmidt, T. C. Prediction of sorption

653

of aromatic and aliphatic organic compounds by carbon nanotubes using poly-parameter

654

linear free-energy relationships. Water Res. 2014, 59, 295-303.

655

29.

656

development for adsorption of aromatic contaminants by multi-walled carbon nanotubes.

657

Environ. Sci. Technol. 2013, 47, (5), 2295-2303.

658

30.

659

from aqueous solution by multi-walled carbon nanotubes. J. Colloid Interface Sci. 2012, 377,

660

355-361.

661

31.

662

aqueous solutions by pristine and surface functionalized single-walled carbon nanotubes.

663

Journal of Environmental Sciences 2016, 43, 187-198.

664

32.

665

Adsorption kinetics, isotherms and thermodynamics of atrazine on surface oxidized

666

multiwalled carbon nanotubes. J. Hazard. Mater. 2009, 169, (1-3), 912-918.

667

33.

668

of trihalomethanes on multiwalled carbon nanotubes. J. Hazard. Mater. 2006, 138, (2), 304-

669

310.

670

34.

671

nanotubes. Environ. Sci. Technol. 2008, 42, (24), 9005-9013.

672

35.

673

Chemosphere 2003, 53, (5), 515-522.

674

36.

675

Affinity of Organic Cations for Soil Organic Matter. Environ. Sci. Technol. 2012, 46, (11),

676

5894-5901.

677

37.

678

Experimentally determined soil organic matter-water sorption coefficients for different classes

679

of natural toxins and comparison with estimated numbers. Environ. Sci. Technol. 2012, 46,

680

(11), 6118-6126.

681

38.

682

dimethylphthalate - Compatibility of batch and column techniques. Soil Sci. Soc. Am. J. 2001,

683

65, (1), 102-111.

Apul, O. G.; Wang, Q. L.; Shao, T.; Rieck, J. R.; Karanfil, T. Predictive model

Hu, J. L.; Tong, Z. L.; Hu, Z. H.; Chen, G. W.; Chen, T. H. Adsorption of roxarsone

Ding, H.; Li, X.; Wang, J.; Zhang, X. J.; Chen, C. Adsorption of chlorophenols from

Chen, G. C.; Shan, X. Q.; Zhou, Y. Q.; Shen, X. E.; Huang, H. L.; Khan, S. U.

Lu, C. S.; Chung, Y. L.; Chang, K. F. Adsorption thermodynamic and kinetic studies

Pan, B.; Xing, B. S.

Adsorption mechanisms of organic chemicals on carbon

Bucheli, T. D.; Gustafsson, O. Soot sorption of non-ortho and ortho substituted PCBs.

Droge, S.; Goss, K. U. Effect of Sodium and Calcium Cations on the Ion-Exchange

Schenzel, J.; Goss, K. U.; Schwarzenbach, R. P.; Bucheli, T. D.; Droge, S. T. J.

Altfelder, S.; Streck, T.; Maraqa, M. A.; Voice, T. C. Nonequilibrium sorption of

23 ACS Paragon Plus Environment

Page 24 of 27

Page 25 of 27

Environmental Science & Technology

684

39.

Li, B. H.; Qian, Y. A.; Bi, E. P.; Chen, H. H.; Schmidt, T. C. Sorption behavior of

685

phthalic acid esters on reference soils evaluated by soil column chromatography. Clean-Soil

686

Air Water 2010, 38, (5-6), 425-429.

687

40.

688

heterocyclic aromatic compounds to soil. Environ. Sci. Technol. 2007, 41, (9), 3172-3178.

689

41.

690

steroid hormones in soils during transport: Column studies and model evaluation. Environ.

691

Sci. Technol. 2004, 38, (5), 1460-1470.

692

42.

693

sorption and nonequilibrium solute transport in aggregated porous media: Experiments,

694

process identification and modeling. J. Contam. Hydrol. 1998, 31, (3-4), 373-407.

695

43.

696

neutral and ionized chlorophenols. Environ. Sci. Technol. 1991, 25, (4), 722-729.

697

44.

