Subscriber access provided by UNIV PRINCE EDWARD ISLAND
Article
Aggregation and Stability of Reduced Graphene Oxide: Complex Roles of Divalent Cations, pH, and Natural Organic Matter Indranil Chowdhury , Nikhita D. Mansukhani, Linda M. Guiney, Mark C Hersam, and Dermont C. Bouchard Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b01866 • Publication Date (Web): 17 Aug 2015 Downloaded from http://pubs.acs.org on August 23, 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 30
Environmental Science & Technology
Aggregation and Stability of Reduced Graphene Oxide: Complex Roles of Divalent Cations, pH, and Natural Organic Matter Indranil Chowdhury1*, Nikhita D. Mansukhani2, Linda M. Guiney2, Mark C. Hersam2, and Dermont Bouchard3*
1 2
Department of Civil and Environmental Engineering, Washington State University, Pullman, WA
Departments of Material Science and Engineering, Chemistry, and Medicine, Northwestern University, Evanston, IL 3
National Exposure Research Laboratory, Ecosystem Research Division, United States Environmental Protection Agency, Athens, GA
*Address correspondence to either author. Phone: 509-335-3721 (I.C.); 706-355-8333 (D.C.B.) E-mail:
[email protected] (I.C.);
[email protected] (D.C.B.)
Submitted to Environmental Science and Technology July 2015.
1 ACS Paragon Plus Environment
Environmental Science & Technology
TOC/ Abstract Art
2 ACS Paragon Plus Environment
Page 2 of 30
Page 3 of 30
1
Environmental Science & Technology
Abstract
2
The aggregation and stability of graphene oxide (GO) and three successively reduced GO
3
(rGO) nanomaterials were investigated. Reduced GO species were partially reduced GO (rGO-
4
1h), intermediately reduced GO (rGO-2h), and fully reduced GO (rGO-5h). Specifically,
5
influence of pH, ionic strength, ion valence and presence of natural organic matter (NOM) were
6
studied. Results show that stability of GO in water decreases with successive reduction of
7
functional groups, with pH having the greatest influence on rGO stability. Stability is also
8
dependent on ion valence and the concentration of surface functional groups. While pH did not
9
noticeably affect stability of GO in the presence of 10 mM NaCl, adding 0.1 mM CaCl2 reduced
10
stability of GO with increased pH. This is due to adsorption of Ca2+ ions on the surface
11
functional groups of GO which reduces the surface charge of GO. As the concentration of rGO
12
functional groups decreased, so did the influence of Ca2+ ions on rGO stability. Critical
13
coagulation concentrations (CCC) of GO, rGO-1h and rGO-2h were determined to be ~200 mM,
14
35 mM and 30 mM NaCl, respectively. In the presence of CaCl2, CCC values of GO and rGO
15
are quite similar, however. Long-term studies show that a significant amount of rGO-1h and
16
rGO-2h remain stable in Call’s Creek surface water, while effluent wastewater readily
17
destabilizes rGO. In the presence NOM and divalent cations (Ca2+, Mg2+), GO aggregates settle
18
from suspension due to GO functional group bridging with NOM and divalent ions. However,
19
rGO-1h and rGO-2h remain suspended due their lower functional group concentration and
20
resultant reduced NOM-divalent cation bridging. Overall, pH, divalent cations and NOM can
21
play complex roles in the fate of rGO and GO.
22 23
3 ACS Paragon Plus Environment
Environmental Science & Technology
24 25 26
1. Introduction
27
Graphene-based nanomaterials have attracted significant attention due to their exceptional
28
properties1-4 having potential electronic, energy, medical, and environmental applications.4-7
29
Graphene-based materials include pristine graphene, graphene oxide (GO) and reduced graphene
30
oxide (rGO),2 with GO being an intermediate product created during the synthesis of rGO. GO
31
can be reduced to rGO to improve electronic properties since its electrical conductivity is
32
significantly higher than that of GO’s. rGO is one of the most common forms of graphene-based
33
nanomaterials due to its superior electrical conductivity and electron mobility. rGO is widely
34
used in electronics, energy storage devices and biomedical applications.2, 8 With many promising
35
applications, rGO will eventually end up in the environment which necessitates a thorough
36
understanding of its environmental fate.
