Chemical Engineering Journal 373 (2019) 1212–1222

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: [Link]/locate/cej

Insight into the ultrasonication of graphene oxide with strong changes in its T
properties and performance for adsorption applications
Giang T.T. Lea, Narong Chanlekb, Jedsada Manyamc, Pakorn Opaprakasita, Nurak Grisdanurakd,
⁎
Paiboon Sreearunothaia,
a
Sirindhorn International Institute of Technology, Thammasat University, Pathum Thani 12121, Thailand
b
Synchrotron Light Research Institute, 111 University Avenue, Muang Nakhon Ratchasima, 30000, Thailand
c
NANOTEC, National Science and Technology Development Agency (NSTDA), 111 Thailand Science Park, Khlong Luang, Pathum Thani 12120, Thailand
d
Faculty of Engineering, Thammasat University, Pathum Thani 12121, Thailand

H I GH L IG H T S G R A P H I C A L A B S T R A C T

• Simple sonication can lead to strong

decrease in graphene oxide(GO) sorp-
tion capacity.
• Observed across all test species: ele-
mental cation, cationic, anionic dye
molecules.
• Sonication causes GO’s various oxygen
groups to be removed with increased
C/O ratios.
• Increased defects on sonicated GO’s
basal plane also decrease π-π attrac-
tions.
• Mechanism based on radicals gener-
ated from sonication react with GO's
facile oxygen proposed.

A R T I C LE I N FO A B S T R A C T

Keywords: A sonochemical approach is widely used as a routine protocol for graphene oxide (GO) preparation prior to
Ultrasonication adsorption applications, but there is a limited understanding of its effects on the GO sorption ability. Herein,
Graphene oxide using different types of sorbates, which are cation, cationic and anionic dye, the changes in sorption capacity
Adsorption properties along with the characteristics of sonicated GO, are shown. The ultrasound treatment unexpectedly removes
oxygen functional groups from GO with a concomitant decrease in GO adsorption capacity, rather than an
increase in its adsorption as expected from the use of ultrasound in dispersing or exfoliating GO. The C/O ratio
increases from 2.1 in pristine GO up to 2.5 in sonicated GO. The loss of oxygen functional groups is also ac-
companied by an increase in disorder and defects on GO sheets, along with the weakening of π-π interaction. The
pathways of GO’s structural and functional groups changed by ultrasound are proposed, based on the generation
of OH· radicals from ultrasonic cavitation. These radicals can react with the facile oxygenated groups of GO, such
as those of epoxide, hydroxyl, and carboxyl, leading to the partial removal of these groups and the breakage of
GO sheets. GO’s adsorption capacities for cation and different dyes can decrease up to 75% after sonication,
which is in agreement with the above changes in GO’s characteristics. The effects are readily observed even in
mild sonication such as using an ultrasonic bath, signifying their importance in many applications employing
ultrasonication to disperse GO.

⁎
Corresponding author.
E-mail address: paiboon_sree@[Link] (P. Sreearunothai).

[Link]
Received 3 March 2019; Received in revised form 29 April 2019; Accepted 17 May 2019
Available online 18 May 2019
1385-8947/ © 2019 Elsevier B.V. All rights reserved.
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

1. Introduction of the ultrasonic effects on the adsorption properties of GO would be

interesting. In this study, two different systems are used, which are
Graphene oxide (GO) is a strongly oxygenated form of graphite, cesium as a representative of an elemental cation, and a cationic dye
with a layered structure decorated by numerous functional groups, such (methylene blue) and anionic dye (methyl orange) as molecular organic
as hydroxyl, carboxyl, epoxide, ketone, and phenolic [1]. The ultra- pollutants. The employment of different sorbates is believed to give
sonication treatment of GO is a simple method to exfoliate the multi- insight into the changes in GO characteristics before and after the ex-
layered GO into a few-layered or a single-layered GO. Ultrasonication is posure to ultrasound.
widely applied in the sample preparation of GO for various applica- Ultrasonic treatment is generally employed in two ways: a probe-
tions, as the single-layered GO may possess a larger active surface area type or a tank-type sonicator (ultrasonic bath). Generally, the probe-
than the unsonicated GO. sonicator possesses a higher localized intensity than the tank-type ul-
Soltani et al. [2] and Yang et al. [3] reported a rapid synthesis of GO trasonic bath. The tank-type sonicator is commonly employed in most
by applying ultrasonication to accelerate oxidation and decrease the use works in the literature for the treatment of GO. However, to measure
of hazardous chemicals during synthesis. Post-synthesis ultrasound ir- the effects of ultrasound at both low and high power on GO and its
radiation also leads to changes in the mechanical properties of GO, as sorption performance, both the tank-type and the probe-type sonicators
reported by Ye et al. [4], with a higher fracture strength and toughness were employed in this study. The use of the probe-type sonicator can be
found in the short-time ultrasonicated GO in mild conditions. Gen- found in many syntheses of graphene oxide composites or graphene-like
erally, the use of an ultrasound treatment at different steps during the materials, as it provides better homogeneity [20,21], while the ultra-
synthesis process can result in different sample properties. Esmaeili sonic bath is commonly employed to disperse GO in a typical aqueous
et al. [5] reported the enhancement of oxygen-containing groups and suspension preparation [22–24].
the surface area of GO when applying ultrasound in the oxidation step In this study, the effects of the sonochemical approach on GO
of Hummers’ method. In the study of Soltani et al. [6], ultrasound was properties are systematically investigated for both the characteristics
applied during the mixing steps of the synthesis by Hummers’ method and the sorption capacity of the original and sonicated GO, obtained
and found to enhance the sorption efficiency of GO for 2-chlorophenol, from the two types of sonicators with varying sonication time. The
due to a greater number of oxygen functional groups and a larger properties of GO before and after ultrasonication are characterized by
surface area. In contrast, GO ultrasonicated after completing all syn- various spectroscopic techniques: Raman spectroscopy, X-ray photo-
thetic steps (from oxidation to drying in a vacuum oven) and re-dis- electron spectroscopy (XPS), UV–vis absorbance spectroscopy, and the
solved in water, may possess a lower number of oxygen functional distinct changes in sorption capacities of sonochemical treated and
groups [7–9]. Soltani et al. [10] proposed a sonochemical reduction of untreated GO are used to observe the properties of GO.
graphene oxide by an ultrasound treatment and reported an increase of
the C/O ratio from 1.1 of the initial GO to 4.2 for the as-produced re-
duced GO, but no effect on the adsorption properties was reported. 2. Materials and methods
Several studies also applied ultrasonication in combination with redu-
cing agents such as ascorbic acid [11], hydrazine hydrate [12], or 2.1. Synthesis and sonication treatment of graphene oxide
ethylene glycol [13] to enhance GO’s reduction.
GO has been widely used as a potential sorbent for heavy metals, Graphene oxide (GO) was synthesized from graphite flakes via the
radionuclides, dyes, and organic pollutants because GO possesses a high typical Hummers' method [25]. Briefly, 3.0 g of graphite powder (Acros
number of oxygen functional groups as well as partial aromatic and Organics, synthetic origin, greater than 99% purity) and 1.5 g of NaNO3
layered structures. In the literature, the use of ultrasound for GO pre- (LOBA, analytical reagent, 99%) were added to 138 mL of concentrated
paration before the sorption process has been widely employed, since it H2SO4 (Carlo Erba, ISO for analysis, 98%) and stirred in an ice bath
enhances the surface area and the adsorption sites due to the exfoliation (less than20 °C) for 1 h to exfoliate graphite into separated layers. 18 g
and fragmentation of GO [14–16]. However, there has been no sys- of KMnO4 (Carlo Erba, ACS for analysis, 99%) was added slowly under
tematic investigation of the sorption ability of GO, before and after continuous agitation. The mixture was kept at 35 °C for 30 min, until it
sonication. In most of the reported studies, GO was treated by ultra- became a gray, thick slurry. Deionized water (DI) (276 mL) was then
sonication prior to the sorption process, and only the sorption capacity gradually added to counter the exothermic heat that was released. The
of the ultrasonicated GO was explored, without evaluating the perfor- mixture was then put on a hot plate at 80 °C for 30 min. 50 mL of H2O2
mance of the unsonicated GO. solution (Sigma-Aldrich, 30% w/w) mixed with 420 mL DI was then
Several works in the literature have reported conflicting perfor- added slowly to reduce the residual KMnO4. The resultant yellow sus-
mances of GO’s adsorption properties. For instance, in the work of Sitko pension was subsequently rinsed with concentrated HCl (Carlo Erba,
et al. [17], they did not apply any sonication to the synthesized GO. analytical reagent, 37%) to remove the residual SO42−, followed by
They obtained high sorption capacities of GO for Cu2+ and Pb2+, which rewashing with DI centrifugation several times at 4000 rpm. The final
were 294 mg/g and 1119 mg/g, respectively, The sorption capacities pH of GO suspension was around 4. A vacuum oven was used to dry the
were much lower in the experiments of Yang et al. [14] (46.6 mg/g for synthesized GO at 40 °C to decrease the amount of trapped water in the
Cu2+), Tan et al. [18] (72.6 mg/g for Cu2+), and Peng et at. [19] GO structure.
(180 mg/g for Pb2+) when the ultrasound treatment was conducted on The as-prepared GO after undergoing the synthesis process was
GO before sorption. The low sorption capacities of GO for Ni2+ ion dissolved in deionized water and diluted to 4 mg/mL. The resulting GO
were also reported in the studies of Najafi et al. [16] (38.6 mg/g) and suspension was subjected to various sonication times, using two dif-
Tan et al. [18] (62.3 mg/g) with the presence of an ultrasonic bath ferent techniques: ultrasonic bath (UB) and probe-sonicator (PS). For
during sample preparation. The ultrasonication in those works was the ultrasonic bath (VGT-2013QTD, GT Sonic Co., 40 kHz, 360 W, and 9
employed without mentioning a specific purpose, but merely used as a L of water medium in the bath), the GO suspension was placed in a 50-
routine protocol to exfoliate GO. mL glass beaker immersed in the water medium. For the probe-soni-
Considering the deviation of sorption results even in just a simple cator (VCX130, VibraCell, Sonics and Materials Inc., 20 kHz, 130 W),
system of copper and lead ions shown above, a systematic investigation the probe was immersed directly into 50 mL of GO suspension. Two sets
of various time treatments were prepared, denoted as UB0–UB120

