CERAMICSINTERNATIONAL

Available online at www.sciencedirect.com

http://dx.doi.org/0272-8842/& 20

nCorrespondenUniversity of Mafax: þ603 7967

E-mail addre

Please cite thishttp://dx.doi.or

Ceramics International ] (]]]]) ]]]–]]]www.elsevier.com/locate/ceramint

Easy preparation of ultrathin reduced graphene oxide sheets at a highstirring speed

Foo Wah Low, Chin Wei Lain, Sharifah Bee Abd Hamid

Nanotechnology & Catalysis Research Centre (NANOCAT), Level 3, IPS Building, University of Malaya (UM), 50603 Kuala Lumpur, Malaysia

Received 21 November 2014; received in revised form 1 January 2015; accepted 3 January 2015

Abstract

This work explored the synthesis of rGO sheets from graphene oxide (GO) using hydrazine solvent as reducing agent through chemicalreduction. Meanwhile, GO films with a 2D structure were prepared from graphite flakes (starting material with an average flake size of 150 nm)by an Improved Hummer's method. Results showed that the chemical oxidation of graphite flakes carried out at room temperature could be usedto prepare GO sheets in the initial stage. The conversion of GO into large-area rGO sheets with �85% of carbon content could then be achievedby chemical reduction. RGO sheets with a lateral dimension of up to �45 nm were obtained, which indicated the formation of an extremely thinlayer of rGO sheets. A high degree of GO reduction was also realized using a high stirring speed (1200 rpm) for 72 h in a mixture of acids andpotassium permanganate, resulting in a high carbon content of rGO with a large lateral dimension and area. Overall, our Improved Hummer'smethod with a high stirring speed (1200 rpm) for 72 h provided an easy approach to the preparation of large-area and ultrathin rGO sheets.& 2015 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

Keywords: Graphene oxide; Reduced graphene oxide; Improved Hummer's method; Chemical reduction

1. Introduction

Graphene is an allotrope of carbon with two-dimensional(2D) crystal structure, which comprises one atom-thick planarsheets of sp2-bonded carbon atoms that are densely packedin a honeycomb crystal lattice [1]. Graphene is the basicstructural element of other carbon allotropes, like graphite,charcoal, carbon nanotubes, and fullerenes. This material hasattracted much scientific interest in global research because ofits exceptional physical properties, such as high electronicconductivity, good thermal stability, and excellent mechan-ical strength [2–6]. An increasing number of research studieshave been devoted to this graphene material because of its 2Dhoneycomb-structured lattice that improves the movement ofcharge carriers. These charge carriers are able to travel sub-micrometer distances without scattering because of their high

10.1016/j.ceramint.2015.01.00815 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

ce to: Level 3, Block A, Institute of Postgraduates Studies,laya, 50603, Kuala Lumpur, Malaysia. Tel.: þ603 7967 6960;6556.ss: cwlai@um.edu.my (C.W. Lai).

article as: F.W. Low, et al., Easy preparation of ultrathin reduced gg/10.1016/j.ceramint.2015.01.008

electrical conductivity. Fundamentally, the high electricalconductivity of graphene is dictated by several form varia-tions, such as wrapped up-fullerene form, rolled up-carbonnanotube form, or stacked up-graphite form. Graphene cangenerally be produced by several disruptive technologies suchas mechanical approach [7], mechanical cleavage of graphite,epitaxial growth on a single crystal silicon carbide, andthermal expansion of graphite [2,8,9]. All of these approachescould be adapted for the synthesis graphene material; how-ever, mechanical approach is considered the most promisingmethod to achieve the best quality of graphene [10]. Crystalswith high structural and electronic quality are produced usingthis method. Unfortunately, the mechanical approach is stillfar from being an effective and efficient technology becauseof low production scale and time consumption issues [10].Thus, continuous efforts have been exerted to search foralternative way to mass produce graphene material that can beadopted for large-scale industrial production. In this work,Park and Ruoff indicated that chemical oxidation of graphiteis a feasible method for the preparation of graphene [11]. Thismethod involves oxidation of graphite to GO using highly

raphene oxide sheets at a high stirring speed, Ceramics International (2015),

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]]2

oxidizing agent (KMnO4) and subsequently reducing GO tographene using hydrazine solvent. The main advantage of thissynthesis method over other methods is that formation of alarge quantity of graphene in powder form could be achieved(scalability). Furthermore, graphene could be dispersed in bothpolar and non-polar solvents by functionalizing the surface ofgraphene, which contains ¼O stretching, C–O–C stretching, C–Ostretching, and broad band for hydroxyl functional group [12]. Asa matter of fact, these electron-donating functional groups couldbe used as a media for the complete exfoliation of GO fromgraphite flakes. Theoretically, chemical exfoliation of graphiteflakes caused weak van der Waals interaction bond between thecarbon layers [13]. By contrast, reduction of GO to graphenecaused all of the oxygen-containing groups to disappear during thechemical reduction stage [14]. Based on literature review, most ofthe graphene reported previously possessed small lateral dimen-sions of about few hundred nanometers to a few microns. Table 1presents the studies conducted by several authors regarding thesynthesis of GO. Much literature summarized that large-areagraphene and GO sheets were difficult to synthesize because ofunavoidable tearing of GO sheets during the extreme oxidationand reduction condition in an Improved Hummer's method andchemical exfoliation process.