37
Besides large production of chemically reduced GO, most graphene-based materials in the
38
environment will occur in reduced form. GO is considered a metastable material which
39
undergoes spontaneous reduction.9, 10 UV irradiation11, natural reductants, bacteria,12 and biota
40
can accelerate its reduction in the environment. Recent studies show that GO is reduced to rGO
41
by sunlight exposure13, 14 and by bacterial degradation.12 Our recent study15 showed that GO
42
readily photo-reacts under simulated sunlight and photo-disproportionates to CO2-reduced
43
materials similar to rGO which are more fragmented low molecular-weight species. Naturally-
44
available metal-based reductants including Al16 and Fe17 can reduce GO to rGO. Sulfur-based
45
inorganic reductants commonly found in anaerobic environments have also been shown to
4 ACS Paragon Plus Environment
Page 4 of 30
Page 5 of 30
Environmental Science & Technology
46
reduce GO.18 Besides bacteria,12, 19 other studies found thatyeast,20 mussel polymer,21 carrot
47
root,22 and caffeic acid23 can transform GO to rGO.
48
To the best of our knowledge, there are no studies systematically investigating the
49
environmental fate and transport of rGO. Most recent work on environmental implications of
50
graphene-based nanomaterials primarily focused on GO24-27 and found that aggregation of GO25
51
follows colloidal theory including Derjaguin-Landau-Verwey-Overbeek (DLVO) theory28 and
52
the Schulze-Hardy rule.29
53
stability of GO due to steric repulsion and long-term stability studies showed GO is highly stable
54
in natural surface water.25 Studies on GO transport through porous media27, 30, 31 found that GO
55
is highly mobile in saturated porous media, that retention is reversible, that increased ionic
56
strength (IS) resulted in greater GO retention primarily due to straining, and that observed
57
transport trends can be explained by DLVO theory. It is still unknown whether findings from
58
GO studies will be applicable to rGO.
Presence of natural organic matter (NOM) noticeably increased the
59
Since most graphene-based nanomaterials in the environment are expected to be in reduced
60
form, understanding fate and transport of rGO is essential. Ours is the first study systematically
61
investigating the fate of reduced GO species in the aquatic environment. We studied the role of
62
surface oxidation state on the fate of rGO as well as long-term stability of rGO in natural and
63
engineered aquatic environments and found in particular that pH, ion valence and NOM play
64
significantly complex roles with rGO species. The overall mechanisms observed in this study
65
have been summarized in Scheme 1 and are described throughout the article.
66 67 68
2. Materials and Methods 2.1 Synthesis and Preparation of rGO 5 ACS Paragon Plus Environment
Environmental Science & Technology
69 70
In this study, GO and three differently-reduced GO (partially reduced GO (rGO-1h),
71
intermediate reduced GO (rGO-2h) and fully reduced GO (rGO-5h)) were used. GO was
72
synthesized using a modified Hummers method described previously and mentioned in the
73
Supporting Information (SI).23 In this study, rGO reduced by the Solvothermal process was used.
74
Chemically reduced GO, rather than rGO reduced via environmental processes, was used to
75
obtain similarly sized rGO sheets with varying oxidation states so that the fundamental role of
76
surface oxidations on the fate of rGO could be investigated. The reduction was achieved by
77
heating a purified GO suspension in N-Methyl-2-pyrrolidone (NMP) to 150°C with constant
78
stirring in a silicone oil bath. The heat reflux was performed for 1, 2, or 5 hours to achieve
79
varying levels of reduction. After solvothermal reduction, the rGO was separated from the NMP
80
using vacuum filtration with 0.1µm alumina filters (Millipore), rinsed heavily with DI water, and
81
re-dispersed in DI water at an approximate concentration of 1 mg/mL. Concentrations of GO and
82
rGO were determined by measuring mass following vacuum filtration. For AFM, and FTIR
83
characterization, the original dispersions in NMP were analyzed.