1213
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

(treated by ultrasonic bath from 0 to 120 min) and PS0–PS120 (treated respectively. For reference, stock solutions without added GO were
by probe-sonicator from 0 to 120 min). All ultrasonic processes were prepared and underwent the same process, to eliminate the interference
performed in an ice/water bath with the temperature kept below 20 °C of the employed vials; there was a negligible adsorption by the vials
to avoid elevated heating, which may lead to the thermal reduction of with all sorbates. All sorption experiments were conducted in triplicate,
GO. using three different vials for each sample. The average values were
calculated and shown in the results section. In order to obtain adsorp-
2.2. Characterization of pristine and sonicated GO tion isotherm, the sorption experiments were done at various sorbate
concentrations with similar condition as above-mentioned, and the
The GO sample for the scanning electron microscope (SEM) mea- results between Qe (amount of sorbate adsorbed on sorbent, mg/g) and
surements was prepared by immersing a silicon wafer into a diluted GO Ce (amount of sorbate remaining in solution, mg/L) were plotted and
suspension (about 1 mg GO/20 mL DI water) followed by drying at also fitted by an adsorption isotherm model. The pH-dependent sorp-
room temperature to yield a randomly deposited GO sheets on a silicon tion capacities of pristine GO and sonicated GO were also evaluated at
wafer. The SEM images of these GO sheets were acquired using an three different pHs: 4, 7, 10, respectively. HCl 0.1 M and NaOH 0.1 M
S3400N SEM (Hitachi High-Tech Europe) operating at 2 kV with a were used for pH adjustment. The three sorbates were prepared as
working distance of about 10 mm. The UV-Vis absorbance of all GO following concentrations for pH-dependent tests: 1 mmol/L for CsCl and
samples in aqueous solution (50 mg/L) was measured using a UV-Vis 5 mg/L for methylene blue and methyl orange, respectively.
spectrometer (Genesys 10S, Thermo Fisher Scientific). The average size
of each GO sample was determined by a dynamic light scattering (DLS) 3. Results and discussion
technique (Partica LA-950 V2). GO samples before and after the ultra-
sound treatment were separated from the suspension by centrifugation 3.1. Size of GO sheets after ultrasonication
at 6000 rpm and dried in a vacuum oven at 40 °C for the following
spectroscopic characterizations: XPS, and FT-Raman. The X-ray pho- Fig. 1 shows the SEM images of GO sheets before and after 120 min
toelectron spectroscopy (XPS) measurements were performed on a PHI of probe sonication. The strong decrease in the size of GO sheets from
5000 Versa Probe IIXPS system (ULVAC-PHI, Japan) with Al K radiation several tens of micrometers in a pristine GO to a few micrometers in the
and 100 µm2 beam size. The data were processed using MultiPak soft- sonicated sample can be clearly observed. This decrease in the size of
ware. Raman spectra of all GO samples were measured by a Fourier- GO with an increase in the ultrasonication time is also in agreement
transform spectrometer Perkin Elmer system 2000 using 532 nm laser with those reported by Golcaves et al. and Kim et al. [7,26]. To obtain
excitation. the quantitative analysis for the GO size after sonication, dynamic light
scattering (DLS) measurements were applied to determine the size
2.3. Adsorption experiments distribution and average size of each sample. Although the size ob-
tained from the DLS technique is the diameter of spherical particles,
Pristine GO and sonicated GO samples were investigated for Cs (I) DLS analysis is widely used to indicate the GO size distribution or the
sorption by dispersing sonicated GO into CsCl (Merck, trace metal basis, trend [7,26–28]. Golcaves et al. [7] suggested that the bent structure of
99.99%) solution with the final concentrations of 2 g GO/L and 1 mmol GO was more stable than the planar structure, and DLS measurement
Cs/L. For dye sorption experiments, methylene blue (Unilab, laboratory was still valid for GO sheets.
reagent, 95%) and methyl orange (LabChem, laboratory reagent, 95%) As shown in Fig. 2, the sonication treatment significantly decreases
were prepared at 5 mg/L, and GO was dispersed at 0.25 g/L for each the average size of GO, from about 33 µm of the initial pristine GO, to
sample. The samples were incubated for 24 h at room temperature 15 µm for UB120 and 4.5 µm for PS120, after 120 ultrasonicated min-
(25 °C) on a shaker with an agitation speed of 200 rpm. The samples utes. The use of the probe-sonicator exhibits strong effects on the
were then centrifuged, and the clear supernatants were collected (our breakage of GO sheets, with the size decreasing from 33 to 12 µm
preliminary studies indicated that the sorption process could reach within only 5 min. After 15 min, the average size of sample PS15 is
equilibrium within 1 h). Cs concentrations before and after sorption about 5 µm and starts to become stable until 120 min without a re-
were measured using Inductively-Coupled Plasma (ICP) spectroscopy markable change in size. In contrast, the size of sample UB5 treated by
(Perkin Elmer Optima 8000, emission mode). Dye concentrations were an ultrasonic bath for 5 min decreases from 30 to 15 µm after 120 min,
measured using UV–Vis spectroscopy. The λmax for determining the dye revealing that an ultrasonic bath can decrease the size of GO but with
concentrations was selected to be the peak absorbance at around not as strong an effect as that of using the probe-sonicator. The frag-
665 nm and 505 nm for methylene blue (MB) and methyl orange (MO), mentation of GO was caused by the shear force created by ultrasonic