698

acidic organic chemicals to soil organic matter. Environ. Sci. Technol. 2009, 43, (24), 9189-

699

9195.

700

45.

701

obtain organic cation sorption isotherms. Environ. Sci. Technol. 2016, 50, (15), 8196-8204.

702

46.

703

aqueous solutions. Ph.D. disserstation, University Duisburg-Essen, Essen, 2010.

704

47.

705

organic chemicals to mineral surfaces. Environ. Sci. Technol. 1997, 31, (4), 1079-1086.

706

48.

707

inorganic anions to wultiwalled carbon nanotubes studied by column chromatography.

708

Environ. Sci. Technol. 2017, 51, (9), 4928-4935.

709

49.

710

groups on desorption of organic compounds from carbon nanotubes into water: Insight into

711

desorption hysteresis. Environ. Sci. Technol. 2013, 47, (15), 8373-8382.

712

50.

713

carbon nanotubes: pH dependence and thermodynamics. Chemosphere 2014, 95, 150-155.

714

51.

715

thermodynamics on surface-modified carbon nanotubes. Environ. Sci. Technol. 2010, 44, (3),

716

978-984.

Bi, E.; Schmidt, T. C.; Haderlein, S. B. Environmental factors influencing sorption of

Das, B. S.; Lee, L. S.; Rao, P. S. C.; Hultgren, R. P. Sorption and degradation of

Fesch, C.; Simon, W.; Haderlein, S. B.; Reichert, P.; Schwarzenbach, R. P. Nonlinear

Lee, L. S.; Rao, P. S. C.; Brusseau, M. L. Nonequilibrium sorption and transport of

Tülp, H. C.; Fenner, K.; Schwarzenbach, R. P.; Goss, K. U. pH-dependent sorption of

Jolin, W. C.; Sullivan, J.; Vasudevan, D.; MacKay, A. K. Column chromatography to

Qian, Y. Sorption of hydrophobic organic compounds (HOCs) to inorganic surfaces in

Mader, B. T.; Goss, K. U.; Eisenreich, S. J. Sorption of nonionic, hydrophobic

Metzelder, F.; Schmidt, T. C.

Environmental conditions influencing sorption of

Wu, W. H.; Jiang, W.; Zhang, W. D.; Lin, D. H.; Yang, K. Influence of functional

Li, H. B.; Zhang, D.; Han, X. Z.; Xing, B. S. Adsorption of antibiotic ciprofloxacin on

Wang, Z. Y.; Yu, X. D.; Pan, B.; Xing, B. S.

24 ACS Paragon Plus Environment

Norfloxacin sorption and its

Environmental Science & Technology

Lu, C. Y.; Chiu, H. S.

Page 26 of 27

717

52.

Adsorption of zinc(II) from water with purified carbon

718

nanotubes. Chem. Eng. Sci. 2006, 61, (4), 1138-1145.

719

53.

720

chromatography for sorption parameter determination. Chemosphere 2010, 80, (7), 787-793.

721

54.

722

degradation of quinoline in unsaturated soil. Environ. Sci. Technol. 1999, 33, (17), 2891-

723

2898.

724

55.

725

adsorption-isotherms from column experiments - an alternative to batch studies. Environ. Sci.

726

Technol. 1993, 27, (5), 943-948.

727

56.

728

antibiotics to multiwalled carbon nanotubes. Langmuir 2009, 25, (19), 11608-11613.

729

57.

730

of aromaticity and substitution of hydroxyl groups. Environ. Sci. Technol. 2008, 42, (19),

731

7254-7259.

732

58.

733

chemicals by carbon nanotubes and activated carbons. Nanotechnology 2012, 23, (29), 13.

734

59.

735

Medina, V. M. O.; Alcala-Jauegui, J. A. Adsorption of organic compounds onto multiwall

736

and nitrogen-doped carbon nanotubes: Insights into the adsorption mechanisms. Water Air

737

and Soil Pollution 2017, 228, (4), 13.

738

60.

739

and desorption of atrazine on carbon nanotubes. J. Colloid Interface Sci. 2008, 321, (1), 30-

740

38.

741

61.