Table 1Summary of GO lateral dimensions.

Authors/year Lateral dimension(nm)

Synthesis technique Reference

Zhang et al./2009

�500 Hummer's Method [15]

Yang et al./2011

�2000 Modified Hummer'sMethod

[16]

Su et al./2009 �1000 Modified Hummer'sMethod

[17]

Gurunathanet al./2013

�5000 Modified Hummer'sMethod

[18]

Ban et al./2012 �150,000 Simplified Hummer'sMethod

[19]

Zhou and Liuet al./2010

�200,000 Hummer's Method [20]

Table 2Summary of GO stirring speed.

Author/year Stirringspeed (rpm)

Stirringduration (h)

Synthesistechnique

Reference

Loryuenyonget al./2013

Notemphasize

1 ModifiedHummer's Method

[30]

Huang et al./2011

Notemphasize

72 Hummer's Method [31]

Thakur et al./2012

Notemphasize

4 ModifiedHummer's Method

[12]

Marcano et al./2010

Notemphasize

12' ImprovedHummer's Method

[21]

Harima et al./2011

Notemphasize

2 ModifiedHummer's Method

[32]

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reducedhttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

The effect of stirring speed on the synthesis of GO or rGO isone of the essential factors that has to be emphasized to obtainhigh-quality graphene. Furthermore, the synthesis process for GOor rGO needs to be simple and cost effective for possible adoptionto large industrial production scale. Based on our literature study inTable 2, we found that most of the researchers who synthesizedGO or rGO were focused in lower or constant speed. To the best ofour knowledge, detailed studies on the formation of nanoscale GOand graphene sheets at a high stirring speed of 1200 rpm arelacking. In this work, the influence of stirring speed on thepreparation of GO and rGO was explored to overcome certaindrawbacks such as extreme tearing of GO sheets and chemicalexfoliation process. This study aims to optimize the stability andscalability based on stirring speed to obtain the desired graphenesheets for material application, such as solar cell, membrane, andtemperature-sensitive device fabrication.

2. Experimental

2.1. Materials

Graphite flakes (o45 μm; Z99.99%), potassium permanga-nate (KMnO4; 97%), and hydrazine (35 wt% solution in water)were purchased from Sigma-Aldrich, Malaysia. Sulfuric acid(H2SO4; 0.5 M), phosphoric acid (H3PO4; 85%), and hydrogenperoxide (H2O2; 30%) were purchased from Chemolab, Malay-sia. hydrochloric acid (HCl; 37%) and acetonitrile (C2H3N,41.05 g/mol) were purchased from Merck, Malaysia.

2.2. GO synthesis

In a typical procedure, GO was synthesized from graphiteflakes by an Improved Hummer's method. A H2SO4:H3PO4

(180:20 mL) solution was added to a mixture of graphite flakes(1.5 g) and KMnO4 (9.0 g). The overall process was reacted inan ice–water bath condition (o20 1C) [21]. Five reactionmixture samples were then stirred for 6, 12, 24, 48, and 72 h atconstant speed (1200 rpm) to allow oxidation of graphite. Thecolor of the mixture changed from dark purplish green todark brown color. The reaction mixture was cooled to roomtemperature and poured into ice (E200 mL) with 3 mL ofH2O2 to stop the oxidation process, and the color of themixture changed to light brownish color, indicating a highoxidation level of graphite [21]. The GO formed was cen-trifuged by using Eppendorf Centrifuge 5430 R [Capacityfor 50 mL conical tubes (7800 rpm with 6 tubes)], and thesupernatant was decanted away. The remaining solid materialwas then washed with HCl and deionized (DI) water [21].

2.3. rGO synthesis

As a continuous step, GO powder was added to purifiedwater (3 mg/mL). Hydrazine (1 mL for 3 mg of GO) was thenadded immediately into the mixture, which was then immersedinto an oil bath at 80 1C to transfer heat more for equilibrium[22]. The mixture was centrifuged (7800 rpm for 1 h) severaltimes to remove unwanted materials such as the liquid lying

graphene oxide sheets at a high stirring speed, Ceramics International (2015),

Fig. 1. Electrolysis deposition and electrical characterization configuration setup.

