Subscriber access provided by CMU Libraries - http://library.cmich.edu
Article
Interaction Between Graphene Oxide Nanoparticles and Quartz Sand Nikolaos P. Sotirelis, and Constantinos V. Chrysikopoulos Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b03496 • Publication Date (Web): 14 Oct 2015 Downloaded from http://pubs.acs.org on October 15, 2015
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 29
Environmental Science & Technology
1 2
Interaction Between Graphene Oxide Nanoparticles and
3
Quartz Sand
4
5 6
by
7
Nikolaos P. Sotirelis, and Constantinos V. Chrysikopoulos*
8 9 10 11 12
School of Environmental Engineering, Technical University of Crete, 73100 Chania, Greece.
13 14 15 16 17 18
Manuscript submitted to the
19
Environmental Science & Technology
20 21 22 23 24
Key words: Nanoparticles, graphene oxide, deposition, attachment, quartz sand, DLVO.
25 26
October 12, 2015
27 28 29 30 31 32 33
*Corresponding author (
[email protected], Fax: +30 28210 37847).
1
ACS Paragon Plus Environment
Environmental Science & Technology
Page 2 of 29
34
ABSTRACT: In this study, the influence of pH, ionic strength (IS), and
35
temperature on graphene oxide (GO) nanoparticles attachment onto quartz
36
sand were investigated. Batch experiments were conducted at three
37
controlled temperatures (4, 12, and 25°C) in solutions with different pH values
38
(pH=4, 7, and 10), and ionic strengths (IS=1.4, 6.4, and 21.4 mM), under static
39
and dynamic conditions. The surface properties of GO nanoparticles and
40
quartz sand were evaluated by electrophoretic mobility measurements.
41
Derjaguin-Landau-Verwey-Overbeek (DLVO) potential energy profiles were
42
constructed for the experimental conditions, using measured zeta potentials.
43
The experimental results showed that GO nanoparticles were very stable
44
under the experimental conditions. Both temperature and pH did not play a
45
significant role in the attachment of GO nanoparticles onto quartz sand. In
46
contrast, IS was shown to influence attachment. The attachment of GO
47
particles onto quartz sand increased significantly with increasing IS. The
48
experimental data were fitted nicely with a Freundlich isotherm, and the
49
attachment kinetics were satisfactorily described with a pseudo-second-order
50
model, which implies that the quartz sand exhibited substantial surface
51
heterogeneity and that GO retention was governed by chemisorption.
52
Furthermore, thermodynamic analysis revealed that the attachment process
53
was non-spontaneous and endothermic, which may be associated with
54
structural changes of the sand surfaces due to chemisorption. Therefore,
55
secondary minimum interaction may not be the dominant mechanism for GO
56
attachment onto the quartz sand under the experimental conditions.
57 58
INTRODUCTION
59
Graphene was first produced in 2004, and is a single atom-thick sheet of
60
carbon packed in a hexagonal lattice.1 For this pioneering discovery, Andre K.
61
Geim and Konstantin S. Novoselov, were awarded the Nobel prize in Physics
62
in 2010. This award acknowledged that graphene, owing to its remarkable
63
physicochemical properties (e.g. high specific surface area, high electron
64
mobility, excellent mechanical strength, excellent structural transformability,
65
and high thermal conductivity) is one of the fastest growing nanomaterials that 2
ACS Paragon Plus Environment
Page 3 of 29
66
Environmental Science & Technology
has changed many fields of science, engineering, and industry.1−4
67 68
Graphene oxide (GO) is a layered nano-material that contains graphene
69
sheets and oxygen-bearing functional groups.5 GO is one of the most
70
important graphene derivatives.6 Due to its excellent electrochemical
71
properties, GO is used in a wide range of applications. Consequently, the
72
production growth of this material is very rapid, and large quantities of GO
73
particles are expected to eventually reach sensitive environmental systems,
74
including subsurface formations.7 Note that the solubility of GO in water is
75
very high.8 Also, the presence of mobile GO particles in surface waters, and
76
groundwater may affect the fate and simultaneous transport (cotransport) of
77
dissolved species and suspended particles (e.g. colloids, biocolloids).9−11
78
Furthermore, various studies have reported that GO may be toxic to a variety
79
of mammalian organisms, as well as to human and bacterial cells.12−17 For this
80
reason, it is important to fully understand the fate of GO in environmental
81
systems, and particularly in groundwater systems, because GO transport in
82
porous media is significantly affected by the interaction between GO and solid
83
matrix (e.g. sand). Numerous studies published in the literature have focused
84
on the transport of GO in porous media.18−21 The results of these studies have
85
shown that GO has great colloidal stability and the main factor that affects GO
86
retention in porous media is the ionic strength (IS).