Fig. 1. SEM images of a) pristine GO and b) PS120, exhibiting the strong decrease in the size of GO sheets after 120 min of probe sonication.

1214
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

Fig. 2. Average size of GO sheets before and after different sonication times.

Fig. 4. Cs sorption capacity of GO that has been sonicated for different times
waves. The bubble cavitation led to the shear distortion of GO sheets,
using an ultrasonic bath or a probe sonicator. Cs initial concentration of 1 mM
weakening the van der Waals interactions. The shear force was strong
or 133 ppm.
enough to break the GO sheets [3].
Fig. 3 shows the size distribution of the two series, UB5–UB120 and
PS5–PS120. The pristine GO exhibits a large size distribution, varying molecules. Both the probe sonicator and the ultrasonic bath show de-
from a few micrometers up to 50 µm. The use of the probe-sonicator creasing GO adsorption capacities across all the test pollutants, but the
helps to narrow the size distribution for GO sheets, with the size ran- probe sonicator has a stronger effect. In the case of Cs+, the sorption
ging from a few hundred nanometers to 10 µm. The more homogeneous capacity decreases from about 16 mg/g in the pristine GO to less than
sizes of the sonicated GO obtained here are in line with several similar 10 mg/g after 120-min probe sonication, equivalent to a 34% decrease
procedures in the literature [7,29,30]. In this experiment, the sonica- in the Cs sorption capacity. In the UB-GO, the trend is also the same.
tion treatment by an ultrasonic bath is less efficient for the breakage of UB-GO exhibits a decreased Cs+ sorption capacity. Even 15 min in an
GO sheets. The size distribution of sample UB120 is still large, as it ultrasonic bath can decrease the Cs+ sorption capacity by about 12%.
varies from several µm up to 45 µm, which suggests that there are still The decrease in the cation adsorption is quite unexpected as soni-
numerous large sheets in UB120 after a long time of the ultrasound cation was shown to help in exfoliating GO structures, as well as de-
treatment. creasing the GO’s size, and thus, should have promoted an increase in
the surface area and the number of active sites for sorption. The de-
3.2. Sorption capacity of GO for Cs cation and dyes crease in the elemental Cs+ adsorption capacity indicates that there
may be a loss of some cation-binding functional groups during the so-
Figs. 4 and 5 show the decline in sorption capacity of sonicated GO nication process, leading to a significant decrease in the sorption ca-
for Cs+ and organic dyes along with an increase in sonication time. It pacity of the sonicated GO. The adsorption of cations onto GO has been
can be clearly observed that the longer the sonication time, the smaller shown to be due to various functional groups on GO sheets, such as the
the adsorption capacities for both the elemental cation and the dye carboxyl, hydroxyl, and epoxide groups [15,32–35]. Compton et al.

Fig. 3. Size distribution of UB and PS series. The probe sonicator yields a smaller size and narrower distribution of GO due to the more intense ultrasound generated
by the probe sonicator [31].

1215
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

numerous hydroxyl groups in the GO structure. In the case of MB, the

H-bonds can be created between the N or S heteroatoms in the aromatic
ring of MB and the –OH group of GO, as shown in Fig. 6. In MO, it can
also form H-bonds via the oxygen atoms in –SO3 with the –OH group of
GO [37,38]. Moreover, the GO surface is negatively charged over a
wide pH range [33,39]. Hence, it can also adsorb MB through elec-
trostatic interactions, which are attractions between species with op-
posite electrical charges. In addition, the sorption pathway of MB and
MO via π−π interactions may possess a slightly different strength.
While MB has three adjacent, joined rings of atoms oriented on the
same plane, the two aromatic rings of MO are separated by an N]N
bond, leading to out-of-plane bending. Hence, the π−π interactions
between MO and the aromatic rings of the GO sheets may be weaker
than that of MB. Therefore, the sorption capacity of GO for MB is much
higher than for MO.
The effects of sonication on the GO’s adsorption performance can
also be observed across a range of sorbate concentrations. This can be
illustrated as the effects of sonication on the GO’s adsorption isotherms.
Three different types of GO suspensions (pristine GO, UB15, PS15) are
incubated with various concentrations of the sorbates. The sonication
Fig. 5. Methylene blue (cationic dye) and methyl orange (anionic dye) sorption time of 15 min was selected here to demonstrate the effect of a typical
capacities of GO before and after sonication. The MB sorption decreases steadily sonication time employed in literature, which usually ranges from 5 to
with increasing sonication time for both the ultrasonic bath and the probe so- 30 min [14,16,18–20]. This sonication time can already have sig-
nicator. Initial concentration = 15 µM for each sorbate. (For interpretation of
nificant effects on the adsorption isotherms and adsorption perfor-
the references to colour in this figure legend, the reader is referred to the web
mance of GO. Fig. 7 shows the adsorption isotherms of GO and soni-
version of this article.)
cated GO on the three sorbates. It can be seen that sonication both the
UB and PS as short as 15 min show clearly the decrease in the ad-
[36] have also reported a decrease of carboxyl groups along the edges sorption performance of GO for all the sorbates. The isotherms can also
of GO sheets after a sonochemical treatment. This decrease in carboxyl be described quite well using Langmuir isotherm, which is expressed as
groups is in agreement with the decrease in the Cs+ sorption observed Qe = (Qm kCe )/(1 + kCe ) where Ce is the equilibrium sorbate concentra-
here as a carboxyl group is known to have a high cation-binding affi- tions in the solution, Qe is the equilibrium amount of sorbate on the
nity. The elimination of some of these groups is indeed observed and is sorbent, Qm is the maximum sorption capacity, and k is the affinity
discussed in the XPS and UV–vis analyses below. constant. Table 1 shows the parameters of the Langmuir model derived
The ability of GO to adsorb organic dye molecules both cationic and from the fitted isotherms of GO and sonicated GO for various sorbates.
anionic is also observed to decrease substantially after sonication as It can be seen that all of the sonicated GO exhibit a decline in the
shown in Fig. 5. Pristine GO shows an adsorption capacity for the ca- adsorption capacity for all the three sorbates. GO treated by a probe-
tionic dye MB of about 19.6 mg/g (61 µmol/g), and about 9.4 mg/g sonicator (PS15) has a stronger decrease in sorption capacity than that
(30 µmol/g) for the anionic dye MO. These values decrease sub- of GO treated by ultrasonic bath (UB15). It should also be noticed that
stantially to 11 mg/g (35 µmol/g) for MB, and 2.4 mg/g (7 µmol/g) for the sorption of the anionic dye, MO, displayed the strongest percentage
MO in the 120-minute PS-GO. decrease in the maximum capacity amongst the three employed sor-
Various possible sorption modes of the dyes onto GO are illustrated bates.
in Fig. 6. Both MB and MO have π-conjugated backbones which likely In addition, there are differences between the decline in sorption
can be adsorbed onto GO through the π−π bond interactions between capacities of sonicated GO for Cs, MB, and MO. While, MO sorption
the carbon basal plane of GO and the aromatic rings of the dyes. Both exhibits the strongest decrease of about 75% from pristine GO to PS120,
MB and MO can also interact with GO through hydrogen bonding with those of MB and Cs show a 44% and 34% decline in adsorption