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]] 3

above the rGO residue after precipitation. The process wasrepeated with stirring rates of 6, 12, 24, 48, and 72 h for eachof the five samples of GO described above (6, 12, 24, 48, and72 h). Thus, a total of 25 samples were prepared. The rGOsamples were dried in bench-tip air oven at 90 1C for 48 h.

2.4. Characterization

The functional groups of GO and rGO samples wereanalyzed over a wave number range of 500–4000 cm�1 usinga Bruker-IFS 66/S Fourier transform infrared (FTIR) spectro-scopy by the KBr pellet method. The spectra were recorded in500–4000 cm�1 range. The phases of the GO and graphenewere determined by X-ray diffraction (XRD) using a D8Advance X-ray diffractometer-Bruker AXS by employing ascanning rate of 0.033 deg/s in a 2θ range from 51 to 601 withCuKα radiation (λ¼1.5418 Å). Furthermore, Raman spectrawere obtained using a Renishaw inVia microscope with HeCdlaser source [laser excitation source (λ¼514.0 nm) at roomtemperature] for functional group and structural identificationpurpose. The band gaps of GO and rGO were calculated usingCary 60 UV–visible spectrophotometer from Agilent Technol-ogies with 400 nm/s scan rate and 200–800 nm wavelengthrange. The surface morphologies of the samples were observedby scanning electron microscopy (SEM) using a TM3030tabletop microscope at a working distance of approximately2.0 mm at high vacuum mode with 5.0 kV and images weremagnified into 30,000� times. The elemental analysis of thegraphite flakes, GO, and graphene were determined usingenergy dispersion X-ray (EDX microanalysis) equipped in theSEM. The morphologies for GO and graphene were furtherconfirmed through high-resolution transmission electronmicroscopy (HRTEM) using a JEM 2100-F with 200 kVaccelerating voltage.

2.5. Deposition of graphite, GO, and rGO setup

For electrical characterization, an electrolysis method setupwas employed for graphite, GO, and rGO deposition process.The overall process involved coating a very thin layer ofgraphite, GO, and rGO on fluorine-doped tin oxide-coatedglass slide (FTO, �7 Ω/sq; Sigma-Aldrich). For depositionpurpose, the two electrodes of FTO glass were immersed into asolution of 100 mL of DI water, 1 mL of acetonitrile, and0.15 g of graphite. The two electrodes were then connected toa power supply with 60.0 V for 15 min processing, whichcreated a circuit through the graphite. The very thin layer ofgraphite, GO, or rGO will coated on anode side (–). Thepurpose of utilize FTO glass slide as both electrodes becauseof make easier to transfer the coated thin layer of graphite, GO,and rGO for anode side to tri-electrode system. Subsequently,the coated sample was connected in a tri-electrode system (redcolor) on an Autolab PGSTAT204 instrument (Fig. 1) and thesamples were characterized under 300 W of solar illuminationlight source. Linear sweep voltammetry was carried out on thegraphite in H2SO4 at various voltages; from �0.3 V reductionat the beginning, a maximum reduction was observed at 0.3 V.

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reduced ghttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

The process was repeated with dark condition, 0.15 g of GO,and rGO instead of graphite. The configuration system isshown in Fig. 1.

3. Results and discussion

In the present study, the effect of stirring duration at a highspeed of 1200 rpm on the formation of nanoscale rGO sheetswas discussed. The present study aims to determine theoptimum duration at high stirring speed that would achievethe desired rGO with highly efficient electrochemical properties.The variation in the transmittance value of the functional groupfor rGO sheet synthesis based on different stirring durationswas studied using the FTIR instrument. Theoretically, GO hasseveral essential acute peaks detected by FTIR patterns suchas C¼O stretching at 1720 cm�1, C–O–C stretching at1204 cm�1, and C–O stretching at 1049 cm�1. These resultsconfirmed that the oxygen molecules (O) were occupied greatlyduring the GO synthesis step. Furthermore, a broad and intenseband of hydroxyl group (–OH) at 3400 cm�1 was observedfrom the FTIR pattern (Fig. 2). Interestingly, the transmittancevalue (intensity) broadened with increased high-speed stirringduration, indicating the increased diffusion rate of oxygenmolecules into the graphite flakes to form carbonyl group,carboxylic group, and epoxy group in GO [12]. Based on basicchemistry knowledge, it might be obtained high-yield large-areaGO of desired result could be expected when 72 h of high-speed

raphene oxide sheets at a high stirring speed, Ceramics International (2015),

Fig. 2. FTIR spectra of GO based on different stirring durations: (a) 6, (b) 12,(c) 24, (d) 48, and (e) 72 h.