84
85 86
2.2 Characterization of rGO
87
Physical dimensions of rGO were measured by AFM; AFM samples were prepared as 32
88
described previously.
XPS and FTIR were utilized to determine surface functionalities of rGO.
89
The area under peaks was used for quantifying the relative concentration of functional groups as
90
shown in Table S1. Details regarding these methods are included in SI.
6 ACS Paragon Plus Environment
Page 6 of 30
Page 7 of 30
Environmental Science & Technology
91
Electrokinetic and hydrodynamic properties of rGO were determined over a range of solution
92
chemistries including different salt types (NaCl, CaCl2), varying ionic strength (IS), and NOM
93
concentrations.
94
Substances Society, MN) was used as standard NOM, and an SRHA stock solution was prepared
95
with accepted procedures.33,
96
(EPM), and zeta potential were measured with a ZetaSizer Nano ZS (Malvern Instrument,
97
Worcestershire, U.K.), following well-established techniques.
98
physicochemical properties of rGO was investigated over a pH range from 4 to 10 in the
99
presence of NaCl, CaCl2 and MgCl2.
Suwannee River Humic Acid Standard II (SRHA) (International Humic
34
rGO hydrodynamic diameter (Dh), electrophoretic mobility
Influence of pH on the
This titration experiment was conducted using an
100
autotitrator (MPT-2, Malvern Instrument, Worcestershire, U.K.); NaOH and HCl were the
101
titrants for controlling pH.
102 103 104
2.3 Aggregation Kinetics and Long Term Stability of rGO
105
Change of rGO Dh as a function of IS, ion valence, and presence of organic matter was
106
measured by time-resolved dynamic light scattering (TR-DLS).35 An rGO concentration of 10
107
mg/L provided a strong DLS signal and was therefore used in all aggregation studies. Both NaCl
108
and CaCl2 were background electrolytes. Details on the aggregation study are included in SI.
109
Long-term stability of GO and rGO was investigated in synthetic and natural waters to relate
110
well-controlled simple solution chemistries to more complex, environmentally relevant
111
conditions; eight types of water were used. Details on water types have been described in our
112
previous study25 and also in SI. 6 mL of 10 mg/L rGO suspended in the treatment water were
113
placed in a 20-mL borosilicate glass bottle (Fisher Scientific, PA) and continuously shaken at
114
100 rpm; rGO concentration and Dh were monitored for 28 days. Concentration was determined 7 ACS Paragon Plus Environment
Environmental Science & Technology
115
using a UV-vis plate reader (Enspire Multimode Reader 2300, PerkinElmer Inc, MA) at 230 nm
116
wavelength, and Dh was determined by a DynaPro Plate Reader II (Wyatt Technology, CA).
117
Calibration curves of GO and rGO using a UV-Vis plate reader at 230 nm are shown in Figure
118
S1. Samples were collected 2 cm below the top of sample for the long term study.
119 120 121
3. Results and Discussion 3.1 Characterization of rGO
122
AFM images (Figure 1) show that average vertical and lateral size of the flakes remained
123
similar from GO to rGO-5h, indicating the reduction process did not significantly change
124
physical dimensions of the graphene flakes. XPS results confirm the reduction of oxygen-
125
containing functional groups since GO was reduced to rGO-5h (Figure 1); FT-IR results show
126
functional groups on GO decreased with the reduction process (Figure S2) and the relative
127
amount of functional groups on GO and rGO, as determined from XPS spectra, is summarized in
128
Table S1. Progressing from unreduced GO to rGO-5h, the relative presence of carbon increases
129
from 40% to 64% while the presence of various functional groups correspondingly decreases.
130
We note that the percent of carboxyl functional groups increases as the reduction process
131
proceeds due to the higher thermal stability of carboxyl functional groups relative to carbonyl
132
and epoxide functional groups. Overall oxygen-to-carbon ratio also decreases as GO is reduced.
133
These results are qualitatively corroborated by FTIR.
