- Email: [email protected]
When supporting electrolyte matters – Tuning capacitive response of graphene oxide via electrochemical reduction in alkali and alkaline earth metal chlorides
Accepted Manuscript When supporting electrolyte matters – tuning capacitive response of graphene oxide via electrochemical reduction in alkali and alkaline earth metal chlorides
Dalibor Karačić, Selma Korać, Ana S. Dobrota, Igor A. Pašti, Natalia V. Skorodumova, Sanjin J. Gutić PII:
S0013-4686(18)32657-4
DOI:
10.1016/j.electacta.2018.11.173
Reference:
EA 33174
To appear in:
Electrochimica Acta
Received Date:
08 September 2018
Accepted Date:
25 November 2018
Please cite this article as: Dalibor Karačić, Selma Korać, Ana S. Dobrota, Igor A. Pašti, Natalia V. Skorodumova, Sanjin J. Gutić, When supporting electrolyte matters – tuning capacitive response of graphene oxide via electrochemical reduction in alkali and alkaline earth metal chlorides, Electrochimica Acta (2018), doi: 10.1016/j.electacta.2018.11.173
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
3
When supporting electrolyte matters – tuning capacitive response of graphene oxide via electrochemical reduction in alkali and alkaline earth metal chlorides
4
Dalibor Karačić1, Selma Korać2, Ana S. Dobrota1, Igor A. Pašti1,3*, Natalia V.
5
Skorodumova3,4, Sanjin J. Gutić2
1 2
6
1University
7
Belgrade, Serbia
8
2University
9
Bosne 33-35, 71000 Sarajevo, Bosnia and Herzegovina
of Belgrade – Faculty of Physical Chemistry, Studentski trg 12-16, 11158 of Sarajevo, Faculty of Science, Department of Chemistry, Zmaja od
10
3Department
11
and Management, KTH - Royal Institute of Technology, Brinellvägen 23, 100 44
12
Stockholm, Sweden
13
4Department
14
Uppsala, Sweden
of Materials Science and Engineering, School of Industrial Engineering
of Physics and Astronomy, Uppsala University, Box 516, 751 20
15
16 17 18 19 20 21
*corresponding author: Dr. Igor A. Pašti University of Belgrade – Faculty of Physical Chemistry Studentski trg 12-16, 11158 Belgrade, Serbia Phone: +381 11 3336 625 Email: [email protected]
ACCEPTED MANUSCRIPT 22
Abstract
23
The ability to tune charge storage properties of graphene oxide (GO) is of utmost
24
importance
25
electrochemical reduction of GO is highly sensitive to the cations present in the
26
solution. GO is reduced at more negative potential in alkali metal chloride solutions
27
than in alkaline earth metal chlorides. During the reduction, the capacitance of GO
28
increases from 10 to 70 times. The maximum capacitances of reduced GO are
29
between 65 and 130 F g−1, depending on the electrolyte and the presence of
30
conductive additive. We propose that different interactions of cations with oxygen
31
functional groups of GO during the reduction are responsible for the observed effect.
32
This hypothesis has been confirmed by Density Functional Theory calculations of
33
alkali and alkaline earth metals interactions with epoxy-functionalized graphene
34
sheet.
for
energy
conversion
applications.
Here
we
show
that
the
35 36
Keywords: graphene oxide; capacitance; electrochemical reduction; capacitance
37
tuning
38 39
ACCEPTED MANUSCRIPT 40
1. Introduction
41
Recent years witnessed rapid increase in knowledge on graphene and
42
graphene-based materials, as well as the emergence of new technologies for a
43
cheap and scalable production of different graphene-based nanostructures [1,2]. This
44
is driven not only by the scientific curiosity, but also by the potential applications of
45
these materials in almost all the areas of modern science and technology. Unique
46
properties of graphene-based materials important for energy storage and conversion
47
are already well acknowledged and used in different systems [3-5]. Among different
48
methods for graphene production, wet chemical methods enable cheap production of
49
large quantities of the „low-quality“ graphene [6]. However, from the electrochemical
50
point of view, this „low-quality“ graphene is the most interesting as the functional
51
groups and vacancies on graphene basal plane act as the active sites for different
52
processes important for the application in electrochemical systems. In terms of
53
charge storage application of graphene based materials, it is accepted that oxygen
54
and nitrogen functional groups [5,7-10] contribute to capacitance through
55
pseudocapacitive processes. Some recent works discuss the attempts for a precise
56
modification of oxygen functionalities on graphene and graphene oxide (GO) basal
57
plane, in order to finely tune its physical, chemical and electrochemical properties
58
[9,11-14]. Among different proposed methods, electrochemical reduction of GO
59
seems to be a competitive approach for such modifications, due to a simple
60
manipulation of the reduction process through the control of the potential and
61
reduction time, scalability and the avoidance of toxic reducing agents [15-18].
62
Number of different experimental conditions and their effects on the reduction
63
process and the properties of final reduced GO (rGO) can be found in literature [15-
64
21].
65
Recently, we have shown general trends in tuning capacitive properties of GO
66
[19]. As GO is reduced it gets de-oxygenated, while its conductivity increases.
67
Capacitance is maximized when the concentration of the oxygen functional groups is
68
properly balanced with the conductivity. The observed trends were shown for a
69
number of different GO materials. While it is known that the reduction of GO is pH
70
dependent [20,21], the question is whether the evolution of GO to rGO during
71
electrochemical reduction can be affected by other factors as well. In particular, we
72
look into the effects of the supporting electrolyte in the process of the electrochemical
73
reduction of GO. While in electrochemical textbooks the supporting electrolytes are
ACCEPTED MANUSCRIPT 74
considered as inert there are reports showing their distinct effect on various
75
electrochemical processes [22-25]. In this contribution, the effects of the alkali metal
76
(Li+, Na+, K+, Rb+ and Cs+) and alkaline earth metal cations (Mg2+, Ca2+, Sr2+ and
77
Ba2+) on the electrochemical reduction of GO film are investigated. These metals
78
have very low standard redox potentials and their cations cannot be reduced in
79
aqueous solutions. However, we show that the reduction of GO is affected by the
80
composition of the inert electrolyte, which also reflects on the capacitive properties of
81
such obtained rGO.
82 83
2. Experiment and theory
84
2.1. Electrochemical measurements
85
Aqueous graphene oxide suspension (standard solution, 4 mg ml−1;
86
Graphenea, Spain [26]) was diluted to obtain 1 mg ml−1 in 6:4 water/ethanol mixture.
87
Also, we prepared analogous GO dispersion containing 0.1 mg ml−1 of Vulcan XC-
88
72R as conductive additive. For each experiment, after intensive sonication, 10 μl of
89
GO the suspension was drop-casted onto the glassy carbon disc (0.196 cm2) and
90
dried under the vacuum at room temperature. Standard three-electrode cell, with Pt
91
foil (1 cm2) as a counter and Ag/AgCl (saturated KCl) as a reference electrode, was
92
used.
93
potentiostat/galvanostat, controlled by PowerSUITE interface, was used.
All
potentials
are
given
versus
this
reference.
PAR
263A
94
Electrochemical measurements were performed in aqueous solutions of alkali
95
(Li, Na, K, Rb, and Cs) and alkaline earth metal (Mg, Ca, Sr, Ba) chlorides (0.1 mol
96
dm−3), with pH adjusted to 5.5±0.1. After the GO electrode was prepared it was
97
transferred to a given solution and its capacitive response is recorded between −0.5
98
and 0.8 V. Then, the electrode was subjected to the reduction (potentiostatic, 10 s) at
99
different potentials (Ered) between −0.8 and −1.6 V. After each reduction capacitance
100
was measured between −0.5 and 0.8 V. Capacitance increment factors (CIF) are
101
evaluated as the capacitance of rGO reduced at given Ered and the capacitance of a
102
starting, non-reduced GO. Knowing the mass of GO loaded on the electrode (m)
103
specific capacitances were calculated as Cspec = q/(2×m×∆V) (q is the charge within
104
the cyclic voltammogram and ∆V is the width of potential window in which
105
capacitance is measured, 1.3 V). We have checked for the repeatability of electrode
106
preparation and confirmed it to be within 5% (by measurement of capacitive response
107
of 5 GO electrodes in the same electrolyte).