Fig. 6. Different sorption mechanisms of GO for methylene blue and methyl orange. The diagram is not meant to display an exact alignment, but to illustrate various
possible modes of interactions. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

1216
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

Fig. 7. Adsorption isotherms of a) Cs; b) methylene blue and c) methyl orange onto pristine GO and sonicated GO (UB15, PS15). (For interpretation of the references
to colour in this figure legend, the reader is referred to the web version of this article.)

Table 1 and epoxide. In addition, the hydroxyl and epoxide groups on GO sheets
Parameters of Langmuir model for Cs, MB and MO adsorption on GO before and are also known to be relatively facile for removal [40–42]. The elim-
after sonication. ination of oxygen functional groups such as the hydroxyl and carboxyl,
Sorbate Sorbent Langmuir isotherm and the change in aromaticity of GO are confirmed by other spectro-
scopic techniques in the following sections.
Qm (mg/g) K (L/mg) R2 Furthermore, the effects of GO sorption capacities at different pH
were also investigated and compared between pristine GO and PS-GO,
Cs GO 113 4.3× 10−3 0.983
UB15 102 3.9× 10−3 0.987 as shown in Fig. 8. The higher sorption capacities of GO toward Cs+
PS15 89 2.4× 10−3 0.992 and the MB with increasing pH are likely due to the deprotonation of
MB GO 139 5.9× 10−2 0.967 various functional groups in GO [33,43,44]. On the contrary, there is
UB15 122 5.6× 10−2 0.975 only a very small decline in the amount of MO adsorbed onto GO along
PS15 106 5.1× 10−2 0.968 with pH changes. As mentioned above, the sorption mechanism of MO –
MO GO 64 2.3× 10−2 0.975 as an anionic dye–is mainly based on π−π interactions with GO rather
UB15 52 2.7× 10−2 0.963 than the oxygen functional groups, hence the deprotonation occurring
PS15 38 2.4× 10−2 0.959 at high pH does not exhibit a significant effect on the sorption of MO.
After sonication. In all these cases, sonicated GO consistently exhibits
lower sorption ability than that of pristine GO at all pHs due to the
capacities after sonication, respectively. This indicates that sonication
removal of oxygen functional groups as well as the sheet fragmentation
may affect functional groups or sites on GO’s structures that are re-
caused by ultrasonication reducing the aromatic backbone. The em-
sponsible for MO adsorption more strongly than the parts on the GO’s
ployment of ultrasound for GO preparation or dispersions should thus
structure that are responsible for the cation sorption. Therefore, the MO
be carefully considered especially in adsorption applications where
adsorption capacity decreases the most after sonication, followed by
oxygen functional groups and pi-pi interactions can play important
that of the cationic dye MB, and the elemental cation Cs. This suggests
roles.
that although sonication may remove some of GO’s cation binding
groups such as carboxyl, the sonication may affect more of the groups
that are capable of making hydrogen bonds such as hydroxyl, phenols,

Fig. 8. pH-dependent sorption capacities of GO toward cesium, methylene blue and methyl orange. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)