Fig. 3. FTIR spectra of (a) graphite, (b) GO with 72 h stirring, and rGOproduced from sample (b) with stirring duration of (c) 6, (d) 12, (e) 24, (f) 48,and (g) 72 h.

Fig. 4. XRD patterns of (a) graphite, (b) GO, and (c) rGO samples.

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]]4

stirring at 1200 rpm because of the significant appearance ofoxygenated bonds during oxidation process [12]. Thus, largeamount of oxygenated molecules participate in the reaction andweaken the van der Waals interaction between the graphiteoxide layers to facilitate their exfoliation [23]. Hence, a highstirring speed (1200 rpm) for 72 h was used as optimumparameter for the rest of the experiments in the beginning stageof GO sheet preparation. As an addition, functional group ofC¼O stretching at 1720 cm�1 for 24 h sample was slightlyshifted to the lower value might due to the defect occurredduring GO synthesis.

Fig. 3 presents the FTIR results for graphite, GO sample thatunderwent 72 h stirring, and 72 h GO sample chosen for rGOproduction with 6, 12, 24, 48, and 72 h stirring duration. Thereason for comparing the variation of rGO stirring duration isto observe the fluctuation of functional groups during thechemical reduction process. All of the essential peaks of therGO samples at 1720, 1204, 1049, and 3400 cm�1 signifi-cantly dropped compared with either graphite or GO samples.These phenomena clearly illustrated that oxygen functionalgroups such as C¼O stretching, C–O–C stretching, C–Ostretching, and hydroxyl groups were mostly removed andbecame rGO during the chemical reduction process. Given thatthe rGO results at 3400 cm–1 peak from 6 h to 72 h stirringduration slowly decreased, the 72 h rGO formation processwas implemented for easy elimination of hydroxyl groupsfrom the carbon basal plane and simple transformation torGO monolayer. Lastly, a disappearance of C¼C bonds at1720 cm�1 along the rGO samples caused by strong oxidationagent such as KMnO4 was observed [24].

In this part of the experiment, XRD analysis was used toinvestigate the crystal structure and orientation and to verify theaverage spacing between GO and rGO layers. Thus, representativesamples were selected for the XRD analysis. Fig. 4(a) shows theXRD patterns of pristine graphite. A sharp and high-intensity peakat 2θ¼26.581 was detected from the XRD pattern (FWHM¼0.3600), which corresponds to the well-arranged layer structurewith d-spacing of 0.335 nm along the (002) orientation [23,24]. In

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reducedhttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

addition, a small Bragg reflection of graphite phase (004) was alsodetected at 2θ value of 54.731 with d-spacing of 0.17 nm(FWHM=0.2952) in the XRD patterns. The presence of graphitephase was identified by ICDD file 9012230 (Reference Pat-tern=01-075-2078). The XRD pattern for GO at a high stirringspeed (after the pristine graphite undergo oxidation process) isshown in Fig. 4(b). The peak (002) shifted to a low angle at2θ¼8.591 with FWHM¼0.6000 (Reference Pattern¼01–074-2329), but the distance between consecutive carbon layers (d-spacing) increased from 0.335 nm to 0.703 nm because of theintroduction of oxide functional groups to the carbon basal planeby chemical oxidation reaction. This phonomena inferred that theinterlayer distance (d-spacing) between consecutive carbon layerswere weakened by the intercalation of GO formation, which haveepoxy, hydroxyl, carbonyl, and carboxyl groups located betweenthe edges of consecutive carbon basal planes [23–25]. The rGOsample exhibited a wide peak at 2θ=24.31 with d-spacing of

graphene oxide sheets at a high stirring speed, Ceramics International (2015),

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]] 5

0.38 nm (FWHM=0.5904) as presented in Fig. 4(c). This resultmay be attributed to the assembly of graphene layers and indicatedthat poor arrangement of graphene sheets along the stackingdirection, which implied this rGO sample was comprised mainlyfrom single or only a few layers of rGO with d-spacing of 0.38 nm[12]. Furthermore, the disappearance of the peak at 2θ¼8.591from the XRD pattern also indicated that the oxygen functionalgroups or the oxygenated molecules were fully removed from therGO sample [12]. A less intense peak with (001) orientation ofgraphene were observed at 2θ¼42.71 (FWHM¼0.2952) and2θ¼44.51 (FWHM¼0.2952) indicated the turbostratic band ofdisordered carbon materials (Reference Pattern¼96-210-1125).