ACCEPTED MANUSCRIPT 108 109
2.2. Calculations
110
The first-principle DFT calculations were performed using the Vienna ab initio
111
simulation code (VASP) [27-30]. We used the generalized gradient approximation
112
(GGA) in the parametrization by Perdew, Burk and Ernzerhof [31] and the projector
113
augmented wave (PAW) method [32,33]. We used the PAW potentials, which include
114
the semi-core states into the valence band. A cut-off energy of 600 eV and Gaussian
115
smearing with a width of σ = 0.025 eV for the occupation of the electronic levels were
116
used. A Monkhorst-Pack Γ-centered 6×6×1 k-point mesh was used. The relaxation of
117
all the atoms within simulation cell was unrestricted. The relaxation procedure was
118
stopped when the Hellmann-Feynman forces on all atoms were below 10−2 eV Å−1.
119
Spin-polarization was taken into account in all calculations.
120 121
3. Results and discussion
122
The GO sample used in this work has a high concentration of oxygen
123
functional groups (C:O ratio close to 1) and a small concentration of heteroatoms (S
124
and N; below 5% in total). Due to a high oxidation degree it is non conductive, as
125
known for GO [19,34]. It undergoes irreversible electrochemical reduction, which is
126
observed under both potentiodynamic and potentiostatic conditions. Reduction can
127
be performed without conductive additive, as can be seen from the cyclic
128
voltammograms of GO reduction in MCl and MCl2 solution (M is alkali or alkaline
129
earth metal) in Fig. 1 (top panel).
130 131
[Figure 1]
132 133
It is clearly seen that the GO reduction takes place at more negative potentials in
134
MCl2 solutions compared to MCl solutions and that it strongly depends on the cation
135
present in the electrolyte. Reduction peak potential is the most negative for LiCl (−1.4
136
V) and shifts to more positive potentials for NaCl and KCl. For earth alkali metals
137
reduction peak potential is shifted to more positive potentials by at least 0.1 V.
138
However, we observed that the charges under reduction peak increase from Li to Rb
139
(from 23.4 mC to 31.5 mC) and remain nearly constant in the series of alkaline earth
140
metals (within experimental error, around 30 mC). Previously we observed that the
141
peak potential of the potentiodynamic reduction correlates with Ered at which the
ACCEPTED MANUSCRIPT 142
highest CIF is observed after potentiostatic reduction when different GOs are used
143
[19]. As can be seen in Fig. 1, such behaviour is preserved when reduction is
144
performed in different electrolytes. When GO is reduced in MCl and MCl2 solutions,
145
the initial capacitance (ranging from 1.5 to 3.2 F g−1, depending on the electrolyte)
146
increases 20 to 70 times (at Ered corresponding to maximum CIF).
147
Beside a clear dependence of the reduction process on the composition of the
148
“inert” electrolyte and separation of the reduction peaks between alkali and alkaline
149
earth metals, it is difficult to observe some distinct trend in Fig. 1. However, when
150
materials for electrochemical capacitors are used it is a common practice to include
151
some conductive additive to the electrode formulation, in order to improve current
152
collection efficiency. In the second set of experiments we used this strategy and
153
investigated GO reduction in MCl and MCl2 solutions when conductive additive is
154
included in the GO film (Fig. 2). Again, reduction in MCl solutions takes place at more
155
negative potentials compared to the MCl2 solutions. When compared to the case
156
where no conductive additive is included (Fig. 1), in all the cases the reduction peaks
157
are shifted to more positive potentials. We ascribe this effect to a higher conductivity
158
of GO films in this series of experiments. In this case, the integrated charges under
159
the GO reduction peaks remain rather constant, within the experimental error (around
160
30 mC for reduction in MCl and 29 mC for MCl2 solutions).
161 162
[Figure 2]
163 164
Moreover, in Fig. 2 it is clearly seen that, in general, the reduction peak
165
potential shifts to more positive potentials as the crystallographic radius of cation
166
increases (i.e., hydrated ion radius decreases [35]) in the series of alkali metals and
167
alkaline earth metals. In the same manner CIFs decrease. Moreover, when
168
compared to the results given in Fig. 1 (bottom panel) it can be observed that the
169
CIFs are significantly lower when conductive additive is used. As previously
170
explained [19] this is not a negative effect of the conductive carbon added to the GO
171
layer, but the consequence of higher initial GO capacitances due to a higher
172
conductivity of the GO/conductive component film. Namely, we measured the initial
173
specific capacitances (i.e. before the reduction) of GO/Vulcan XC-72R films around
174
10 F g−1.
ACCEPTED MANUSCRIPT 175
In Fig. 3(a) we show the evolution of capacitive response of the GO film
176
reduced in KCl solution at different potentials. Development of pseudocapacitance is
177
seen as a wide hump observed in the potential window −0.5 to 0.3 V. The evaluated
178
specific capacitances for different Ered are shown in Figs.3(b) and (c).
179 180
[Figure 3]
181 182
Cspec obtained upon the reduction show dispersion between 65 and 130 F g−1 (values
183
corresponding to the maximum CIFs, Figs. 1 and 2). This demonstrates an efficient
184
capacitance tuning via electrochemical reduction. The capacitances measured after
185
the reduction in MCl solutions are typically lower than the ones measured in MCl2
186
solutions. As in all the cases maximum CIFs correspond to the peak potential of the
187
potentiodynamic GO reduction, it is not surprising why in the alkali metal chlorides
188
and alkaline earth metal chlorides series we observe similar values of capacitances.
189
In all the cases the maximum capacitance is reached when the conductivity is
190
balanced with the concentration of oxygen functional groups, which contributes to
191
capacitance through the pseudofaradaic processes. Here we observe that this state
192
corresponds to roughly (55±5) % of oxygen functional groups removed from the GO
193
surface, irrespective of the electrolyte used (evaluated through the charge under the
194
cyclic voltammograms of GO reduction). The variations of Cspec are due to the
195
specific interactions of M+ and M2+ ions with the oxygen functional groups remaining
196
on the reduced GO surface. Also, Cspec measured in the presence of conductive
197
additive are higher than the ones measured upon the reduction of pure GO films. We
198
explain this result by the higher conductivity of reduced GO/Vulcan XC-72R films and
199
more open structure of the films containing Vulcan XC-72R. Namely, 50 nm globular
200
particles of Vulcan XC-72R [36] prevent extensive stacking of GO layers and provide
201
more access to the electrolyte during the potentiodynamic cycling. It should be noted
202
that a deep understanding of capacitive response of electrochemically produced GO
203
would benefit from the knowledge of the specific surface area of produced rGO,
204
usually determined using Brunauer-Emmett-Teller (BET) surface area analysis.
205
However, due to very small amounts of obtained GO (below 10 μg per experiment)
206
such kind of measurement is very difficult, while production of large quantities of rGO
207
would introduce significant uncertainties. In fact, the existing literature lacks of the
208
data regarding the change of the specific surface of GO upon electrochemical
ACCEPTED MANUSCRIPT 209
reduction. However, we strongly believe that short reduction processes (< 10 s)
210
during which GO film is immobilized on the electrode in the form of film, does not
211
allow the GO sheets to migrate and/or reorient significantly. Thus we do not expect
212
significant changes of the surface area due to restacking or graphene layers, in
213
contrast to such effect observed for the cases of chemical or thermal reduction of
214
GO.
215
Here we emphasize clear differences in the observed GO reduction peak
216
potentials (Figs. 1 and 2, upper panels) in different electrolytes. Irrespective of the
217
addition of the conductive component, reduction takes place at more negative
218
potentials in earth alkali metal chlorides. Moreover, in the presence of Vulcan XC-
219
72R the reduction potential shifts to more positive potentials when going from LiCl to
220
NaCl and KCl, while in RbCl and CsCl the peak potentials are close to the one in KCl.