1217
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

3.3. Raman spectra about 45%, with 10% remaining for the C = O species (such as car-
bonyl, carboxyl, ketone). The amount of the C-O species slightly de-
Raman spectroscopy is a powerful characterization technique that creases after the sonication treatment, and exhibits the lowest amount
can give insight into the structural properties of carbon-based materials in the sample PS120, at about 31%. However, there is no significant
[45]. The Raman spectra of pristine GO and all ultrasonicated GO change observed in the C]O species. A loss of oxygen-containing
samples contain two prominent bands. The first band, localized around groups during sonication is also observed in the study of Nam et al. [9],
1350 cm−1, is called the D-mode band. This band arises from the pre- using the changes in the zeta potential of GO suspension before and
sence of disordered structures in the sp2-hybridized carbon network of after a sonochemical approach. Fewer oxygen groups lead to a less
GO, such as zone boundaries and inter-vacancies. The second band, the negative zeta potential of GO in the solution.
G-mode band appearing at 1590 cm−1, is the primary in-plane vibra- In addition, the C/O ratio increases from 2.1 in pristine GO to 2.4 in
tion mode of the sp2 carbon atoms [46]. PS120 (see Table 2), revealing the elimination of some oxygen func-
With increasing sonication time, GO shows a continuous red-shift of tional groups after the ultrasound treatment. An increase in the C/O
the G band, from 1599 cm−1 in the pristine GO to 1590 cm−1 in PS120 ratio is also observed in the study of Goncalves et al. [7] which showed
(see Fig. 9) and 1596 cm−1 in UB120, which suggest a noticeable that ultrasonication can cause a decrease or conversion of the C–O–C/
transformation in the GO lattice structure. This phenomenon can be C–OH into either C–H or C]O species on GO sheets, leading to an in-
assigned to the change in the number of layers [47] when multilayer crease in the C/O ratio.
GO is exfoliated to be a few-layer or mono-layer GO. Martinez et al. [8],
investigating the effects of ultrasonic cavitation, found that the number 3.5. UV–Vis absorbance spectra
of layers in the GO sheets can decrease from more than 30 layers to less
than 5 layers after 20 min of the ultrasound treatment. The shift of the G The absorbance of GO before and after ultrasonication treatment
band originates from the tensile strain of GO sheets derived from the was measured by UV–Vis spectroscopy. GO was prepared at a con-
ultrasonic waves; as reported by Huang et al. [48]. The G band was centration of 50 mg/L for each sample.
found to exhibit a significant red-shift with increasing strain. Shang There is a distinct band appearing at 230 nm and a broad shoulder
et al. [49] show an obvious difference between the red-shift trends of appears at 300 nm, which can be assigned to the π- π* transition of the
the high strain and low strain levels exhibited by the G band. Hence, the C]C bond and the n-π* transition of the C]O bond (carbonyl or car-
G band is considered as a relevant component to indicate the tensile boxyl group), respectively [53]. There are no observed peak shifts for
strain in GO sheets. Due to the difference in localized intensity between all sonicated samples. However, the intensity of both UB-GO and PS-GO
the tank-type and probe-type sonicators, the probe-sonicator generates increases remarkably, due to the fragmentation and exfoliation of GO
a higher strain, leading to a more down-shift of the G band in the PS120 sheets, which create more particles in the suspension and absorb more
(1590 cm−1) than that of the ultrasonic bath UB120 (1596 cm−1). light from the excitation source. In addition, the normalized spectra
The ID/IG ratio is an efficient probe to determine the level of dis- exhibited in the insets of Fig. 12 show a slight decrease in the amount of
order in graphene-like materials [50]. The ID/IG of each sample was the C]O bond fraction for all sonicated GO samples, which is in line
calculated by the intensity of the D band and G band, exhibited in with the XPS results. However, PS-GO reveals a faster and stronger
Fig. 6. According to Table 1, an increase of sonication time leads to a decrease than those of UB-GO, as the C]O peak fraction immediately
decrease in the ID/IG ratio. In a highly disordered carbon system like decreases just after 5 min of probe sonication in the PS5 sample, while
graphene oxide, a decrease in the ID/IG ratio also suggests an increase in the UB-GO series shows a gradually systematic decrease of 300-nm
the number of defects [45]. The appearance of structural defects in
sonicated GO is also reported in several studies [4,8,51]. Bracamonte
et al. [52] found that short-time sonication can create defects, parti-
cularly at the layer edges, while a longer sonication time leads to de-
fects that are created in the bulk.
In PS-GO, the ID/IG ratio drops from 1.23 in pristine GO to 1.13 in
PS120, revealing a high number of defects that have been created by
the ultrasonication, similar to the lowest sorption capacity of PS120,
while in the UB-GO series, the ID/IG ratio does not show a significant
change after sonication. A broadening of the G band is also observed.
For instance, the FWHH of the G band increases from 62 cm−1 in
pristine GO to 66 cm−1 in the sample PS120, which can be ascribed to
defects and to the breaking of the conjugated π system in GO sheets.
The broadened G band suggests that there are more defects and less sp2
hybridization in the GO structure after the ultrasound treatment. These
significant changes in the lattice structure of GO reduce the π-con-
jugation on the GO sheets and contribute to the decrease in the MO and
MB sorption of ultrasonicated GO, as shown in Fig. 4.

3.4. X-ray photoelectron spectroscopy (XPS) spectra

Fig. 10 shows the deconvoluted C1s spectra of various GO samples

using the Gaussian–Lorentzian function. All samples exhibit 3 major
peaks at 284, 286.5, and 288 eV. The peak appearing at 284 eV is as-
signed to C–C and C = C. The peak at 286.5 eV is assigned to C-O
species, and the remaining shoulder at 288 eV corresponds to the C = O
binding energy.
The component fractions of each sample were calculated from the Fig. 9. Raman spectra of various GO samples and the red-shift of the G band in
area of the fitted peaks, displayed in Fig. 11. In pristine GO, the C–C/ sonicated samples. (For interpretation of the references to colour in this figure
C = C and C-O species (such as hydroxyl, ethers, carboxyl, epoxide) are legend, the reader is referred to the web version of this article.)

1218
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

Fig. 10. XPS spectra of different GO samples.

species. These C]O bond species may be assigned to various groups

such as carboxyl or ketone, but carboxyl is more unstable [54,55].
Hence, the decline of C]O here is likely due to carboxyl at the GO sheet
edges, possibly through decarboxylation [40]. This is also in agreement
with the decrease in the cation adsorption capacity of sonicated GO,
which is probably from a decrease in the cation-binding carboxyl
groups from the GO sheet after sonication.
In general, the peak shift from 230 nm to about 255–270 nm can
indicate the formation of reduced GO, as it is assigned to the restoration
of electronic conjugation within the GO sheets [56–58]. Loryuenyong
et al. also suggested that the red-shift of the 230 nm-peak is due to a
decrease in oxygen functional groups and an increase in the aromatic
rings in GO sheets [59]. Soltani et al. [10] observed a peak shift of up to
261 nm for ultrasonicated GO, however, their synthesis process was
conducted at 80 °C. In this study, the oxygen-containing groups are
reduced or eliminated from GO's structure, but without the restoration
of the sp2 aromatic structure, unlike the reduction of GO using high
temperature or chemical reducing agents. Therefore, instead of the
formation of aromatic rings after oxygen atoms leave the GO sheets as
typically observed in graphene oxide reduction [40,60,61], the sheet
Fig. 11. Component fractions of various carbon species in sonicated GO sam- edges or defect edges of sonicated GO are likely terminated by hy-
ples. drogen atoms after the oxygen functional groups leave. These hydrogen
atoms lead to an increase in the C/O ratio (as shown in XPS results) but
Table 2 without the restoration of aromaticity, as observed in both the Raman
Element contents and C/O ratio of GO before and after sonication. and the UV–vis results. The hydrogen atoms or radicals could come
Sample %C %O C/O ratio
from the dissociation of water by the application of ultrasonication.
Combining the results obtained from the above adsorption experi-
GO 67.6 32.4 2.1 ments and characterizations, the changes of GO structure through ul-
UB5 67.7 32.3 2.1 trasonication have been identified. The lower sorption capacities of
UB120 68.9 31.1 2.2
PS5 68.6 31.4 2.2
sonicated-GO toward cationic sorbate imply the decline of various
PS120 70.8 29.2 2.4 oxygen functional groups, while the sorption performance with organic
dyes exhibits the loss of sp2 hybridization. These changes are also

1219
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

Fig. 12. UV-Vis absorbance spectra of a) UB-GO and b) PS-GO.