Raman analysis was applied to determine and understandthe number and orientation of layer, quality, and types of edgeof the GO and rGO samples. For the GO sample, the majorband position in the Raman spectra was significantly changedafter treating with a reducing agent (hydrazine solvent)because of elastically scattering charge carriers [26]. Fromthe spectra of GO and rGO in Fig. 5, two major vibrationbands in the range between 1100 and 1700 cm�1 weredetected. The D vibration bands of GO and rGO were recordedat 1353.85 and 1349.25 cm�1, respectively, with the breathingmode of κ-point photons of A1g symmetry [12]. On thecontrary, the G vibration bands of GO and rGO were observedat 1588.40 and 1603.31 cm�1 to comply with first-orderscattering of E2g phonons by their sp2-hybridized carbon[12]. Generally, the intensity ratio of the D band to the Gband (ID/IG) increased the intensity vibration band by morethan double after the reduction of GO. The D vibration band ofrGO increased mainly because of the sp2 carbon orientation.The high intensity of G vibration band also indicated thatgraphene obtained more isolated domains in rGO than in GO[27]. The G vibration band of rGO increased because of theirremoval of oxygen moieties from GO [28]. To verify themonolayer or multilayer stacking of graphene layers on rGO,the two-phonon (2D) Raman scattering of graphene-basedmaterials is an essential peak to analyze. The significant peakthat appeared at 2711.76 cm�1 determined that rGO is a

Fig. 5. Raman spectra of (a) GO and (b) rGO samples.

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reduced ghttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

Lorentzian peak of the 2D vibration of monolayer graphenecompared with GO with a smaller peak at 2669.00 cm�1.Consequently, the 2D bands of both GO and rGO peaks wereshifted to a high wave number, and the peaks indicatedmultilayer rGO. The 2D vibration bands of GO and rGO wererecorded at 2943.27 and 2945.77 cm�1, respectively. The 2Dpeaks of both GO and rGO indicated that they have multilayerstructures. The incident occurred may be because GO consistsof oxygenated molecules that have the ability to preventgraphene layers from stacking with one another. Conversely,the intensity of rGO slightly increased may be because of theabsence of functional groups, and only a few graphene layersstack with one another to form a multilayer rGO [12].Ultraviolet–visible spectroscopy (UV–vis) measurement was

carried out to monitor the degree of oxidation of the GO andrGO samples. In addition, the absorption of GO and rGOtransition from the ground state to excited state were deter-mined. The result was strongly based on its reflectance in thevisible range depending on the color of the GO and rGOsamples involved. The spectra for the GO and rGO samples areshown in Fig. 6, with absorption peaks at 272 and 284 nm,respectively. Based on the absorbance spectra of GO, the mainabsorption peak at 272 nm indicated the π–πn transition of thearomatic C–C ring, whereas the weak absorption at 319 nmwas due to the n–πn transition of C¼O bonds [12]. When GOwas reduced to rGO, a red shift within the UV region wasobserved at 284 nm; the decrease in oxygenated functionalgroups and the increase of aromatic C–C bonds facilited theelectrons to be easily excited at low energy [29]. In addition,the energy band gap of rGO (4.37 eV) is much lower than thatof GO (4.56 eV), which proved that the energy required to freean outer shell electron from its orbit shell to become mobilecharge carrier is much better for rGO than for GO. Besides, thedata obtained around the sharp peak were rough becausehigh concentrations of GO and rGO were detected during theUV–vis test. The band gap measurement of GO and rGO is

Fig. 6. UV–vis spectra of (a) GO and (b) rGO samples.

raphene oxide sheets at a high stirring speed, Ceramics International (2015),

Fig. 7. SEM images of (a) graphite, (b) GO, and (c) rGO samples.

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]]6

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reducedhttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

shown below:

Eg ¼1240λ

ð1Þ

Light brownish GO solution and light black rGO solution wereused for the UV–vis test (as shown in the inset of Fig. 6). Theoptical absorption of rGO was expected to be higher thanthat of GO. In this case, rGO solution was taken in lesserconcentration than GO to obtain a stable dispersion of rGO inabsolute ethanol. In fact, UV–vis spectra totally depend on theconcentration of the solution involved; therefore, rGO exhib-ited higher optical absorption than GO.The morphology of the graphite, GO, and rGO samples

were illustrated using SEM as presented in Fig. 7. Asillustrated in Fig. 7(a), graphite particles are in a platelet-likecrystalline form of carbon [30]. The layer of graphite sheet hasa thickness of 154 nm. Interestingly, the thickness of GOsheets became small with a thickness of 89.2 nm [Fig. 7(b)].The main reason was that the consecutive carbon basal planes(d-spacing) were increased and the distance between carbonlayers increased simultaneously. The morphology of the rGOsample [Fig. 7(c)] has a similar appearance to the graphite andGO samples, but the average wall thickness of the sheet layerwas further reduced to E46 nm. This result suggests that theoxygenated molecules were removed from the rGO orbital.When oxidation was reduced by hydrazine, the rGO sheetsfully obtained a monolayer structure with 46 nm (Fig. 7(c)).Subsequently, corresponding EDX analysis were conducted