221
In the case of the MCl2 solutions the peak potentials are close to each other for
222
MgCl2, CaCl2 and SrCl2, while it is more positive for BaCl2. As reduction is performed
223
at relatively deep cathodic potentials, the explanation of this behaviour can be sought
224
in the interactions of cations with the oxygen functional groups of GO. The theoretical
225
considerations of the alkali metals-oxygen functional groups interactions [8] have
226
shown that the charge transferred from the metal atom localizes in the vicinity of the
227
oxygen functional group which interacts with alkali metal atom. Based on these
228
results, we expect that doubly charged alkaline earth metal cations strongly localize
229
negative charge brought to the GO layer and in this way activate the reduction of the
230
oxygen functional groups and their removal from the GO surface more efficiently than
231
singly charged alkali metal cations.
232
In order to test this hypothesis we investigated the interaction of studied alkali
233
and alkaline earth metal atoms with an epoxy-functionalized graphene basal plane by
234
means of DFT calculations. We opted for the epoxy group as the used GO sample
235
has the largest contribution of C‒O moiety in the high resolution C1s XPS spectra
236
[26]. One epoxy group was introduced into the graphene simulation cell containing 32
237
carbon atoms (4×4 cell). The energy of interaction was quantified as the adsorption
238
energy (Eads) defined as:
239 240 241
Eads = EM@epoxy-graphene – Eepoxy-graphene – EM
(1)
ACCEPTED MANUSCRIPT 242
where EM@epoxy-graphene, Eepoxy-graphene and EM stand for the total energy of metal M
243
interacting with epoxy-graphene, the total energy of bare epoxy-graphene surface
244
and the total energy of isolated atom M. The obtained results (Fig. 4) show that alkali
245
metals, when brought to epoxy-graphene, open the epoxy group, moving the O atom
246
from its initial C‒C bridge position to a C-top site (Fig. 4). The situation is the same
247
for Mg, but other alkaline earth metal atoms spontaneously reduce epoxy graphene.
248
For the cases of Ca, Sr and Ba pristine graphene is restored and the MO fragment is
249
formed (Fig. 4). This process is followed by deliberation of significant amounts of
250
energy, much higher compared to the cases when the opening of epoxy group is
251
observed. Considering significant effect of alkali and alkaline earth metal cations on
252
the electrochemical reduction of GO it is interesting to draw a parallel to the effects of
253
the same cations on the cellulose pyrolysis process [37]. Shimada et al. [37]
254
demonstrated that the addition of NaCl and KCl did not change the weight loss
255
temperature of bulk cellulose, while the addition of MgCl2 and CaCl2 significantly
256
reduced this temperature. Hence, in both cases, the electrochemical reduction of GO
257
and pyrrolysis of cellulose, a lower energy input is required for the reaction in the
258
presence of alkaline earth metal cations.
259 260
[Figure 4]
261 262
Even though here we present the analysis of metal atoms interactions with epoxy-
263
graphene in the absence of any electrolyte, which could have a significant impact on
264
calculated Eads, the presented results help to rationalize the experimentally observed
265
behaviour, in particular the shift of the GO reduction peak to more positive potentials
266
in MCl2 solutions. Considering the observed trends in the series of MCl and MCl2, we
267
expect that ions of the same charge (1+ or 2+) with larger hydratation spheres
268
localize charge to lower extent, as they are located at greater distance from the GO
269
surface. For this reason, reduction in LiCl commences at more negative potentials
270
compared to the cases of NaCl and KCl. Cs+ and Rb+ have similar radii of hydrated
271
ions like K+ (around 230 pm [35]) and for this reason the reduction peak potentials
272
are close in these three cases (Fig. 2, upper panel). Finally we would like to
273
emphasize that the present work can also reconcile some opposing reports regarding
274
the evolution of oxygen functional groups upon electrochemical reduction [17,38].
275
Namely, due to the effects of pH and also different cations present in the electrolyte,
ACCEPTED MANUSCRIPT 276
the electrochemical reduction of GO can be significantly affected and results with
277
reduced GO samples with different surface chemical groups.
278 279
4. Conclusion
280
We have shown that the electrochemical reduction of GO is highly sensitive to
281
the type of cations present in the electrolyte. Reduction of GO takes place at more
282
negative potentials in alkali metal chloride solution, compared to the earth alkali metal
283
chlorides. Upon reduction capacitance increases dramatically, but Ered of the
284
maximum capacitance is dependent on the composition of the electrolyte. Using
285
electrochemical reduction maximum capacitances in the range 65 to 130 F g−1 are
286
obtained, which also depended on the presence of conductive component in the GO
287
film. We propose that the observed effect is due to specific interactions of cations
288
with present oxygen, which can activate the removal of oxygen functional groups at
289
different potentials.
290 291
Acknowledgements
292
A.S.D. and I.A.P. acknowledges the support provided by Serbian Ministry of
293
Education, Science and Technological Development (project III45014). N.V.S.
294
acknowledges the support provided by Swedish Research Council through the
295
project No. 2014-5993. We also acknowledge the support from Carl Tryggers
296
Foundation for Scientific Research. The computations were performed on resources
297
provided by the Swedish National Infrastructure for Computing (SNIC) at the High
298
Performance Computing Center North (HPC2N) at Umeå University. Authors are
299
grateful to Denis Sačer from Faculty of Chemical Engineering and Technology,
300
University of Zagreb, for providing us with the graphene oxide sample.
301 302
References
303
[1] Q. Tang, Z. Zhou, Z. Chen, Graphene-related nanomaterials: tuning properties by
304
functionalization, Nanoscale 5 (2013) 4541–4583
305
[2] K.S. Mali, J. Greenwood, J. Adisoejoso, R. Phillipson, S. De Feyter,
306
Nanostructuring
307
Nanoscale 7 (2015) 1566–1585
graphene
for
controlled
and
reproducible
functionalization,
ACCEPTED MANUSCRIPT 308
[3] B.L. Dasari, J.M.Nouri, D. Brabazon, S. Naher, Graphene and derivatives –
309
Synthesis techniques, properties and their energy applications, Energy 140 (2017)
310
766–778
311
[4] N.A.A. Ghany, S.A. Elsherif, H.T. Handal, Revolution of Graphene for different
312
applications: State-of-the-art, Surfaces and Interfaces 9 (2017) 93–106
313
[5] W. Yang, M. Ni, X. Ren, Y. Tian, N. Li, Y. Su, X. Zhang, Graphene in
314
Supercapacitor Applications, Curr. Opin.Colloid Interface Sci.20 (2015) 416–428
315
[6] H. Yu, B. Zhang, C. Bulin, R. Li, R. Xing, High-efficient Synthesis of Graphene
316
Oxide Based on Improved Hummers Method, Sci. Rep. 6 (2016) Article number:
317
36143
318
[7] Y.-H. Lee, K.-H. Chang, C.-C. Hu, Differentiate the pseudocapacitance and
319
double-layer capacitance contributions for nitrogen-doped reduced graphene oxide in
320
acidic and alkaline electrolytes, J. Power Sources 227 (2013) 300-308
321
[8] A.S. Dobrota, S. Gutić, A. Kalijadis, M. Baljozović, S.V. Mentus, N.V.
322
Skorodumova, I.A. Pašti, Stabilization of alkali metal ions interaction with OH-
323
functionalized graphene via clustering of OH groups – implications in charge storage
324
applications, RSC Adv. 6 (2016) 57910–57919
325
[9] Z. Liu, L. Jiang, L. Sheng, Q. Zhou, T. Wei, B. Zhang, Z. Fan, Oxygen Clusters
326
Distributed in Graphene with “Paddy Land” Structure: Ultrahigh Capacitance and
327
Rate Performance for Supercapacitors, Adv. Funct. Mater. 28 (2018) 1705258
328
[10] B. Xu, S. Yue, Z. Sui, X. Zhang, S. Hou, G. Cao, Y. Yang, What is the choice for
329
supercapacitors: graphene or graphene oxide?, Energy Environ. Sci. 4 (2011) 2826–
330
2830
331
[11] M. Wei, L Qiao, H. Zhang, S. Karakalos, K. Ma, Z. Fu, M.T. Swihart, G. Wu,
332
Engineering reduced graphene oxides with enhanced electrochemical properties
333
through multiple-step reductions, Electrochim. Acta 258 (2017) 735–743
334
[12] Z. Jovanovic, D. Bajuk-Bogdanović, S. Jovanović, Ž. Mravik, J. Kovač, I.