87
The objective of this work was to improve our understanding of the
88
mechanisms responsible for deposition and attachment of GO nanoparticles
89
onto quartz sand. Batch experiments were conducted, to investigate the effect
90
of temperature, pH, and IS on GO attachment onto quartz sand, and GO
91
aggregation, under static and dynamic conditions. The attachment kinetics,
92
related attachment isotherms of GO interactions with quartz sand under
93
different temperatures, and the corresponding thermodynamics were
94
examined. Also, the aggregation of GO and the attachment behavior of GO
95
onto quartz sand were related to theoretically determined DLVO energy
96
interaction profiles. To our knowledge, no previous study has investigated the
97
effect of temperature and pH on GO attachment onto quartz sand. 3
ACS Paragon Plus Environment
Environmental Science & Technology
Page 4 of 29
98 99
MATERIALS AND METHODS
100
Graphene oxide. High purity SP-1 graphite powder (Bay Carbon Inc,
101
Bay City, MI) was used to produce graphite oxide based on the procedures
102
reported by Hummers et al.22 The graphite oxide was exfoliated by sonication
103
and centrifugation to obtain graphene oxide flakes.23 The GO suspensions
104
were prepared by mixing 3 mg of graphene oxide flakes with 250 mL of a
105
phosphate buffered solution (PBS) with low ionic strength (IS=1.4 mM). A
106
transmission electron microscopy (TEM) system, JEOL (JEM-2100, operated
107
at 200 kV), equiped with an Erlangshen CCD camera (Model 782 ES500W),
108
was used to record images of GO nanoparticles. The morphology of
109
representative GO flakes is shown in Figure SI1. The suspensions with
110
different ionic strengths were adjusted with NaCl; whereas, the suspensions
111
with different pH values were adjusted with either H2PO4 or NaOH.
112
Subsequently, the suspensions were sonicated (Elmasonic S 30/(H), Elma
113
Schmidbauer GmbH, Singen, Germany) for 2 h to ensure that the dispersion
114
is thoroughly uniform. It was confirmed that prolonged sonication did not to
115
have negative effects (e.g. re-aggregation) on GO particles (see Figure SI2).
116
All of the solutions were prepared with distilled deionized water (ddH2O).
117
Furthermore, all chemicals employed in this study were of analytical reagent
118
grade, employed without any additional purification.
119
Calibration curves were prepared, for each set of solution chemistry (pH
120
and IS) examined in this study, in order to establish the relationship between
121
absorbance, Abs [-], and GO concentration, CGO [M/L3], in the range 0 to 30
122
mg/L. A series of diluted samples were prepared from an aqueous solution
123
with known GO concentration, and the absorbance of each diluted sample
124
was measured at the wavelength of λmax=231 nm with a UV-Visible
125
spectrophotometer (Cary 400 BIO, Varian, Palo Alto, California). Also, a
126
zetasizer (Nano ZS90, Malvern Instruments, Southborough, MA) was used to
127
measure the zeta potential and hydrodynamic diameter of GO nanoparticles
128
under the various experimental conditions of this study at 25 °C. All zeta
129
potential and hydrodynamic diameter measurements were obtained in 4
ACS Paragon Plus Environment
Page 5 of 29
Environmental Science & Technology
130
triplicates. The sizes of GO nanoparticles under the various experimental
131
conditions of this study are presented in Table SI1.
132 133
Sand. Quartz sand with grain diameter ranging from 0.425 to 0.600 mm
134
(sieve no. 40) was used in this study for the GO attachment experiments.