134 135 136 137
3.2 Electrokinetic and Hydrodynamic Properties of rGO
8 ACS Paragon Plus Environment
Page 8 of 30
Page 9 of 30
Environmental Science & Technology
138
EPM increased from (-5.49 ± 0.12) × 10-8 m2V-1s-1 to (-3.89 ± 0.19) × 10-8 m2V-1s-1 as GO
139
was successively reduced (Figure 2). Interestingly, EPM decreased initially from GO to rGO-1h,
140
then remained similar from rGO-1h to rGO-5h, indicating that further reduction of GO does not
141
influence surface potential of rGO although the hydrodynamic size of graphene changed
142
noticeably. Dh of rGO increased from 373 nm to 1049 nm as GO was reduced fully to rGO-5h
143
(Figure 2). Even though EPM remained quite similar from rGO-1h to rGO-5h, increased Dh with
144
rGO indicates surface functional groups are playing major roles in aggregation. XPS results
145
show that functional groups of GO decrease as GO is reduced to rGO-5h which leads to higher
146
hydrophobicity of the materials and increased aggregation.
147
EPM of all rGO increases with greater IS due to charge screening, and remains quite similar
148
across NaCl concentrations (Figure S3). The influence of NaCl concentration on hydrodynamic
149
size of GO and rGO is significantly different, however. For GO, Dh increased from 489.9 ± 82.9
150
nm in 0.1 mM NaCl to 669.4 ± 94.5 nm in 100 mM NaCl (Figure S3); for both rGO-1h and rGO-
151
2h, hydrodynamic size increased above 2000 nm in 100 mM NaCl. Higher aggregation of GO
152
and rGO was observed in the presence of CaCl2 although EPM was significantly less than in
153
NaCl.
154
concentration on hydrodynamic size was similar for GO and rGO.
155 156 157
3.3 Influence of pH on Physicochemical Properties of rGO
158
Increased CaCl2 concentration resulted greater EPM, but the influence of CaCl2
3.3.1 pH Effect on GO In the presence of 10 mM NaCl, the EPM of GO decreased from (-3.44 ± 0.03) × 10-8 m2V-
159 160
1 -1
s to (-5.46 ± 0.15) × 10-8 m2V-1s-1 since pH increased from 4 to 10 (Figure 3) due to the
161
deprotonation of surface functional groups of GO with increased pH.25, 26 GO Dh was relatively 9 ACS Paragon Plus Environment
Environmental Science & Technology
Page 10 of 30
162
unchanged (~500 nm) throughout the pH range, however, which has been observed earlier. 10, 11
163
In the presence of 10 mM NaCl with 0.1 mM CaCl2, the EPM of GO increased from (-2.61 ±
164
0.06) × 10-8 m2V-1s-1 to (-1.72 ± 0.22) × 10-8 m2V-1s-1 as pH increased from 4 to 10, which is
165
completely opposite to the trend observed in the presence of NaCl alone. GO Dh increased from
166
460.2 ± 50.9 nm to 1354.5 ± 275.6 nm as pH increased from 4 to 10 in the presence of 0.1 mM
167
CaCl2 and 10 mM NaCl. This trend indicates a small amount of background CaCl2 (0.1 mM) can
168
significantly affect the physicochemical properties of GO as a function of pH. Since CaCl2
169
concentrations in natural aquatic environments often exceed 0.1 mM, the presence of Ca2+ ions
170
will play a major role in GO stability, an effect that may be due to binding capacity of Ca2+ ions
171
to functional groups of GO (Scheme 1).36, 37 Higher pH causes increased deprotonation of GO
172
surface functional groups, resulting in higher binding of Ca2+ ions with GO functional groups
173
and increased surface charge of GO (Scheme 1). Increased EPM also results in greater Dh due to
174
charge screening (Figure 3D). In the presence of 10 mM NaCl with 0.1 mM MgCl2 background,
175
EPM of GO decreased slightly from (-2.92 ± 0.06) × 10-8 m2V-1s-1 to (-3.20 ± 0.46) × 10-8 m2V-
176
1 -1
177
observed in the presence of 10 mM NaCl alone, indicating Mg2+ ions are also binding with GO
178
surface functional groups, but not as strongly as Ca2+ ions. GO Dh values remained similar across
179
the pH range in the presence of background 0.1 mM MgCl2.