335
Holclajtner-Antunović, M. Vujković, The role of surface chemistry in the charge
336
storage properties of graphene oxide, Electrochim. Acta 258 (2017) 1228–1243
ACCEPTED MANUSCRIPT 337
[13] D. D’Angelo, C. Bongiorno, M. Amato, I. Deretzis, A. La Magna, E. Fazio, S.
338
Scalese, Oxygen Functionalities Evolution in Thermally Treated Graphene Oxide
339
Featured by EELS and DFT Calculations, J. Phys. Chem. C 121 (2017) 5408–5414
340
[14] A. Ganguly, S. Sharma, P. Papakonstantinou, J. Hamilton, Probing the Thermal
341
Deoxygenation of Graphene Oxide Using High-Resolution In Situ X-ray-Based
342
Spectroscopies, J. Phys. Chem. C 115 (2011) 17009–17019
343
[15] S.Y. Toh, K.S. Loh, S.K. Kamarudin, W.R.W. Daud, Graphene production via
344
electrochemical reduction of graphene oxide: Synthesis and characterisation, Chem.
345
Eng. J. 251 (2014) 422–434
346
[16] Y. Zhang, H. Hao, L. Wang, Effect of morphology and defect density on electron
347
transfer of electrochemically reduced graphene oxide, Appl. Surf. Sci. 390 (2016)
348
385-392
349
[17] S.Y. Toh, K.S. Loh, S.K. Kamarudin, W.R.W. Daud, The impact of
350
electrochemical reduction potentials on the electrocatalytic activity of graphene oxide
351
toward the oxygen reduction reaction in an alkaline medium, Electrochim. Acta 199
352
(2016) 194–203
353
[18] G. Liu, Y. Wang, D. Sun, Investigation of conductivity and catalytic ability at an
354
electrochemically reduced graphene oxide film modified electrode, J. Mater. Sci.
355
Mater. Electron. 26 (2015) 943–949
356
[19] S.J. Gutić, Dž. Kozlica, F. Korać, D. Bajuk-Bogdanović, M. Mitrić, V.M. Mirsky,
357
S.V. Mentus, I.A. Pašti, Electrochemical tuning of capacitive response of graphene
358
oxide, Phys. Chem. Chem. Phys. (2018) DOI: 10.1039/c8cp03631d
359
[20] M. Hilder, B. Winther-Jensen, D. Li, M. Forsythc, D.R. MacFarlane, Direct
360
electro-deposition of graphene from aqueous suspensions, Phys. Chem. Chem.
361
Phys. 13 (2011) 9187–9193
362
[21] J. Kauppila, P. Kunnas, P. Damlin, A. Viinikanoja, C. Kvarnström,
363
Electrochemical reduction of graphene oxide films in aqueous and organic solutions,
364
Electrochim. Acta 89 (2013) 84–89
ACCEPTED MANUSCRIPT 365
[22] M.R. Thorson, K.I. Siil, P.J. Kenis, Effect of Cations on the Electrochemical
366
Conversion of CO2 to CO, J. Electrochem. Soc. 160(1) (2013) F69–F74
367
[23] P.P. Lopes, D. Strmcnik, J.S. Jirkovsky, J.G.Connell, V. Stamenkovic, N.
368
Markovic, Double layer effects in electrocatalysis: The oxygen reduction reaction and
369
ethanol oxidation reaction on Au(111), Pt(111) and Ir(111) in alkaline media
370
containing Na and Li cations, Catal. Today 262 (2016) 41–47
371
[24] Z. Zainal, C.Y. Lee, M.Z. Hussein, A. Kassim, N.A. Yusof, Effect of supporting
372
electrolytes in electrochemically-assisted photodegradation of an azo dye, J.
373
Photoch. Photobio. A 172(3) (2005) 316–321
374
[25] N. Gavrilov, I.A. Pašti, M. Mitrić, J. Travas-Sejdić, G. Ćirić-Marjanović, S.V.
375
Mentus, Electrocatalysis of oxygen reduction reaction on polyaniline-derived
376
nitrogen-doped carbon nanoparticle surfaces in alkaline media, J. Power Sources
377
220 (2012) 306–316
378
[26]
379
oxide-4-mg-ml-water-dispersion-1000-ml, accessed August 12th 2018
380
[27] G. Kresse, J. Hafner, Ab initio molecular dynamics for liquid metals, Phys. Rev.
381
B 47 (1993) 558–561
382
[28] G. Kresse, J. Hafner. Ab initio molecular-dynamics simulation of the liquid-metal-
383
amorphous-semiconductor transition in germanium, Phys. Rev. B 49 (1994) 14251–
384
14269
385
[29] G. Kresse, J. Furthmüller, Efficiency of ab-initio total energy calculations for
386
metals and semiconductors using a plane-wave basis set, Comput. Mat. Sci. 6 (1996)
387
15–50
388
[30] G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy
389
calculations using a plane-wave basis set, Phys. Rev. B, 54 (1996) 11169–11186
390
[31] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made
391
simple, Phys. Rev. Lett. 77(18) (1996) 3865–3868
392
[32] P. E. Blochl, Projector augmented-wave method, Phys. Rev. B 50 (1994) 17953–
393
17979
https://www.graphenea.com/collections/graphene-oxide/products/graphene-
ACCEPTED MANUSCRIPT 394
[33] G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector
395
augmented-wave method, Phys. Rev. B 59 (1999) 1758–1775
396
[34] V.B. Mohan, R. Brown, K. Jayaraman, D. Bhattacharyya, Characterisation of
397
reduced graphene oxide: Effects of reduction variables on electrical conductivity, Mat.
398
Sci. Eng. B 193 (2015) 49–60
399
[35] G.W. Vanloon, S.J. Duffy, Environmental Chemistry: A global perspective, third
400
ed., Oxford University Press, New York, 2011, pp. 218–219
401
[36] http://www.fuelcellstore.com/vulcan-xc-72r accessed August 12th 2018.
402
[37] N. Shimada, H. Kawamoto, S. Saka, Different action of alkali/alkaline earth metal
403
chlorides on cellulose pyrolysis, J. Anal. Appl. Pyrol. 81(1) (2008) 80–87
404
[38] A. Ambrosi, M. Pumera, Precise tuning of surface composition and
405
electron‐transfer properties of graphene oxide films through electroreduction, Chem.
406
Eur. J., 19(15) (2013) 4748–4753
407
ACCEPTED MANUSCRIPT 408
Figure captions
409 410
Figure 1. top panel: cyclic voltammograms of GO reduction in MCl and MCl2
411
solutions (0.1 mol dm−3) (full lines – alkali metals, dashed lines – alkaline earth
412
metals); bottom panel: capacitance increment factors as the function of reduction
413
potential. For the presented measurements no conductive additive is used.
414 415
Figure 2. top panel: cyclic voltammograms of GO reduction in MCl and MCl2
416
solutions (0.1 mol dm−3) (full lines – alkali metals, dashed lines – alkaline earth
417
metals); bottom panel: capacitance increment factors as the function of reduction
418
potential. GO films contained Vulcan XC-72R as conductive additive.
419 420
Figure 3. (a) evolution of the capacitive response of GO upon reduction in KCl
421
solution at different reduction potentials (Ered; no conductive additive); (b) Maximum
422
specific capacitances (Cspec) of GO (corresponding to capacitances measured at Ered
423
which corresponds to the maximum CIF) and Cspec measured after reduction at −1.5
424
V in the absence of conductive additive; (c) the same as for (b) but with conductive
425
additive in the GO film.
426 427
Figure 4. Optimized structures of alkali and alkaline earth metal atoms interacting
428
with epoxy-graphene. Calculated adsorption energies (in eV per atom) are included
429
as well as M‒O and O‒C bond lengths (in Å).