135
Following the procedures reported by Chrysikopoulos and Aravantinou,24 the
136
particle-size distribution value determined by sieve analysis was used to
137
calculate the coefficient of uniformity, Cu=d60/d10=1.21 (were d10, and d60 is the
138
diameter of a sand grain that is barely too large to pass through a sieve that
139
allows 10%, and 60%, respectively, of the material (by weight) to pass
140
through). The quartz sand employed in this study was relatively uniform
141
because the smaller the value of Cu the more uniform the sand. Note that
142
Cu=1 corresponds to uniform sand.25 The chemical composition of the quartz
143
sand as reported by the manufacturer (Filcom, Netherlands) was: 96.2% SiO2,
144
0.15% Na2O, 0.11% CaO, 0.02% MgO, 1.75% Al2O3, 0.78% K2O, 0.06% SO3,
145
0.46% Fe2O3, 0.03% P2O5, 0.02% BaO, 0.01% Mn3O4, and 0.28% loss on
146
ignition. The quartz sand was thoroughly cleaned with 0.1 M nitric acid HNO3
147
(70%) for 3 h to remove surface impurities (e.g. metal hydroxides and organic
148
coatings), rinsed with distilled deionized water (ddH2O), then soaked in 0.1 M
149
NaOH for 3 h, and rinsed with ddH2O again.26,27 Finally, the quartz sand was
150
dried in an oven at 80 °C. The sand grains were relatively large for direct zeta
151
potential measurement by a zetasizer. Consequently, the sand grains were
152
crushed into fine powder. Several stable suspensions were formed with the
153
crushed sand, each having a desired ionic strength, which were used for zeta
154
potential measurements.28,29
155 156
Batch experiments. Both static and dynamic batch experiments were
157
conducted under various solution chemistry conditions at 4, 12 and 25 °C, in
158
order to examine the effect of pH, IS, and temperature on GO aggregation and
159
GO attachment on quartz sand. All batch experiments were performed in 20
160
mL Pyrex glass screw-cap tubes (Fisher Scientific). Glass tubes were washed
161
with detergent, rinsed thoroughly in ddH2O, autoclave sterilized, and oven 5
ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 29
162
dried at 80°C overnight. A PBS solution with IS=1.35 mM was prepared with
163
0.001 M phosphate buffer salts in ddH2O and adjusted to a pH 7.2 with NaOH.
164
The PBS solution was used to stabilize the pH of GO dispersion.5 Note that
165
the relatively wide temperature range (4 to 25°C), and pH range (4 to 10)
166
considered in this study is representative of the conditions observed in various
167
ground and surface waters.30
168
For each experiment, 16 glass tubes were employed, which were divided
169
into two groups. Each group consisted of 8 glass tubes. The glass tubes of the
170
first group (experimental tubes) contained 14 mL of GO dispersion with 14 g
171
of sand, and the glass tubes of the second group (control tubes) contained 20
172
mL of GO dispersion without sand. All glass tubes were filled to the top.
173
However, a small air bubble was always trapped within the tubes when the
174
caps were screwed onto the tubes. Both groups were treated in the same
175
manner. The experiments at 4 and 12°C were conducted in an incubator (Foc
176
120E, Velp Scientifica, Italy). The dynamic batch experiments were performed
177
with the tubes attached to a rotator (Selecta, Agitador orbit), operated at 12
178
rpm, in order to allow the sand to mix within the GO suspension. Control
179
tubes, in the absence of sand, were used to monitor GO aggregation and
180
possible GO attachment onto the walls of the glass tubes. A sample of the
181
PBS solution (3.0 mL) was removed from each selected glass tube at different
182
preselected times (0, 5, 10, 20, 30, 60, 90, 120, 180, 240 minutes) and the
183
GO concentration was measured in triplicates. All used glass tubes were
184
discarded. The experimental conditions of the various batch experiments
185
conducted in this study are summarized in Table SI2.
186 187
THEORETICAL ANALYSIS OF GO-SAND INTERACTIONS
188
Assuming that GO particles are practically colloids, the classical theory
189
developed by Derjaguin-Landau-Verwey-Overbeek (DLVO) is applicable.