to determine the average atomic percent (at%) of the elementspresent in graphite, GO, and rGO samples. The results fromEDX spectra are presented in Table 3. For the pristine graphitesample, the peaks of C and O were detected, which indicate theexistence of C (96.30 at%) and O (3.70 at%) elements. Forthe GO sample, the intensity of the C peak from the EDXspectrum decreased significantly from 96.30 at% to 65.41 at%,whereas the O element increased tremendously from 3.70% to34.59%. This result was understandable because of thepresence of oxygenated molecules from KMnO4 during theoxidation process. A high C peak was found from the rGOsample (88.01 at%), but the O peak decreased to 11.99%. Theresult was attributed to the strong reduction reactions involvedusing hydrazine agent. Thus, carbon content significantlyincreased, and carbon–carbon atoms well bonded witheach other.An atomic scale morphological study was carried out to

examine the structural layers of GO and rGO using theHRTEM analysis as exhibited in Fig. 9. Fig. 9(a) shows the

Table 3EDX results of graphite, GO, and rGO samples.

Materials Carbon, C (at%) Oxygen, O (at%) Total (at%)

Graphite 96.30 3.70 100.00GO 65.41 34.59 100.00rGO 88.01 11.99 100.00

graphene oxide sheets at a high stirring speed, Ceramics International (2015),

Fig. 8. A schematic of the silk-like, three-dimensional appearance of rGOmonolayer.

Fig. 9. HRTEM images of (a) GO and (b) rGO samples.

Fig. 10. I–V characteristic curves of (a) FTO default glass, (b) graphite, (c) GOwith 72 h stirring duration, and (d) rGO with 72 h stirring duration.

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]] 7

HRTEM image of the GO sample after high-speed stirring.The image shows that large amount of white spots weredeposited on the surface of the GO sample. This observationindicates the presence of a bonding network between theoxygenated molecules and the carbon planes. The HRTEMimage for rGO sample is exhibited in Fig. 9(b). The silk-likeappearance of rGO after the reduction process was clearlyobserved. A simple three-dimensional schematic of a mono-layer rGO sample is shown in Fig. 8. This image manifests thatultrathin layer of rGO can be formed using a stirrer at a highstirring speed.

A scanning potentiostat was used to measure current underan applied potential (I–V) characteristic curves for the graphite,GO, rGO samples (Fig. 10). Under dark conditions, all samplesexhibited non-significant current of less than 10 mA. However,the current increased under solar illumination. A maximumcurrent of 7.6 mA was observed for the rGO sample, which isrelatively higher than those of GO (6.6 mA) and graphite(1.9 mA). These results are consistent with literature, whichsuggested that the ultrathin monolayer rGO sample couldenhance the transportation of photo-induced charge carriersrapidly. This result reveals that the rGO sample exhibitedbetter electrical properties than GO and graphite samples. Inthe same way, current (I) exhibited an increasing trend fromrGO (7.6 mA)4GO (6.6 mA)4graphite (1.9 mA). This dra-matically increasing trend proved that electrochemical rGOshowed a much higher carbon to oxygen ratio than GO andgraphite.

Overall, a schematic diagram illustrating the formation ofgraphene-like materials (rGO) through exfoliation and reductionof GO sample is presented in Fig. 11. In brief, the efficientproduction of GO could be realized by an Improved Hummer'smethod, whereas the exfoliation of GO to form rGO could beperformed by chemical reduction. Based on an Improved Hum-mer's method, the main function of H2SO4 was to add functionalgroups to the graphite surface in the presence of KMnO4 asoxidizing agent. Subsequently, the synthesized GO sample created

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reduced ghttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

functionalized oxygenated bonds with carbon atoms. When GOwas formed, the crystalline structure of d-spacing was increasedcompared with the graphite flake structure. Lastly, GO was reduced

raphene oxide sheets at a high stirring speed, Ceramics International (2015),

Fig. 11. Mechanism of rGO formation: (a) graphite undergoing oxidation process to (b) GO and exfoliation of GO to (c) rGO under chemical reduction process.

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]]8

with hydrazine solvent to form rGO sample with ultrathinmonolayer crystalline structure. The oxygenated bonds was fullyexpelled or reduced from rGO orientation.