confirmed by spectroscopic techniques. XPS results demonstrate the dyes. This mechanism is proposed for both PS-GO and UB-GO, but PS-
decrease of oxygen content, especially in the form of C-O groups. In GO may exhibit a higher level of reduction and fragmentation, due to its
addition, Raman spectra together with UV–vis spectra help to indicate higher localized ultrasonic intensity.
the changes in GO lattice structure, with a decline in sp2 and an in-
crease in sp3 hybridization.
4. Conclusions
The trend of decreasing adsorption performance with sonication
time is also very clear and helps to postulate a possible mechanism of
The unexpected effects of ultrasonication on GO’s adsorption
GO’s evolution during sonication process.
properties were observed during the ultrasonic pre-treatment of gra-
It is well known that pure water under ultrasound can dissociate
phene oxide, leading to a decrease of sorption performance. The effects
into OH· and H· radicals, due to the high local pressure and temperature
were observed in as little as 15 min in an ultrasonic bath and became
generated in a collapsing cavitation bubble. These radicals can also
increasingly stronger when using a probe sonicator to disperse GO.
interact and react to form other derived products such as O·, H2O·, and
Cation and organic dyes were used to investigate the changes in sorp-
H2O2 [62]. Therefore, these free radicals are likely the main factors
tion capacity of ultrasonicated GO. These dyes showed similar results
causing the partial reduction of GO, as they can attack the oxygen
that the sorption capacities were significantly reduced after sonication.
atoms contained in the various functional groups of GO and generate
Ultrasonication was shown to cause the partial removal of the oxyge-
different reactions. The proposed mechanisms of GO ultrasonication
nated groups on GO sheets, as well as fragmentation. Ultrasonication
and reactions with free radicals are shown in Fig. 13.
also increases the defects on the basal plane as shown by the XPS,
The attacks of these radicals on various oxygen functional groups
Raman, DLS, and UV–Vis analyses. The decreases in the oxygen groups
follow different reactions. The carboxyl groups, which appear along the
and aromatic backbone decrease GO’s ability to adsorb organic dye
GO sheet edges, could be easily attacked by the OH· radicals, turning
molecules through hydrogen bonding and π-π interactions, resulting in
them into carboxylate radicals and further decomposing to form CO2
a significant decrease in the dye adsorption capacity. There is decreased
[40]. A hydroxyl group can also react with the OH· radical, and the C
adsorption of the elemental cation Cs+, which is adsorbed mainly
atom, on which the hydroxyl group is bonded to, is taken from the GO
through electrostatic interactions. This decreased adsorption indicates
plane and leaves as CO2 [63]. In the case of epoxide, ring opening may
that some of the negatively charged groups on GO’s surface such as
occur to form trans-diols consisting of two hydroxyls, followed by the
carboxyl groups have been partially removed by ultrasonication, but to
attack of OH·, degrading to CO2 as mentioned for the hydroxyl case
a lesser extent compared to the decrease in the dye adsorption capa-
above. The use of ultrasound to assist the ring opening reaction of ep-
cities.
oxide has been reported in several studies [64,65]. The elimination of
The possible pathways for GO’s structural and functional groups
these in-plane groups also causes the formation of disorder and nu-
change by ultrasonication, leading to decreased adsorption capacities,
merous defects on GO sheets, as revealed by Raman spectra, disrupting
are proposed in this paper. Ultrasound is well known to be able to
the conjugated carbon network.
generate cavitation inside an aqueous solution, which can cause water
The breakage of a GO sheet likely happens at the defects generated
molecules to dissociate into H· and OH· radicals. These radicals then
after some groups have been removed. There is also the harsh shear
react with various oxygenated groups on the GO structure, such as
force of the ultrasonic waves. This shear force contributes to the
epoxide, hydroxyl, and carboxyl, leading to the partial removal of these
breaking of the sheet. The process continuously repeats, with more
groups, the breakage of the GO sheets, and disorder or numerous de-
defects generated with increased sonication time, as revealed by the
fects created on GO’s carbon network. The sonochemical treatment of
Raman analysis, leading to fragmentation with smaller GO sheets.
GO provides an understanding on the use of ultrasound for GO sample
During this process, the oxygen-containing groups are also eliminated
preparation, in addition to the perceived advantages of using an ul-
by the ultrasound. The final sample has a lower oxygen content as
trasonic treatment for better dispersion, homogenization, or for ex-
shown by XPS atomic concentrations, with detrimental effects on the
foliation to increase the surface area. GO dispersed by ultrasonication
sample's adsorption properties for the elemental cation and the organic
could be suitable for several applications which require a low oxygen

1220
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

Fig. 13. Schematic diagram illustrating a possible evolution of GO through the ultrasonic treatment. 1, 2, and 3 demonstrate the potential paths for the elimination of
oxygen functional groups. 4 shows the formation of defects after the in-plane groups are eliminated. 5 illustrates the breakage of GO sheets which may start along the
region of oxygen functional groups, after they are removed.

content. This method can also be carefully optimized for applications Appendix A. Supplementary data
which require the presence of numerous oxygen functional groups on
the surface of GO. Applications that require the incorporation of GO in Supplementary data to this article can be found online at https://
their design can be helped by this method. [Link]/10.1016/[Link].2019.05.108.

References
Acknowledgments
[1] M. Acik, C. Mattevi, C. Gong, G. Lee, K. Cho, M. Chhowalla, Y.J. Chabal, The role of
The authors gratefully acknowledge the support from the Center of intercalated water in multilayered graphene oxide, ACS Nano 4 (2010) 5861–5868.
[2] T. Soltani, B.-K. Lee, Low intensity-ultrasonic irradiation for highly efficient, eco-
Excellence in Materials and Plasma Technology (CoE M@P Tech) and friendly and fast synthesis of graphene oxide, Ultrason. Sonochem. 38 (2017)
the Integrated Budget for Research and Innovation for fiscal year 2562, 693–703.
[3] H. Yang, H. Li, J. Zhai, L. Sun, H. Yu, Simple synthesis of graphene oxide using
Thammasat University (Thailand). The partial support from the Center
ultrasonic cleaner from expanded graphite, Ind. Eng. Chem. Res. 53 (2014)
of Excellence in Materials Science, Construction and Maintenance 17878–17883.
Technology, Thammasat University is also gratefully acknowledged. [4] S. Ye, J. Feng, The effect of sonication treatment of graphene oxide on the me-
chanical properties of the assembled films, RSC Adv. 6 (2016) 39681–39687.
G.L. acknowledges the financial support from the Excellent Foreign [5] A. Esmaeili, M.H. Entezari, Facile and fast synthesis of graphene oxide nanosheets
Student (EFS) doctoral scholarship from Sirindhorn International Via bath ultrasonic irradiation, J. Colloid Interface Sci. 432 (2014) 19–25.
Institute of Technology (SIIT), Thammasat University. Authors are [6] T. Soltani, B.-K. Lee, Mechanism of highly efficient adsorption of 2-chlorophenol
onto ultrasonic graphene materials: comparison and equilibrium, J. Colloid
grateful to the SUT-NANOTEC-SLRI joint research facility (BL5.3), Interface Sci. 481 (2016) 168–180.
Synchrotron Light Research Institute (SLRI), Thailand for XPS facility. [7] G. Gonçalves, M. Vila, I. Bdikin, A. de Andrés, N. Emami, R.A.S. Ferreira,