190
Based on the DLVO theory, the total interaction energy between two surfaces
191
(here GO and quartz sand) equals the arithmetic sum of the van der Waals
192
ΦvdW, double layer, Φdl, and Born, ΦBorn, potential energies:26
6
ACS Paragon Plus Environment
Page 7 of 29
Environmental Science & Technology
193
ΦDLVO (h) = Φ vdW (h) + Φdl (h) + ΦBorn (h)
194
where h [m] is the separation distance between the approaching surfaces.
(1)
195
Following the work by Chrysikopoulos and Syngouna,31 for the case of
196
two approaching surfaces, the ΦvdW [J] interactions were calculated with the
197
expression provided by Gregory,32 using λ≈10-7 m for the characteristic
198
wavelength of the sphere-plate or plate-plate interactions, and A123=6.26×10-
199
21
200
sand.18,19 The Φdl for sphere-plate interactions were calculated with the
201
expression provided by Hogg et al.,33 using NA=6.02×1023 [1/mol] for the
202
Avogadro’s number, e=1.602×10−19 [C] for the elementary charge, and
203
kB=1.38×10−23 [J/K] for the Boltzmann constant. The ΦBorn [J] for sphere-plate
204
was estimated by the relationship provided by Ruckenstein et al.34 Note that
205
ΦBorn can easily be neglected if h>1 nm. The effect of Born interaction may not
206
be of great significance in aqueous systems due to the possible presence of
207
hydrated ions, which prevent surface-surface separation distances to
208
approach h~0.3 nm.31
[J] for the combined Hamaker constant for the system GO-water-quartz
209
For the case of two approaching surfaces, both with planar geometries
210
(plate-plate), the ΦvdW [J], Φdl [J], and ΦBorn [J] interactions were calculated
211
with the equations provited by Gregory,32 Hogg et al.,33 and Mahmood et al.,35
212
respectively. All the details associated with the DLVO calculations are
213
presented in the Supporting Information section.
214 215
RESULTS AND DISCUSSION
216
Zeta potentials and DLVO interactions. Figure SI3 illustrates
217
graphically the measured zeta potential values of GO and quartz sand at
218
various pH values. Both GO and quartz sand are negatively charged for all the
219
pH values examined in this study (2 ≤ pH ≤ 10), suggesting that GO-sand and
220
GO-GO interactions are repulsive. Similar findings for GO-sand and GO-GO
221
interactions have been reported in the literature by several other
222
investigators.18−20 Zeta potential values can also be used to evaluate the
223
stability of suspended particles. Note that the zeta potentials of both GO and 7
ACS Paragon Plus Environment
Environmental Science & Technology
Page 8 of 29
224
quartz sand decrease with increasing pH, suggesting that the suspended
225
particles become more stable with increasing pH. It is worthy to note that the
226
stability of suspended particles is expected to increase with increasing
227
absolute zeta potential values.36 At relatively low absolute zeta potential
228
values aggregates may form, because existing attractive forces may be
229
stronger than the repulsive forces.37 The commonly used threshold for
230
absolute zeta potential value for stable colloidal suspensions is considered to
231
be >30 mV.38,39 Therefore, based on the data of Figure SI3 and preliminary
232
laboratory observations, particle suspensions with pH0.3 nm, rendering the
258
effect of Born interaction as insignificant. Consequently, the interaction energy
259
profiles constructed do not exhibit a Φmin₁.