s as pH increased from 4 to 10. This decrease in EPM is not as significant (p > 0.05) as that
180
3.3.2 pH Effect on rGO
181
In the presence of 10 mM NaCl only, EPM of rGO-1h decreased from (-3.28 ± 0.08) × 10-8
182
m2V-1s-1 to (-4.37 ± 0.08) × 10-8 m2V-1s-1 as pH increased from 4 to 10 (Figure 3B). The
183
decrease for rGO-1h as a function of pH is not as significant as that observed for GO because
184
rGO-1h contains fewer surface functional groups (Table S1), and deprotonation of surface 10 ACS Paragon Plus Environment
Page 11 of 30
Environmental Science & Technology
185
functional groups with increased pH is lower for rGO-1h (Scheme 1). rGO-1h Dh was also
186
relatively insensitive to pH changes in the presence of 10 mM NaCl. The EPM of rGO-1h
187
remained almost constant (~-2 × 10-8 m2V-1s-1) from pH 4 to 10 in the presence of 10 mM NaCl
188
with 0.1 mM CaCl2, and was significantly higher than in NaCl alone, indicating that binding of
189
Ca2+ ions with surface functional groups of rGO-1h influences electrokinetic properties of rGO-
190
1h. Increase of EPM due to binding of Ca2+ ions with rGO-1h is not as significant as that
191
observed for GO because rGO-1h has fewer surface functional groups (Scheme 1). Similarly, the
192
Dh of rGO-1h increased slightly from pH 4 to 10 in the presence of background 0.1 mM CaCl2,
193
but is not as significant as that observed for GO. In the presence of 10 mM NaCl and 0.1 mM
194
MgCl2, EPM of rGO-1h decreases as pH increased from 4 to 10. In the presence of background
195
0.1 mM MgCl2, Dh of rGO-1h remained fairly constant (~500 nm) throughout the pH range
196
investigated in this study.
197
In the presence of 10 mM NaCl only, both EPM and Dh of rGO-2h decreased from as pH
198
increased from 4 to 10 (Figure 3C, F). In the presence of 10 mM NaCl and 0.1 mM CaCl2, EPM
199
of rGO-2h decreased from (-1.29 ± 0.15) × 10-8 m2V-1s-1 to (-3.20 ± 0.17) × 10-8 m2V-1s-1 as pH
200
increased from 4 to 10, which is the opposite of the trend observed for GO in the presence of
201
background 0.1 mM CaCl2. Since rGO-2h contains a very low number of surface functional
202
groups (Table S1), binding capacity of Ca2+ ions with them is minimal. Hence, increased pH
203
does not result in further binding of Ca2+ ions with rGO-2h, and an increase in EPM as a function
204
of pH was not observed (Scheme 1). Dh of rGO-2h decreased from 2072.7 ± 310.8 nm to 741.7 ±
205
154.1 nm as a function of pH, which is also completely opposite to the trend for GO. In the
206
presence of 10 mM NaCl with 0.1 mM MgCl2, EPM of rGO-2h decreased from (-2.32 ± 0.03) ×
207
10-8 m2V-1s-1 to (-3.55 ± 0.14) × 10-8 m2V-1s-1 as pH increased from 4 to 10. Dh of rGO-2h 11 ACS Paragon Plus Environment
Environmental Science & Technology
208
decreased from 1071.1 ± 74.4 nm to 726.6 ± 43.2 nm as a function of pH in the presence of 0.1
209
mM MgCl2.