1221
G.T.T. Le, et al. Chemical Engineering Journal 373 (2019) 1212–1222

L.D. Carlos, J. Grácio, P.A.A.P. Marques, Breakdown into nanoscale of graphene S.T. Nguyen, Additive-free hydrogelation of graphene oxide by ultrasonication,
oxide: confined hot spot atomic reduction and fragmentation, Sci. Rep. 4 (2014) Carbon 50 (2012) 3399–3406.
6735. [37] H. Chen, A. Zhong, J. Wu, J. Zhao, H. Yan, Adsorption behaviors and mechanisms of
[8] P. Pérez-Martínez, J.M. Galvan-Miyoshi, J. Ortiz-López, Ultrasonic cavitation effects methyl orange on heat-treated palygorskite clays, Ind. Eng. Chem. Res. 51 (2012)
on the structure of graphene oxide in aqueous suspension, J. Mater. Sci. 51 (2016) 14026–14036.
10782–10792. [38] J. Ma, F. Yu, L. Zhou, L. Jin, M. Yang, J. Luan, Y. Tang, H. Fan, Z. Yuan, J. Chen,
[9] S.-W. Nam, C. Jung, H. Li, M. Yu, J.R.V. Flora, L.K. Boateng, N. Her, K.-D. Zoh, Enhanced adsorptive removal of methyl orange and methylene blue from aqueous
Y. Yoon, Adsorption characteristics of diclofenac and sulfamethoxazole to graphene solution by alkali-activated multiwalled carbon nanotubes, ACS Appl. Mater.
oxide in aqueous solution, Chemosphere 136 (2015) 20–26. Interfaces 4 (2012) 5749–5760.
[10] T. Soltani, B. Kyu Lee, A benign ultrasonic route to reduced graphene oxide from [39] B. Konkena, S. Vasudevan, Understanding aqueous dispersibility of graphene oxide
pristine graphite, J. Colloid Interface Sci. 486 (2017) 337–343. and reduced graphene oxide through Pka measurements, J. Phys. Chem. Lett. 3
[11] A. Abulizi, K. Okitsu, J.J. Zhu, Ultrasound assisted reduction of graphene oxide to (2012) 867–872.
graphene in L-ascorbic acid aqueous solutions: kinetics and effects of various factors [40] M. Acik, G. Lee, C. Mattevi, A. Pirkle, R.M. Wallace, M. Chhowalla, K. Cho,
on the rate of graphene formation, Ultrason Sonochem. 21 (2014) 1174–1181. Y. Chabal, The role of oxygen during thermal reduction of graphene oxide studied
[12] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu, by infrared absorption spectroscopy, J. Phys. Chem. C 115 (2011) 19761–19781.
S.T. Nguyen, R.S. Ruoff, Synthesis of graphene-based nanosheets via chemical re- [41] E. Tegou, G. Pseiropoulos, M.K. Filippidou, S. Chatzandroulis, Low-temperature
duction of exfoliated graphite oxide, Carbon 45 (2007) 1558–1565. thermal reduction of graphene oxide films in ambient atmosphere: infra-red spec-
[13] S. Anandan, A. Manivel, M. Ashokkumar, One-step sonochemical synthesis of re- troscopic studies and gas sensing applications, Microelectron. Eng. 159 (2016)
duced graphene oxide/Pt/Sn hybrid materials and their electrochemical properties, 146–150.
Fuel Cells 12 (2012) 956–962. [42] Y.J. Oh, J.J. Yoo, Y.I. Kim, J.K. Yoon, H.N. Yoon, J.-H. Kim, S.B. Park, Oxygen
[14] S.-T. Yang, Y. Chang, H. Wang, G. Liu, S. Chen, Y. Wang, Y. Liu, A. Cao, Folding/ functional groups and electrochemical capacitive behavior of incompletely reduced
aggregation of graphene oxide and its application in Cu2+ removal, J. Colloid graphene oxides as a thin-film electrode of supercapacitor, Electrochim. Acta 116
Interface Sci. 351 (2010) 122–127. (2014) 118–128.
[15] W. Peng, H. Li, Y. Liu, S. Song, A review on heavy metal ions adsorption from water [43] W. Zhang, C. Zhou, W. Zhou, A. Lei, Q. Zhang, Q. Wan, B. Zou, Fast and con-
by graphene oxide and its composites, J. Mol. Liq. 230 (2017) 496–504. siderable adsorption of methylene blue dye onto graphene oxide, Bull. Environ.
[16] F. Najafi, O. Moradi, M. Rajabi, M. Asif, I. Tyagi, S. Agarwal, V.K. Gupta, Contamin. Toxicol. 87 (2011) 86–90.
Thermodynamics of the adsorption of nickel ions from aqueous phase using gra- [44] X. Mi, G. Huang, W. Xie, W. Wang, Y. Liu, J. Gao, Preparation of graphene oxide
phene oxide and glycine functionalized graphene oxide, J. Mol. Liq. 208 (2015) aerogel and its adsorption for Cu2+ ions, Carbon 50 (2012) 4856–4864.
106–113. [45] A.C. Ferrari, Raman spectroscopy of graphene and graphite: disorder, electron-
[17] R. Sitko, E. Turek, B. Zawisza, E. Malicka, E. Talik, J. Heimann, A. Gagor, B. Feist, phonon coupling, doping and nonadiabatic effects, Solid State Commun. 143 (2007)
R. Wrzalik, Adsorption of divalent metal ions from aqueous solutions using gra- 47–57.
phene oxide, Dalton Trans. 42 (2013) 5682–5689. [46] R. Beams, L. Gustavo Cancado, L. Novotny, Raman characterization of defects and
[18] P. Tan, J. Sun, Y. Hu, Z. Fang, Q. Bi, Y. Chen, J. Cheng, Adsorption of Cu2+, Cd2+ dopants in graphene, J. Phys. Conden. Matter: an Instit. Phys. J. 27 (2015) 083002.
and Ni2+ from aqueous single metal solutions on graphene oxide membranes, J. [47] L.M. Malard, M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, Raman spectroscopy
Hazard. Mater. 297 (2015) 251–260. in graphene, Phys. Rep. 473 (2009) 51–87.
[19] W. Peng, H. Li, Y. Liu, S. Song, Comparison of Pb(Ii) adsorption onto graphene [48] M. Huang, H. Yan, C. Chen, D. Song, T.F. Heinz, J. Hone, Phonon softening and
oxide prepared from natural graphites: diagramming the Pb(Ii) adsorption sites, crystallographic orientation of strained graphene studied by Raman Spectroscopy,
Appl. Surf. Sci. 364 (2016) 620–627. PNAS 106 (2009) 7304–7308.
[20] F.-F. Ma, D. Zhang, T. Huang, N. Zhang, Y. Wang, Ultrasonication-assisted de- [49] J. Shang, et al., Effect of folded and crumpled morphologies of graphene oxide
position of graphene oxide on electrospun poly(Vinylidene Fluoride) membrane and platelets on the mechanical performances of polymer nanocomposites, Polymer 68
the adsorption behavior, Chem. Eng. J. 358 (2019) 1065–1073. (2015) 131–139.
[21] D. Huang, J. Wu, L. Wang, X. Liu, J. Meng, X. Tang, C. Tang, J. Xu, Novel insight [50] A. Kaniyoor, S. Ramaprabhu, A Raman Spectroscopic investigation of graphite
into adsorption and co-adsorption of heavy metal ions and an organic pollutant by oxide derived graphene, AIP Adv. 2 (2012) 032183.