ACCEPTED MANUSCRIPT
Highlights
Electrochemical reduction of graphene oxide is sensitive to cations GO is reduced at more negative potentials in alkali metal chloride solutions Capacitance of graphene oxide is maximized when reduced at optimal potential Capacitance is tuned between 65 and 130 F g−1. The effect is due to the interactions of cations with oxygen functional groups
Dalibor Karačić, Selma Korać, Ana S. Dobrota, Igor A. Pašti, Natalia V. Skorodumova, Sanjin J. Gutić PII:
S0013-4686(18)32657-4
DOI:
10.1016/j.electacta.2018.11.173
Reference:
EA 33174
To appear in:
Electrochimica Acta
Received Date:
08 September 2018
Accepted Date:
25 November 2018
Please cite this article as: Dalibor Karačić, Selma Korać, Ana S. Dobrota, Igor A. Pašti, Natalia V. Skorodumova, Sanjin J. Gutić, When supporting electrolyte matters – tuning capacitive response of graphene oxide via electrochemical reduction in alkali and alkaline earth metal chlorides, Electrochimica Acta (2018), doi: 10.1016/j.electacta.2018.11.173
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
3
When supporting electrolyte matters – tuning capacitive response of graphene oxide via electrochemical reduction in alkali and alkaline earth metal chlorides
4
Dalibor Karačić1, Selma Korać2, Ana S. Dobrota1, Igor A. Pašti1,3*, Natalia V.
5
Skorodumova3,4, Sanjin J. Gutić2
1 2
6
1University
7
Belgrade, Serbia
8
2University
9
Bosne 33-35, 71000 Sarajevo, Bosnia and Herzegovina
of Belgrade – Faculty of Physical Chemistry, Studentski trg 12-16, 11158 of Sarajevo, Faculty of Science, Department of Chemistry, Zmaja od
10
3Department
11
and Management, KTH - Royal Institute of Technology, Brinellvägen 23, 100 44
12
Stockholm, Sweden
13
4Department
14
Uppsala, Sweden
of Materials Science and Engineering, School of Industrial Engineering
of Physics and Astronomy, Uppsala University, Box 516, 751 20
15
16 17 18 19 20 21
*corresponding author: Dr. Igor A. Pašti University of Belgrade – Faculty of Physical Chemistry Studentski trg 12-16, 11158 Belgrade, Serbia Phone: +381 11 3336 625 Email: [email protected]
ACCEPTED MANUSCRIPT 22
Abstract
23
The ability to tune charge storage properties of graphene oxide (GO) is of utmost
24
importance
25
electrochemical reduction of GO is highly sensitive to the cations present in the
26
solution. GO is reduced at more negative potential in alkali metal chloride solutions
27
than in alkaline earth metal chlorides. During the reduction, the capacitance of GO
28
increases from 10 to 70 times. The maximum capacitances of reduced GO are
29
between 65 and 130 F g−1, depending on the electrolyte and the presence of
30
conductive additive. We propose that different interactions of cations with oxygen
31
functional groups of GO during the reduction are responsible for the observed effect.
32
This hypothesis has been confirmed by Density Functional Theory calculations of
33
alkali and alkaline earth metals interactions with epoxy-functionalized graphene
34
sheet.
for
energy
conversion
applications.
Here
we
show
that
the
35 36
Keywords: graphene oxide; capacitance; electrochemical reduction; capacitance
37
tuning
38 39
ACCEPTED MANUSCRIPT 40
1. Introduction
41
Recent years witnessed rapid increase in knowledge on graphene and
42
graphene-based materials, as well as the emergence of new technologies for a
43
cheap and scalable production of different graphene-based nanostructures [1,2]. This
44
is driven not only by the scientific curiosity, but also by the potential applications of
45
these materials in almost all the areas of modern science and technology. Unique
46
properties of graphene-based materials important for energy storage and conversion
47
are already well acknowledged and used in different systems [3-5]. Among different
48
methods for graphene production, wet chemical methods enable cheap production of
49
large quantities of the „low-quality“ graphene [6]. However, from the electrochemical
50
point of view, this „low-quality“ graphene is the most interesting as the functional
51
groups and vacancies on graphene basal plane act as the active sites for different
52
processes important for the application in electrochemical systems. In terms of
53
charge storage application of graphene based materials, it is accepted that oxygen
54
and nitrogen functional groups [5,7-10] contribute to capacitance through
55
pseudocapacitive processes. Some recent works discuss the attempts for a precise
56
modification of oxygen functionalities on graphene and graphene oxide (GO) basal
57
plane, in order to finely tune its physical, chemical and electrochemical properties
58
[9,11-14]. Among different proposed methods, electrochemical reduction of GO
59
seems to be a competitive approach for such modifications, due to a simple
60
manipulation of the reduction process through the control of the potential and
61
reduction time, scalability and the avoidance of toxic reducing agents [15-18].
62
Number of different experimental conditions and their effects on the reduction
63
process and the properties of final reduced GO (rGO) can be found in literature [15-
64
21].
65
Recently, we have shown general trends in tuning capacitive properties of GO
66
[19]. As GO is reduced it gets de-oxygenated, while its conductivity increases.
67
Capacitance is maximized when the concentration of the oxygen functional groups is
68
properly balanced with the conductivity. The observed trends were shown for a
69
number of different GO materials. While it is known that the reduction of GO is pH
70
dependent [20,21], the question is whether the evolution of GO to rGO during
71
electrochemical reduction can be affected by other factors as well. In particular, we
72
look into the effects of the supporting electrolyte in the process of the electrochemical
73
reduction of GO. While in electrochemical textbooks the supporting electrolytes are
ACCEPTED MANUSCRIPT 74
considered as inert there are reports showing their distinct effect on various
75
electrochemical processes [22-25]. In this contribution, the effects of the alkali metal
76
(Li+, Na+, K+, Rb+ and Cs+) and alkaline earth metal cations (Mg2+, Ca2+, Sr2+ and
77
Ba2+) on the electrochemical reduction of GO film are investigated. These metals
78
have very low standard redox potentials and their cations cannot be reduced in
79
aqueous solutions. However, we show that the reduction of GO is affected by the
80
composition of the inert electrolyte, which also reflects on the capacitive properties of
81
such obtained rGO.
82 83
2. Experiment and theory
84
2.1. Electrochemical measurements
85
Aqueous graphene oxide suspension (standard solution, 4 mg ml−1;
86
Graphenea, Spain [26]) was diluted to obtain 1 mg ml−1 in 6:4 water/ethanol mixture.
87
Also, we prepared analogous GO dispersion containing 0.1 mg ml−1 of Vulcan XC-
88
72R as conductive additive. For each experiment, after intensive sonication, 10 μl of
89
GO the suspension was drop-casted onto the glassy carbon disc (0.196 cm2) and
90
dried under the vacuum at room temperature. Standard three-electrode cell, with Pt
91
foil (1 cm2) as a counter and Ag/AgCl (saturated KCl) as a reference electrode, was
92
used.
93
potentiostat/galvanostat, controlled by PowerSUITE interface, was used.
All
potentials
are
given
versus
this
reference.
PAR
263A
94
Electrochemical measurements were performed in aqueous solutions of alkali
95
(Li, Na, K, Rb, and Cs) and alkaline earth metal (Mg, Ca, Sr, Ba) chlorides (0.1 mol
96
dm−3), with pH adjusted to 5.5±0.1. After the GO electrode was prepared it was
97
transferred to a given solution and its capacitive response is recorded between −0.5
98
and 0.8 V. Then, the electrode was subjected to the reduction (potentiostatic, 10 s) at
99
different potentials (Ered) between −0.8 and −1.6 V. After each reduction capacitance
100
was measured between −0.5 and 0.8 V. Capacitance increment factors (CIF) are
101
evaluated as the capacitance of rGO reduced at given Ered and the capacitance of a
102
starting, non-reduced GO. Knowing the mass of GO loaded on the electrode (m)
103
specific capacitances were calculated as Cspec = q/(2×m×∆V) (q is the charge within
104
the cyclic voltammogram and ∆V is the width of potential window in which
105
capacitance is measured, 1.3 V). We have checked for the repeatability of electrode
106
preparation and confirmed it to be within 5% (by measurement of capacitive response
107
of 5 GO electrodes in the same electrolyte).