260
The simulated GO-sand (sphere-plate) interactions energy profiles (see
261
Figures 1a,b) exhibit a shallow Φmin₂, and a very high Φmax₁. Note that the
262
lowest Φmax₁ and deepest Φmin₂ were observed for pH=4 (see Figure 1a). Note
263
that, Φmax₁ was smallest for the highest Is examined in this study (see Figure
264
3b), similar trend has been observed by other investigators.19 Also, as
265
expected, Φmin₂ was deepest for the highest Is. Furthermore, the simulated
266
GO-GO (plate-plate) interaction energy profiles (see Figures 1c,d) suggest
267
that there is a very shallow Φmin₂, and very high Φmax₁, suggesting the
268
presence of strong repulsive forces between GO nanoparticles. Clearly,
269
fluctuations in both pH and Is do not play a significant role in the GO-GO
270
interactions energy profiles. This is in agreement with previous findings that
271
for typical environmental systems (pH=5-9), GO aggregation is not
272
significantly affected by pH.19 It is worthy to note that Φmin₂ is very shallow for
273
all cases examined here (see Figure 1). Therefore, secondary minimum
274
interaction is an unlike mechanism for GO attachment onto the quartz sand
275
and GO aggregation. However, DLVO interaction energy profiles do not
276
account for surface roughness, angularity, chemical impurities, and surface
277
charge heterogeneity of the sand, which are known to produce local areas of
278
favorable interaction that lead to attachment even under unfavorable
279
conditions.43-45 Although in this study the quartz sand was thoroughly cleaned
280
to eliminate surface charge heterogeneity, charge heterogeneities from the
281
sand surfaces cannot be ruled out.46,47
282 283
Isotherm batch experiments. The experimental data from the GO
284
equilibrium attachment onto quartz sand at three different temperatures where
285
fitted nicely with a Freundlich isotherm (the linear and Langmuir isotherm
286
models were also tested, see Table SI3), which has been employed in various
287
contaminant, colloid and biocolloid attachment studies of environmental 9
ACS Paragon Plus Environment
Environmental Science & Technology
Page 10 of 29
288
significance.24,31,48,49 The Freundlich isotherm is a non-linear relationship
289
between the aqueous phase GO concentration at equilibrium, Ceq [Mn/L3], in
290
units of [mg GO/Liter of solution], and the GO concentration adsorbed onto
291
the quartz sand at equilibrium, C∗eq [Mn/Ms], in units of [mg GO/g sand]:50
292
Ceq = K f Ceq
293
where Kf [L3+m/Ms Mm−1 n ] is the Freundlich constant in units of [(Liter of
294
solution)m/(g sand)(mg GO)m-1], m [-] is the Freundlich exponent, which is
295
equal to one for linear deposition and attachment. Also, for notational
296
convenience Mn was introduced for the mass of nanoparticles (GO), and Ms
297
for the mass of solids (quartz sand). Note that the Freundlich isotherm
298
describes equilibrium attachment onto heterogeneous sorbent surfaces, and
299
contrary to Langmuir isotherm, does not assume monolayer deposition.24 The
300
Freundlich parameter m is a measure of the surface heterogeneity of the
301
quartz sand (the smaller the value of m the higher the surface heterogeneity
302
of the quartz sand), and Kf is directly proportional to the deposition and
303
attachment capacity of the quartz sand. For the estimation of the Freundlich
304
parameters the linearized form of the Freundlich isotherm was used:
305
log C∗eq = log K f + m log Ceq
306
The parameters m and logKf were estimated by the slope and ordinate
307
(vertical axis intercept), respectively, of the linear plot of the experimental data
308
in the form of log C∗eq versus logCeq. The static and dynamic attachment data
309
for the experiments conducted in this study at three different temperatures are
310
presented in Figure 2, and the corresponding Freundlich isotherm parameters
311
are listed in Table 1. The model fittings of the experimental data were
312
obtained with the graphical statistical software “IGOR-Pro” (WaveMetrics Inc.).
313
Based on the calculated R2 values (see Figure 2 and Table 1), the
314
Freundlich isotherm model fits very well the experimental data under static as
315
well as dynamic conditions. The number of accessible attachment sites is
316
much higher in dynamic than static experiments due to agitation, which
317
improves the contact of quartz sand grains with the liquid and decreases the
318
resistance to mass transfer.51–53 Therefore, more GO particles were adsorbed
∗
m
(2)
(3)
10
ACS Paragon Plus Environment
Page 11 of 29
Environmental Science & Technology
319
onto the quartz sand under dynamic than static conditions (compare the Kf
320
values listed in Table 1). This finding is in agreement with other attachment
321
studies.24,31 The attachment of GO nanoparticles onto quartz sand was a
322
favorable process because all of the estimated m values were less than unity
323
(m