magnetic graphene nanomaterials in water, Chem. Eng. J. 358 (2019) 1399–1409. [51] Y. Li, R. Umer, Y.A. Samad, L. Zheng, K. Liao, The effect of the ultrasonication pre-
[22] R. Zare-Dorabei, S.M. Ferdowsi, A. Barzin, A. Tadjarodi, Highly efficient simulta- treatment of graphene oxide (Go) on the mechanical properties of go/polyvinyl
neous ultrasonic-assisted adsorption of Pb(Ii), Cd(Ii), Ni(Ii) and Cu (Ii) ions from alcohol composites, Carbon 55 (2013) 321–327.
aqueous solutions by graphene oxide modified with 2,2′-dipyridylamine: central [52] M.V. Bracamonte, G.I. Lacconi, S.E. Urreta, L.E.F. Foa Torres, On the nature of
composite design optimization, Ultrason. Sonochem. 32 (2016) 265–276. defects in liquid-phase exfoliated graphene, J. Phys. Chem. C 118 (2014)
[23] Y. Gao, Y. Li, L. Zhang, H. Huang, J. Hu, S.M. Shah, X. Su, Adsorption and removal 15455–15459.
of tetracycline antibiotics from aqueous solution by graphene oxide, J. Colloid [53] B. Haghighi, M.A. Tabrizi, Green-synthesis of reduced graphene oxide nanosheets
Interface Sci. 368 (2012) 540–546. using rose water and a survey on their characteristics and applications, RSC Adv. 3
[24] C.J. Madadrang, H.Y. Kim, G. Gao, N. Wang, J. Zhu, H. Feng, M. Gorring, (2013) 13365–13371.
M.L. Kasner, S. Hou, Adsorption behavior of edta-graphene oxide for Pb (Ii) re- [54] A. Ganguly, S. Sharma, P. Papakonstantinou, J. Hamilton, Probing the thermal
moval, ACS Appl. Mater. Interfaces 4 (2012) 1186–1193. deoxygenation of graphene oxide using high-resolution in situ X-ray-based spec-
[25] W.S. Hummers, R.E. Offeman, Preparation of graphitic oxide, J. Am. Chem. Soc. 80 troscopies, J. Phys. Chem. C 115 (2011) 17009–17019.
(1958) 1339. [55] L. Li, X. Yao, H. Li, Z. Liu, W. Ma, X. Liang, Thermal stability of oxygen-containing
[26] S.G. Kim, S.H. Wang, C.M. Ok, S.Y. Jeong, H.S. Lee, Lateral diffusion of graphene functional groups on activated carbon surfaces in a thermal oxidative environment,
oxides in water and the size effect on the orientation of dispersions and electrical J. Chem. Eng. Jpn. 47 (2014) 21–27.
conductivity, Carbon 125 (2017) 280–288. [56] X. Zhu, Q. Liu, X. Zhu, C. Li, M. Xu, Y. Liang, Reduction of graphene oxide via
[27] D. Joung, V. Singh, S. Park, A. Schulte, S. Seal, S.I. Khondaker, Anchoring ceria ascorbic acid and its application for simultaneous detection of dopamine and as-
nanoparticles on reduced graphene oxide and their electronic transport properties, corbic acid, Int. J. Electrochem. Sci. 7 (2012) 5172–5184.
J. Phys. Chem. C 115 (2011) 24494–24500. [57] B. Li, X. Zhang, P. Chen, X. Li, L. Wang, C. Zhang, W. Zheng, Y. Liu, Waveband-
[28] E.J. Frankberg, L. George, A. Efimov, M. Honkanen, J. Pessi, E. Levänen, Measuring dependent photochemical processing of graphene oxide in fabricating reduced
synthesis yield in graphene oxide synthesis by modified hummers method. graphene oxide film and graphene oxide–Ag nanoparticles film, RSC Adv. 4 (2014)
Fullerenes, Nanotubes, Carbon Nanostruct. 23 (2015) 755–759. 2404–2408.
[29] X. Qi, T. Zhou, S. Deng, G. Zong, X. Yao, Q. Fu, Size-specified graphene oxide sheets: [58] D. Li, M.B. Müller, S. Gilje, R.B. Kaner, G.G. Wallace, Processable aqueous disper-
ultrasonication assisted preparation and characterization, J. Mater. Sci. 49 (2014) sions of graphene nanosheets, Nat. Nanotechnol. 3 (2008) 101.
1785–1793. [59] V. Loryuenyong, K. Totepvimarn, P. Eimburanapravat, W. Boonchompoo, A. Buasri,
[30] C. Cai, N. Sang, Z. Shen, X. Zhao, Facile and size-controllable preparation of gra- Preparation and characterization of reduced graphene oxide sheets via water-based
phene oxide nanosheets using high shear method and ultrasonic method, J. Exp. exfoliation and reduction methods, Adv. Mater. Sci. Eng. 2013 (2013) 5.
Nanosci. 12 (2017) 247–262. [60] S. Pei, H.-M. Cheng, The reduction of graphene oxide, Carbon 50 (2012)
[31] R. Durge, R.V. Kshirsagar, P. Tambe, Effect of sonication energy on the yield of 3210–3228.
graphene nanosheets by liquid-phase exfoliation of graphite, Procedia Eng. 97 [61] A. Bagri, C. Mattevi, M. Acik, Y.J. Chabal, M. Chhowalla, V.B. Shenoy, Structural
(2014) 1457–1465. evolution during the reduction of chemically derived graphene oxide, Nat. Chem. 2
[32] A. Gopalakrishnan, R. Krishnan, S. Thangavel, G. Venugopal, S.-J. Kim, Removal of (2010) 581.
heavy metal ions from pharma-effluents using graphene-oxide nanosorbents and [62] L. de Castro, F.P. Capote, Chapter 1 Introduction: Fundamentals of Ultrasound and
study of their adsorption kinetics, J. Ind. Eng. Chem. 30 (2015) 14–19. Basis of Its Analytical Uses. In Techniques and Instrumentation in Analytical
[33] P. Kaewmee, J. Manyam, P. Opaprakasit, G.T. Truc Le, N. Chanlek, Chemistry, Elsevier, 2007 pp. 1 34.
P. Sreearunothai, Effective removal of cesium by pristine graphene oxide: perfor- [63] C. Yu, B. Zhang, F. Yan, J. Zhao, J. Li, L. Li, J. Li, Engineering nano-porous graphene
mance, characterizations and mechanisms, RSC Adv. 7 (2017) 38747–38756. oxide by hydroxyl radicals, Carbon 105 (2016) 291–296.
[34] G.T.T. Le, J. Manyam, P. Opaprakasit, N. Chanlek, N. Grisdanurak, [64] N. Azizi, E. Gholibeglo, M. Maryami, S.D. Nayeri, S.M. Bolourtchian, Ultrasound
P. Sreearunothai, Divergent mechanisms for thermal reduction of graphene oxide mediated efficient ring opening of epoxides by in situ generated dithiocarbamates
and their highly different ion affinities, Diam. Relat. Mater. 89 (2018) 246–256. in green reaction media, C. R. Chim. 16 (2013) 412–418.
[35] J. Wang, B. Chen, Adsorption and coadsorption of organic pollutants and a heavy [65] M.M. Nasef, M. Zakeri, J. Asadi, E. Abouzari-Lotf, A. Ahmad, R. Malakooti,
metal by graphene oxide and reduced graphene materials, Chem. Eng. J. 281 (2015) Environmentally benign and highly regioselective ring opening of epoxides ac-
379–388. celerated by ultrasound irradiation, Green Chem. Lett. Rev. 9 (2016) 76–84.
[36] O.C. Compton, Z. An, K.W. Putz, B.J. Hong, B.G. Hauser, L. Catherine Brinson,

1222