ACCEPTED MANUSCRIPT 108 109
2.2. Calculations
110
The first-principle DFT calculations were performed using the Vienna ab initio
111
simulation code (VASP) [27-30]. We used the generalized gradient approximation
112
(GGA) in the parametrization by Perdew, Burk and Ernzerhof [31] and the projector
113
augmented wave (PAW) method [32,33]. We used the PAW potentials, which include
114
the semi-core states into the valence band. A cut-off energy of 600 eV and Gaussian
115
smearing with a width of σ = 0.025 eV for the occupation of the electronic levels were
116
used. A Monkhorst-Pack Γ-centered 6×6×1 k-point mesh was used. The relaxation of
117
all the atoms within simulation cell was unrestricted. The relaxation procedure was
118
stopped when the Hellmann-Feynman forces on all atoms were below 10−2 eV Å−1.
119
Spin-polarization was taken into account in all calculations.
120 121
3. Results and discussion
122
The GO sample used in this work has a high concentration of oxygen
123
functional groups (C:O ratio close to 1) and a small concentration of heteroatoms (S
124
and N; below 5% in total). Due to a high oxidation degree it is non conductive, as
125
known for GO [19,34]. It undergoes irreversible electrochemical reduction, which is
126
observed under both potentiodynamic and potentiostatic conditions. Reduction can
127
be performed without conductive additive, as can be seen from the cyclic
128
voltammograms of GO reduction in MCl and MCl2 solution (M is alkali or alkaline
129
earth metal) in Fig. 1 (top panel).
130 131
[Figure 1]
132 133
It is clearly seen that the GO reduction takes place at more negative potentials in
134
MCl2 solutions compared to MCl solutions and that it strongly depends on the cation
135
present in the electrolyte. Reduction peak potential is the most negative for LiCl (−1.4
136
V) and shifts to more positive potentials for NaCl and KCl. For earth alkali metals
137
reduction peak potential is shifted to more positive potentials by at least 0.1 V.
138
However, we observed that the charges under reduction peak increase from Li to Rb
139
(from 23.4 mC to 31.5 mC) and remain nearly constant in the series of alkaline earth
140
metals (within experimental error, around 30 mC). Previously we observed that the
141
peak potential of the potentiodynamic reduction correlates with Ered at which the
ACCEPTED MANUSCRIPT 142
highest CIF is observed after potentiostatic reduction when different GOs are used
143
[19]. As can be seen in Fig. 1, such behaviour is preserved when reduction is
144
performed in different electrolytes. When GO is reduced in MCl and MCl2 solutions,
145
the initial capacitance (ranging from 1.5 to 3.2 F g−1, depending on the electrolyte)
146
increases 20 to 70 times (at Ered corresponding to maximum CIF).
147
Beside a clear dependence of the reduction process on the composition of the
148
“inert” electrolyte and separation of the reduction peaks between alkali and alkaline
149
earth metals, it is difficult to observe some distinct trend in Fig. 1. However, when
150
materials for electrochemical capacitors are used it is a common practice to include
151
some conductive additive to the electrode formulation, in order to improve current
152
collection efficiency. In the second set of experiments we used this strategy and
153
investigated GO reduction in MCl and MCl2 solutions when conductive additive is
154
included in the GO film (Fig. 2). Again, reduction in MCl solutions takes place at more
155
negative potentials compared to the MCl2 solutions. When compared to the case
156
where no conductive additive is included (Fig. 1), in all the cases the reduction peaks
157
are shifted to more positive potentials. We ascribe this effect to a higher conductivity
158
of GO films in this series of experiments. In this case, the integrated charges under
159
the GO reduction peaks remain rather constant, within the experimental error (around
160
30 mC for reduction in MCl and 29 mC for MCl2 solutions).
161 162
[Figure 2]
163 164
Moreover, in Fig. 2 it is clearly seen that, in general, the reduction peak
165
potential shifts to more positive potentials as the crystallographic radius of cation
166
increases (i.e., hydrated ion radius decreases [35]) in the series of alkali metals and
167
alkaline earth metals. In the same manner CIFs decrease. Moreover, when
168
compared to the results given in Fig. 1 (bottom panel) it can be observed that the
169
CIFs are significantly lower when conductive additive is used. As previously
170
explained [19] this is not a negative effect of the conductive carbon added to the GO
171
layer, but the consequence of higher initial GO capacitances due to a higher
172
conductivity of the GO/conductive component film. Namely, we measured the initial
173
specific capacitances (i.e. before the reduction) of GO/Vulcan XC-72R films around
174
10 F g−1.
ACCEPTED MANUSCRIPT 175
In Fig. 3(a) we show the evolution of capacitive response of the GO film
176
reduced in KCl solution at different potentials. Development of pseudocapacitance is
177
seen as a wide hump observed in the potential window −0.5 to 0.3 V. The evaluated
178
specific capacitances for different Ered are shown in Figs.3(b) and (c).
179 180
[Figure 3]
181 182
Cspec obtained upon the reduction show dispersion between 65 and 130 F g−1 (values
183
corresponding to the maximum CIFs, Figs. 1 and 2). This demonstrates an efficient
184
capacitance tuning via electrochemical reduction. The capacitances measured after
185
the reduction in MCl solutions are typically lower than the ones measured in MCl2
186
solutions. As in all the cases maximum CIFs correspond to the peak potential of the
187
potentiodynamic GO reduction, it is not surprising why in the alkali metal chlorides
188
and alkaline earth metal chlorides series we observe similar values of capacitances.
189
In all the cases the maximum capacitance is reached when the conductivity is
190
balanced with the concentration of oxygen functional groups, which contributes to
191
capacitance through the pseudofaradaic processes. Here we observe that this state
192
corresponds to roughly (55±5) % of oxygen functional groups removed from the GO
193
surface, irrespective of the electrolyte used (evaluated through the charge under the
194
cyclic voltammograms of GO reduction). The variations of Cspec are due to the
195
specific interactions of M+ and M2+ ions with the oxygen functional groups remaining
196
on the reduced GO surface. Also, Cspec measured in the presence of conductive
197
additive are higher than the ones measured upon the reduction of pure GO films. We
198
explain this result by the higher conductivity of reduced GO/Vulcan XC-72R films and
199
more open structure of the films containing Vulcan XC-72R. Namely, 50 nm globular
200
particles of Vulcan XC-72R [36] prevent extensive stacking of GO layers and provide
201
more access to the electrolyte during the potentiodynamic cycling. It should be noted
202
that a deep understanding of capacitive response of electrochemically produced GO
203
would benefit from the knowledge of the specific surface area of produced rGO,
204
usually determined using Brunauer-Emmett-Teller (BET) surface area analysis.
205
However, due to very small amounts of obtained GO (below 10 μg per experiment)
206
such kind of measurement is very difficult, while production of large quantities of rGO
207
would introduce significant uncertainties. In fact, the existing literature lacks of the
208
data regarding the change of the specific surface of GO upon electrochemical
ACCEPTED MANUSCRIPT 209
reduction. However, we strongly believe that short reduction processes (< 10 s)
210
during which GO film is immobilized on the electrode in the form of film, does not
211
allow the GO sheets to migrate and/or reorient significantly. Thus we do not expect
212
significant changes of the surface area due to restacking or graphene layers, in
213
contrast to such effect observed for the cases of chemical or thermal reduction of
214
GO.
215
Here we emphasize clear differences in the observed GO reduction peak
216
potentials (Figs. 1 and 2, upper panels) in different electrolytes. Irrespective of the
217
addition of the conductive component, reduction takes place at more negative
218
potentials in earth alkali metal chlorides. Moreover, in the presence of Vulcan XC-
219
72R the reduction potential shifts to more positive potentials when going from LiCl to
220
NaCl and KCl, while in RbCl and CsCl the peak potentials are close to the one in KCl.