742

thermodynamics of sorption of nitroaromatic compounds to as-grown and oxidized

743

multiwalled carbon nanotubes. J. Colloid Interface Sci. 2009, 330, (1), 1-8.

744

62.

745

aqueous solutions by carbon nanotube hybrid polymer adsorbents. Chemical Engineering

746

Research & Design 2017, 123, 76-83.

747

63.

748

multi-walled carbon nanotubes for the removal of olaquindox from aqueous solutions. J.

749

Hazard. Mater. 2011, 197, 389-396.

Bi, E. P.; Schmidt, T. C.; Haderlein, S. B. Practical issues relating to soil column

Thomsen, A. B.; Henriksen, K.; Gron, C.; Moldrup, P.

Sorption, transport, and

Bürgisser, C. S.; Cernik, M.; Borkovec, M.; Sticher, H. Determination of nonlinear

Ji, L. L.; Chen, W.; Zheng, S. R.; Xu, Z. Y.; Zhu, D. Q. Adsorption of sulfonamide

Lin, D. H.; Xing, B. S. Adsorption of phenolic compounds by carbon nanotubes: Role

Brooks, A. J.; Lim, H. N.; Kilduff, J. E. Adsorption uptake of synthetic organic

Diaz-Flores, P. E.; Arcibar-Orozco, J. A.; Perez-Aguilar, N. V.; Rangel-Mendez, J. R.;

Yan, X. M.; Shi, B. Y.; Lu, J. J.; Feng, C. H.; Wang, D. S.; Tang, H. X. Adsorption

Shen, X. E.; Shan, X. Q.; Dong, D. M.; Hua, X. Y.; Owens, G.

Kinetics and

Xu, L. H.; Wang, Z. M.; Ye, S. A.; Sui, X. Y. Removal of p-chlorophenol from

Zhang, L.; Xu, T. C.; Liu, X. Y.; Zhang, Y. Y.; Jin, H. J. Adsorption behavior of

25 ACS Paragon Plus Environment

Page 27 of 27

Environmental Science & Technology

750

64.

Zhu, X.; Wang, J. L. Comparison of sulfamethazine adsorption by multi-walled

751

carbon nanotubes (MWCNTs) and magnetic MWCNTs from aqueous solution. Fresenius

752

Environmental Bulletin 2017, 26, (1A), 579-589.

753

65.

754

carbon nanotubes: Effects of solution chemistry. Environ. Pollut. 2016, 213, 90-98.

755

66.

756

sorbent/water systems and sorption coefficients for selected pollutants. A review. J. Phys.

757

Chem. Ref. Data 2001, 30, (1), 187-439.

758

67.

759

gas chromatography to determine thermodynamic parameters of aroma-starch interactions. J.

760

Chromatogr. A 2002, 969, (1-2), 9-16.

761

68.

762

soils - Processes, measurements and experience with the applicability of the modified OECD-

763

guideline 106. Chemosphere 1991, 22, (3-4), 285-304.

764

69.

Li, Y. H.; Ding, J.; Luan, Z. K.; Di, Z. C.; Zhu, Y. F.; Xu, C. L.; Wu, D. H.; Wei, B.

765

Q.

Competitive adsorption of Pb2+, Cu2+ and Cd2+ ions from aqueous solutions by

766

multiwalled carbon nanotubes. Carbon 2003, 41, (14), 2787-2792.

767

70.

768

Z.; Dalmacija, B. Adsorption of chlorinated phenols on multiwalled carbon nanotubes. Rsc

769

Advances 2015, 5, (32), 24920-24929.

Engel, M.; Chefetz, B. Adsorption and desorption of dissolved organic matter by

Delle Site, A.

Factors affecting sorption of organic compounds in natural

Boutboul, A.; Lenfant, F.; Giampaoli, P.; Feigenbaum, A.; Ducruet, V. Use of inverse

von Oepen, B.; Kordel, W.; Klein, W. Sorption of nonpolar and polar compounds to

Kragulj, M.; Trickovic, J.; Kukovecz, A.; Jovic, B.; Molnar, J.; Roncevic, S.; Konya,

770

26 ACS Paragon Plus Environment