Page 12 of 30
210 211
3.4 Aggregation Kinetics of rGO
212
Aggregation profiles of rGO in 50 mM NaCl in the presence and absence of SRHA are
213
presented in Figure S4. As a function of time in 50 mM NaCl, the increase in GO Dh is quite
214
low, indicating GO is highly stable at this IS. rGO Dh values increase significantly as a function
215
of time, however, which indicates all rGOs are highly unstable at this IS. Aggregation rates
216
increase from rGO-1h to rGO-5h due to the loss of surface functional groups, as observed from
217
XPS and FTIR data. Adding SRHA reduces the aggregation of GO and all rGO (Figure S4)
218
although fully reduced rGO-5h is highly unstable even in the presence of SRHA. A similar trend
219
is observed in 1 mM CaCl2 (Figure S5).
220
Aggregation attachment efficiency (α) of GO and rGO increases with NaCl concentration
221
(Figure 4A) due to charge screening, however, GO α requires higher NaCl concentration to reach
222
1 (diffusion-limited region) than does rGO. α of rGO-1h and rGO-2h overlap as a function of
223
NaCl concentration, yielding critical coagulation concentrations (CCC) of around ~30-35 mM
224
NaCl. The GO CCC is near 200 mM NaCl and GO and rGO CCC’s are very similar (~0.7 mM
225
CaCl2) in the presence of CaCl2. A previous study38 on multi-walled carbon nanotubes
226
(MWNTs) observed a similar trend in the presence of CaCl2 for highly- and lowly-oxidized
227
MWNTs which the authors attributed to bidentate versus monodentate carboxylate complex
228
formation.
229 230
12 ACS Paragon Plus Environment
Page 13 of 30
231
Environmental Science & Technology
3.5 Long-Term Fate of rGO in the Natural Environment
232 233
Long-term stability of rGO in natural and artificial waters is summarized in Figure 5. In
234
Call’s Creek water, GO remains highly stable over a month-long study, which is likely due to the
235
low IS (conductivity 26 µS/cm) and presence of NOM (2.8 ± 0.57 mg/L). Moreover, Call’s
236
Creek primarily contains monovalent ions (Na+, K+) and a negligible amount of divalent ions
237
(Mg2+, Ca2+) (Table S2). Additionally, all rGO samples were less stable than GO in Call’s Creek
238
water which is consistent with the aggregation study. The suspended concentration of fully-
239
reduced GO (rGO-5h) decreases below 25% within one day, and below 10% within five days.
240
After one day, about 60% of both rGO-1h and rGO-2h were suspended in Call’s Creek water,
241
and after almost one month, ~15% of rGO-1h, ~15% rGO-2h and ~8% rGO-5h remain
242
suspended which indicates reduced GO may not be completely removed from the water column
243
in natural surface waters.
244
Long-term stability of rGO in synthetic surface water in the absence of NOM is shown in
245
Figure 5B. After one day, GO and all types of rGO concentrations decrease to less than 10% of
246
the initial concentration which indicates that graphene-based materials are not stable in surface
247
water in the absence of NOM. The addition of 5 mg/ L SRHA to the synthetic surface water
248
increased stability of all graphene-based materials investigated in this study (Figure 5C); after
249
one day, 80% GO, 53% rGO-1h and 45% rGO-2h remained suspended in synthetic surface water
250
in the presence of SRHA, but 90% suspended) in synthetic surface water in the
270
presence of SRHA. Stability of GO and rGO in synthetic surface water, with and without
271
divalent ions in the presence of SRHA after 28 days appears in Figure S6. It is obvious that
272
removing divalent ions significantly increases the stability of GO, confirming the role of divalent
273
ions in GO’s fate. rGO-1h and rGO-2h, on the other hand,
274
Ca2+ ions with SRHA than without, as observed from Figure S6. In fact, after just one day,
275
almost 50% of rGO-1h and rGO-2h remained suspended in synthetic surface water with SRHA,
276
while removing divalent ions drastically reduced the stability of rGO-1h and rGO-2h to ~10%.
are more stable in the presence of
14 ACS Paragon Plus Environment
Page 15 of 30
Environmental Science & Technology
277
This indicates that divalent ions can increase stability of rGO in the presence of NOM. For rGO,
278
Ca2+ ions can help binding of rGO with SRHA and enhance stability via steric repulsion. Similar
279
effects were not observed Call’s Creek water since the sample has a very low concentration of
280
divalent ions (