221
In the case of the MCl2 solutions the peak potentials are close to each other for
222
MgCl2, CaCl2 and SrCl2, while it is more positive for BaCl2. As reduction is performed
223
at relatively deep cathodic potentials, the explanation of this behaviour can be sought
224
in the interactions of cations with the oxygen functional groups of GO. The theoretical
225
considerations of the alkali metals-oxygen functional groups interactions [8] have
226
shown that the charge transferred from the metal atom localizes in the vicinity of the
227
oxygen functional group which interacts with alkali metal atom. Based on these
228
results, we expect that doubly charged alkaline earth metal cations strongly localize
229
negative charge brought to the GO layer and in this way activate the reduction of the
230
oxygen functional groups and their removal from the GO surface more efficiently than
231
singly charged alkali metal cations.
232
In order to test this hypothesis we investigated the interaction of studied alkali
233
and alkaline earth metal atoms with an epoxy-functionalized graphene basal plane by
234
means of DFT calculations. We opted for the epoxy group as the used GO sample
235
has the largest contribution of C‒O moiety in the high resolution C1s XPS spectra
236
[26]. One epoxy group was introduced into the graphene simulation cell containing 32
237
carbon atoms (4×4 cell). The energy of interaction was quantified as the adsorption
238
energy (Eads) defined as:
239 240 241
Eads = EM@epoxy-graphene – Eepoxy-graphene – EM
(1)
ACCEPTED MANUSCRIPT 242
where EM@epoxy-graphene, Eepoxy-graphene and EM stand for the total energy of metal M
243
interacting with epoxy-graphene, the total energy of bare epoxy-graphene surface
244
and the total energy of isolated atom M. The obtained results (Fig. 4) show that alkali
245
metals, when brought to epoxy-graphene, open the epoxy group, moving the O atom
246
from its initial C‒C bridge position to a C-top site (Fig. 4). The situation is the same
247
for Mg, but other alkaline earth metal atoms spontaneously reduce epoxy graphene.
248
For the cases of Ca, Sr and Ba pristine graphene is restored and the MO fragment is
249
formed (Fig. 4). This process is followed by deliberation of significant amounts of
250
energy, much higher compared to the cases when the opening of epoxy group is
251
observed. Considering significant effect of alkali and alkaline earth metal cations on
252
the electrochemical reduction of GO it is interesting to draw a parallel to the effects of
253
the same cations on the cellulose pyrolysis process [37]. Shimada et al. [37]
254
demonstrated that the addition of NaCl and KCl did not change the weight loss
255
temperature of bulk cellulose, while the addition of MgCl2 and CaCl2 significantly
256
reduced this temperature. Hence, in both cases, the electrochemical reduction of GO
257
and pyrrolysis of cellulose, a lower energy input is required for the reaction in the
258
presence of alkaline earth metal cations.
259 260
[Figure 4]
261 262
Even though here we present the analysis of metal atoms interactions with epoxy-
263
graphene in the absence of any electrolyte, which could have a significant impact on
264
calculated Eads, the presented results help to rationalize the experimentally observed
265
behaviour, in particular the shift of the GO reduction peak to more positive potentials
266
in MCl2 solutions. Considering the observed trends in the series of MCl and MCl2, we
267
expect that ions of the same charge (1+ or 2+) with larger hydratation spheres
268
localize charge to lower extent, as they are located at greater distance from the GO
269
surface. For this reason, reduction in LiCl commences at more negative potentials
270
compared to the cases of NaCl and KCl. Cs+ and Rb+ have similar radii of hydrated
271
ions like K+ (around 230 pm [35]) and for this reason the reduction peak potentials
272
are close in these three cases (Fig. 2, upper panel). Finally we would like to
273
emphasize that the present work can also reconcile some opposing reports regarding
274
the evolution of oxygen functional groups upon electrochemical reduction [17,38].
275
Namely, due to the effects of pH and also different cations present in the electrolyte,
ACCEPTED MANUSCRIPT 276
the electrochemical reduction of GO can be significantly affected and results with
277
reduced GO samples with different surface chemical groups.
278 279
4. Conclusion
280
We have shown that the electrochemical reduction of GO is highly sensitive to
281
the type of cations present in the electrolyte. Reduction of GO takes place at more
282
negative potentials in alkali metal chloride solution, compared to the earth alkali metal
283
chlorides. Upon reduction capacitance increases dramatically, but Ered of the
284
maximum capacitance is dependent on the composition of the electrolyte. Using
285
electrochemical reduction maximum capacitances in the range 65 to 130 F g−1 are
286
obtained, which also depended on the presence of conductive component in the GO
287
film. We propose that the observed effect is due to specific interactions of cations
288
with present oxygen, which can activate the removal of oxygen functional groups at
289
different potentials.
290 291
Acknowledgements
292
A.S.D. and I.A.P. acknowledges the support provided by Serbian Ministry of
293
Education, Science and Technological Development (project III45014). N.V.S.
294
acknowledges the support provided by Swedish Research Council through the
295
project No. 2014-5993. We also acknowledge the support from Carl Tryggers
296
Foundation for Scientific Research. The computations were performed on resources
297
provided by the Swedish National Infrastructure for Computing (SNIC) at the High
298
Performance Computing Center North (HPC2N) at Umeå University. Authors are
299
grateful to Denis Sačer from Faculty of Chemical Engineering and Technology,
300
University of Zagreb, for providing us with the graphene oxide sample.
301 302
References
303
[1] Q. Tang, Z. Zhou, Z. Chen, Graphene-related nanomaterials: tuning properties by
304
functionalization, Nanoscale 5 (2013) 4541–4583
305
[2] K.S. Mali, J. Greenwood, J. Adisoejoso, R. Phillipson, S. De Feyter,
306
Nanostructuring
307
Nanoscale 7 (2015) 1566–1585
graphene
for
controlled
and
reproducible
functionalization,
ACCEPTED MANUSCRIPT 308
[3] B.L. Dasari, J.M.Nouri, D. Brabazon, S. Naher, Graphene and derivatives –
309
Synthesis techniques, properties and their energy applications, Energy 140 (2017)
310
766–778
311
[4] N.A.A. Ghany, S.A. Elsherif, H.T. Handal, Revolution of Graphene for different
312
applications: State-of-the-art, Surfaces and Interfaces 9 (2017) 93–106
313
[5] W. Yang, M. Ni, X. Ren, Y. Tian, N. Li, Y. Su, X. Zhang, Graphene in
314
Supercapacitor Applications, Curr. Opin.Colloid Interface Sci.20 (2015) 416–428
315
[6] H. Yu, B. Zhang, C. Bulin, R. Li, R. Xing, High-efficient Synthesis of Graphene
316
Oxide Based on Improved Hummers Method, Sci. Rep. 6 (2016) Article number:
317
36143
318
[7] Y.-H. Lee, K.-H. Chang, C.-C. Hu, Differentiate the pseudocapacitance and
319
double-layer capacitance contributions for nitrogen-doped reduced graphene oxide in
320
acidic and alkaline electrolytes, J. Power Sources 227 (2013) 300-308
321
[8] A.S. Dobrota, S. Gutić, A. Kalijadis, M. Baljozović, S.V. Mentus, N.V.
322
Skorodumova, I.A. Pašti, Stabilization of alkali metal ions interaction with OH-
323
functionalized graphene via clustering of OH groups – implications in charge storage
324
applications, RSC Adv. 6 (2016) 57910–57919
325
[9] Z. Liu, L. Jiang, L. Sheng, Q. Zhou, T. Wei, B. Zhang, Z. Fan, Oxygen Clusters
326
Distributed in Graphene with “Paddy Land” Structure: Ultrahigh Capacitance and
327
Rate Performance for Supercapacitors, Adv. Funct. Mater. 28 (2018) 1705258
328
[10] B. Xu, S. Yue, Z. Sui, X. Zhang, S. Hou, G. Cao, Y. Yang, What is the choice for
329
supercapacitors: graphene or graphene oxide?, Energy Environ. Sci. 4 (2011) 2826–
330
2830
331
[11] M. Wei, L Qiao, H. Zhang, S. Karakalos, K. Ma, Z. Fu, M.T. Swihart, G. Wu,
332
Engineering reduced graphene oxides with enhanced electrochemical properties
333
through multiple-step reductions, Electrochim. Acta 258 (2017) 735–743
334
[12] Z. Jovanovic, D. Bajuk-Bogdanović, S. Jovanović, Ž. Mravik, J. Kovač, I.