4. Conclusion

The present study demonstrated that a high stirring speed of1200 rpm for 72 h with hydrazine solvent as reducing agent iseffective in synthesizing ultrathin monolayer rGO with anaverage thickness of 46 nm. High-yield large-area GO wasobtained after 72 h of high-speed stirring at 1200 rpm becauseof the significant appearance of oxygenated bonds duringoxidation. Based on the FTIR analysis, elimination of hydroxylgroups from the carbon basal plane was achieved significantlyto form monolayer rGO sheets after high-speed stirring for72 h. The ultrathin monolayer rGO sample could enhance thetransportation of photo-induced charge carriers rapidly basedon the I–V characteristic curves as compared to GO andgraphite samples. Such a mechanistic understanding is veryimportant for the controlled growth of ultrathin rGO sheets,which may be used in many socio-economic applications, suchas solar cells, touch sensors, light-emitting diodes (LEDs),ultrafast transistors, and others.

Acknowledgement

This research is supported by High Impact Research Chan-cellory Grant UM.C/625/1/HIR/228 (J55001-73873) from theUniversity of Malaya.

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reducedhttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

Reference

[1] H.P. Boehm, R. Setton, E. Stumpp, Nomenclature and terminology of

graphite intercalation compounds (IUPAC Recommendations 1994), Pure

Appl. Chem. 66 (9) (1994) 1893–1901.[2] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang,

S.V. Dubonos, I.V. Grigorieva, A.A. Firsov, Electric field effect in

atomically thin carbon films, Science 306 (5696) (2004) 666–669.[3] R.R. Nair, P. Blake, A.N. Grigorenko, K.S. Novoselov, T.J. Booth,

T. Stauber, N.M.R. Peres, A.K. Geim, Fine structure constant defines

visual transparency of graphene, Science 320 (5881) (2008) 1308. 10.

1126/science.1156965.[4] J.S. Bunch, A.M. Van Der Zande, S.S. Verbridge, I.W. Frank,

D.M. Tanenbaum, J.M. Parpia, H.G. Craighead, P.L. McEuen, Electromecha-

nical resonators from graphene sheets, Science 315 (5811) (2007) 490–493.[5] A. Yu, P. Ramesh, M.E. Itkis, E. Bekyarova, R.C. Haddon, Graphite

nanoplatelet-epoxy composite thermal interface materials, J. Phys. Chem.

C 111 (21) (2007) 7565–7569.[6] M.D. Stoller, S. Park, Y. Zhu, J. An, R.S. Ruoff, Graphene-based

ultracapacitors, Nano Lett. 8 (10) (2008) 3498–3502.[7] A.K. Geim, K.S. Novoselov, The rise of graphene, Nat. Mater. 6 (3)

(2007) 183–191.[8] K.V. Emtsev, A. Bostwick, K. Horn, J. Jobst, G.L. Kellogg, L. Ley,

J.L. McChesney, T. Ohta, S.A. Reshanov, J. Röhrl, E. Rotenberg,

A.K. Schmid, D. Waldmann, H.B. Weber, T. Seyller, Towards wafer-

size graphene layers by atmospheric pressure graphitization of silicon

carbide, Nat. Mater. 8 (3) (2009) 203–207.[9] G. Chen, W. Weng, D. Wu, C. Wu, J. Lu, P. Wang, X. Chen, Preparation

and characterization of graphite nanosheets from ultrasonic powdering

technique, Carbon 42 (4) (2004) 753–759.[10] A.K. Geim, Graphene: status and prospects, Science 324 (5934) (2004)

1530–1534.[11] S. Park, R.S. Ruoff, Chemical methods for the production of graphenes,

Nat. Nanotechnol. 4 (4) (2009) 217–224.[12] S. Thakur, N. Karak, Green reduction of graphene oxide by aqueous

phytoextracts, Carbon 50 (14) (2012) 5331–5339.

graphene oxide sheets at a high stirring speed, Ceramics International (2015),

F.W. Low et al. / Ceramics International ] (]]]]) ]]]–]]] 9

[13] P. Zhu, M. Shen, S. Xiao, D. Zhang, Experimental study on thereducibility of graphene oxide by hydrazine hydrate, Phys. B: Condens.Matter 406 (3) (2011) 498–502.

[14] T. Kuila, S. Bose, P. Khanra, A.K. Mishra, N.H. Kim, J.H. Lee, A greenapproach for the reduction of graphene oxide by wild carrot root, Carbon50 (3) (2012) 914–921.

[15] T. Zhang, D. Zhang, M. Shen, A low-cost method for preliminaryseparation of reduced graphene oxide nanosheets, Mater. Lett. 63 (23)(2009) 2051–2054.