335
Holclajtner-Antunović, M. Vujković, The role of surface chemistry in the charge
336
storage properties of graphene oxide, Electrochim. Acta 258 (2017) 1228–1243
ACCEPTED MANUSCRIPT 337
[13] D. D’Angelo, C. Bongiorno, M. Amato, I. Deretzis, A. La Magna, E. Fazio, S.
338
Scalese, Oxygen Functionalities Evolution in Thermally Treated Graphene Oxide
339
Featured by EELS and DFT Calculations, J. Phys. Chem. C 121 (2017) 5408–5414
340
[14] A. Ganguly, S. Sharma, P. Papakonstantinou, J. Hamilton, Probing the Thermal
341
Deoxygenation of Graphene Oxide Using High-Resolution In Situ X-ray-Based
342
Spectroscopies, J. Phys. Chem. C 115 (2011) 17009–17019
343
[15] S.Y. Toh, K.S. Loh, S.K. Kamarudin, W.R.W. Daud, Graphene production via
344
electrochemical reduction of graphene oxide: Synthesis and characterisation, Chem.
345
Eng. J. 251 (2014) 422–434
346
[16] Y. Zhang, H. Hao, L. Wang, Effect of morphology and defect density on electron
347
transfer of electrochemically reduced graphene oxide, Appl. Surf. Sci. 390 (2016)
348
385-392
349
[17] S.Y. Toh, K.S. Loh, S.K. Kamarudin, W.R.W. Daud, The impact of
350
electrochemical reduction potentials on the electrocatalytic activity of graphene oxide
351
toward the oxygen reduction reaction in an alkaline medium, Electrochim. Acta 199
352
(2016) 194–203
353
[18] G. Liu, Y. Wang, D. Sun, Investigation of conductivity and catalytic ability at an
354
electrochemically reduced graphene oxide film modified electrode, J. Mater. Sci.
355
Mater. Electron. 26 (2015) 943–949
356
[19] S.J. Gutić, Dž. Kozlica, F. Korać, D. Bajuk-Bogdanović, M. Mitrić, V.M. Mirsky,
357
S.V. Mentus, I.A. Pašti, Electrochemical tuning of capacitive response of graphene
358
oxide, Phys. Chem. Chem. Phys. (2018) DOI: 10.1039/c8cp03631d
359
[20] M. Hilder, B. Winther-Jensen, D. Li, M. Forsythc, D.R. MacFarlane, Direct
360
electro-deposition of graphene from aqueous suspensions, Phys. Chem. Chem.
361
Phys. 13 (2011) 9187–9193
362
[21] J. Kauppila, P. Kunnas, P. Damlin, A. Viinikanoja, C. Kvarnström,
363
Electrochemical reduction of graphene oxide films in aqueous and organic solutions,
364
Electrochim. Acta 89 (2013) 84–89
ACCEPTED MANUSCRIPT 365
[22] M.R. Thorson, K.I. Siil, P.J. Kenis, Effect of Cations on the Electrochemical
366
Conversion of CO2 to CO, J. Electrochem. Soc. 160(1) (2013) F69–F74
367
[23] P.P. Lopes, D. Strmcnik, J.S. Jirkovsky, J.G.Connell, V. Stamenkovic, N.
368
Markovic, Double layer effects in electrocatalysis: The oxygen reduction reaction and
369
ethanol oxidation reaction on Au(111), Pt(111) and Ir(111) in alkaline media
370
containing Na and Li cations, Catal. Today 262 (2016) 41–47
371
[24] Z. Zainal, C.Y. Lee, M.Z. Hussein, A. Kassim, N.A. Yusof, Effect of supporting
372
electrolytes in electrochemically-assisted photodegradation of an azo dye, J.
373
Photoch. Photobio. A 172(3) (2005) 316–321
374
[25] N. Gavrilov, I.A. Pašti, M. Mitrić, J. Travas-Sejdić, G. Ćirić-Marjanović, S.V.
375
Mentus, Electrocatalysis of oxygen reduction reaction on polyaniline-derived
376
nitrogen-doped carbon nanoparticle surfaces in alkaline media, J. Power Sources
377
220 (2012) 306–316
378
[26]
379
oxide-4-mg-ml-water-dispersion-1000-ml, accessed August 12th 2018
380
[27] G. Kresse, J. Hafner, Ab initio molecular dynamics for liquid metals, Phys. Rev.
381
B 47 (1993) 558–561
382
[28] G. Kresse, J. Hafner. Ab initio molecular-dynamics simulation of the liquid-metal-
383
amorphous-semiconductor transition in germanium, Phys. Rev. B 49 (1994) 14251–
384
14269
385
[29] G. Kresse, J. Furthmüller, Efficiency of ab-initio total energy calculations for
386
metals and semiconductors using a plane-wave basis set, Comput. Mat. Sci. 6 (1996)
387
15–50
388
[30] G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy
389
calculations using a plane-wave basis set, Phys. Rev. B, 54 (1996) 11169–11186
390
[31] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made
391
simple, Phys. Rev. Lett. 77(18) (1996) 3865–3868
392
[32] P. E. Blochl, Projector augmented-wave method, Phys. Rev. B 50 (1994) 17953–
393
17979
https://www.graphenea.com/collections/graphene-oxide/products/graphene-
ACCEPTED MANUSCRIPT 394
[33] G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector
395
augmented-wave method, Phys. Rev. B 59 (1999) 1758–1775
396
[34] V.B. Mohan, R. Brown, K. Jayaraman, D. Bhattacharyya, Characterisation of
397
reduced graphene oxide: Effects of reduction variables on electrical conductivity, Mat.
398
Sci. Eng. B 193 (2015) 49–60
399
[35] G.W. Vanloon, S.J. Duffy, Environmental Chemistry: A global perspective, third
400
ed., Oxford University Press, New York, 2011, pp. 218–219
401
[36] http://www.fuelcellstore.com/vulcan-xc-72r accessed August 12th 2018.
402
[37] N. Shimada, H. Kawamoto, S. Saka, Different action of alkali/alkaline earth metal
403
chlorides on cellulose pyrolysis, J. Anal. Appl. Pyrol. 81(1) (2008) 80–87
404
[38] A. Ambrosi, M. Pumera, Precise tuning of surface composition and
405
electron‐transfer properties of graphene oxide films through electroreduction, Chem.
406
Eur. J., 19(15) (2013) 4748–4753
407
ACCEPTED MANUSCRIPT 408
Figure captions
409 410
Figure 1. top panel: cyclic voltammograms of GO reduction in MCl and MCl2
411
solutions (0.1 mol dm−3) (full lines – alkali metals, dashed lines – alkaline earth
412
metals); bottom panel: capacitance increment factors as the function of reduction
413
potential. For the presented measurements no conductive additive is used.
414 415
Figure 2. top panel: cyclic voltammograms of GO reduction in MCl and MCl2
416
solutions (0.1 mol dm−3) (full lines – alkali metals, dashed lines – alkaline earth
417
metals); bottom panel: capacitance increment factors as the function of reduction
418
potential. GO films contained Vulcan XC-72R as conductive additive.
419 420
Figure 3. (a) evolution of the capacitive response of GO upon reduction in KCl
421
solution at different reduction potentials (Ered; no conductive additive); (b) Maximum
422
specific capacitances (Cspec) of GO (corresponding to capacitances measured at Ered
423
which corresponds to the maximum CIF) and Cspec measured after reduction at −1.5
424
V in the absence of conductive additive; (c) the same as for (b) but with conductive
425
additive in the GO film.
426 427
Figure 4. Optimized structures of alkali and alkaline earth metal atoms interacting
428
with epoxy-graphene. Calculated adsorption energies (in eV per atom) are included
429
as well as M‒O and O‒C bond lengths (in Å).
ACCEPTED MANUSCRIPT
Highlights
Electrochemical reduction of graphene oxide is sensitive to cations GO is reduced at more negative potentials in alkali metal chloride solutions Capacitance of graphene oxide is maximized when reduced at optimal potential Capacitance is tuned between 65 and 130 F g−1. The effect is due to the interactions of cations with oxygen functional groups