[16] S.T. Yang, S. Chen, Y. Chang, A. Cao, Y. Liu, H. Wang, Removal ofmethylene blue from aqueous solution by graphene oxide, J. ColloidInterf. Sci. 359 (1) (2011) 24–29.

[17] C.Y. Su, Y. Xu, W. Zhang, J. Zhao, X. Tang, C.H. Tsai, L.J. Li,Electrical and spectroscopic characterizations of ultra-large reducedgraphene oxide monolayers, Chem. Mater. 21 (23) (2009) 5674–5680.

[18] S. Gurunathan, J.W. Han, J.H. Kim, Green chemistry approach for thesynthesis of biocompatible graphene, Int. J. Nanomedicine 8 (1) (2013)2719–2732.

[19] F.Y. Ban, S.R. Majid, N.M. Huang, H.N. Lim, Graphene oxide and itselectrochemical performance, Int. J. Electrochem. Sci. 7 (5) (2012)4345–4351.

[20] X. Zhou, Z. Liu, A scalable, solution-phase processing route to grapheneoxide and graphene ultralarge sheets, Chem. Commun. 46 (15) (2010)2611–2613.

[21] D.C. Marcano, D.V. Kosynkin, J.M. Berlin, A. Sinitskii, Z. Sun,A. Slesarev, L.B. Alemany, W. Lu, J.M. Tour, Improved synthesis ofgraphene oxide, ACS Nano 4 (8) (2010) 4806–4814.

[22] S. Park, J. An, J.R. Potts, A. Velamakanni, S. Murali, R.S. Ruoff,Hydrazine-reduction of graphite- and graphene oxide, Carbon 49 (9)(2011) 3019–3023.

[23] F. Zeng, Z. Sun, X. Sang, D. Diamond, K.T. Lau, X. Liu, D.S. Su, In situone‐step electrochemical preparation of graphene oxide nanosheet‐modifiedelectrodes for biosensors, Chem. Sus. Chem. 4 (11) (2011) 1587–1591.

Please cite this article as: F.W. Low, et al., Easy preparation of ultrathin reduced ghttp://dx.doi.org/10.1016/j.ceramint.2015.01.008

[24] G.I. Titelman, V. Gelman, S. Bron, R.L. Khalfin, Y. Cohen, H. Bianco-Peled,

Characteristics and microstructure of aqueous colloidal dispersions of graphite

oxide, Carbon 43 (3) (2005) 641–649.[25] M. Wojtoniszak, X. Chen, R.J. Kalenczuk, A. Wajda, J. Łapczuk,

M. Kurzewski, M. Drozdzik, P.K. Chu, E. Borowiak-Palen, Synthesis,

dispersion, and cytocompatibility of graphene oxide and reduced gra-

phene oxide, Colloids Surf. B. Biointerfaces 1 (89) (2012) 79–85.[26] L.G. Cancado, M.A. Pimenta, B.R.A. Neves, M.S.S. Dantas, A. Jorio,

Influence of the atomic structure on the raman spectra of graphite edges,

Phys. Rev. Lett. 93 (247401) (2004) 1–4.[27] Y. Wang, Z. Shi, J. Yin, Facile synthesis of soluble graphene via a green

reduction of graphene oxide in tea solution and its biocomposites, ACS

Appl. Mater. Interfaces 3 (4) (2011) 1127–1133.[28] P. Cui, J. Lee, E. Hwang, H. Lee, One-pot reduction of graphene

oxide at subzero temperatures, Chem. Commun. 47 (45) (2011)

12370–12372.[29] F.T. Thema, M.J. Moloto, E.D. Dikio, N.N. Nyangiwe, L. Kotsedi,

M. Maaza, M. Khenfouch, Synthesis and characterization of graphene

thin films by chemical reduction of exfoliated and intercalated graphite

oxide, J. Chem. 2013 (2013) (2013) 1–6.[30] V. Loryuenyong, K. Totepvimarn, P. Eimburanapravat, W. Boonchompoo,

A. Buasri, Preparation and characterization of reduced graphene oxide sheets

via water-based exfoliation and reduction methods, Adv. Mater. Sci. Eng. 2013

(2013) (2013) 1–5.[31] N.M. Huang, H.N. Lim, C.H. Chia, M.A. Yarmo, M.R. Muhamad,

Simple room-temperature preparation of high-yield large-area graphene

oxide, Int. J. Nanomedicine 6 (2011) (2011) 3443–3448.[32] Y. Harima, S. Setodoi, I. Imae, K. Komaguchi, Y. Ooyama, J. Ohshita,

H. Mizota, J. Yano, Electrochemical reduction of graphene oxide in

organic solvents, Electrochim. Acta 56 (2011) (2011) 5363–5368.

raphene oxide sheets at a high stirring speed, Ceramics International (2015),