Synthesis of graphene via ultra-sonic exfoliation of graphite oxide andits electrochemical characterization
Hassnain Asgar a, K.M. Deen b, Usman Riaz a, Zia Ur Rahman c, Umair Hussain Shah a,Waseem Haider a, c, *
a School of Engineering and Technology, Central Michigan University, Mt. Pleasant, MI 48859, USAb Department of Materials Engineering, University of British Columbia, Vancouver, BC V6T 1Z4, Canadac Science of Advanced Materials, Central Michigan University, Mt. Pleasant, MI 48859, USA
h i g h l i g h t s
� A relatively direct synthesis method for production of graphene is presented.� IR, Raman and XRD analyses confirmed formation of graphene starting from graphite.� XPS and TEM characterization validated the formation of graphene.� Electrochemical response of GO and graphene was evaluated in de-aerated 0.5M KOH.� Presence of functional groups in GO resulted in improved values of Rct and Ceff,p.
a b s t r a c t
A direct method of producing graphene from graphite oxide (GO) via ultra-sonication is presented in this work. The synthesis of graphene was validatedthrough IR, XRD, Raman, and XPS analyses. Moreover, the diffraction pattern obtained from TEM also validated the formation of graphene with char-acteristics (002) plane. The electrochemical behavior of GO and graphene was evaluated by electrochemical impedance spectroscopy and linear sweepvoltammetry in 0.5M KOH solution. The relatively larger effective pseudocapacitance and broad current peak exhibited by GO in the LSV plots was relatedwith the dominant adsorption of ‘Hads’ during reduction of water. It has been considered that large overpotential and relatively higher current responseexhibited by GO compared to graphene was associated with the preferential adsorption of Hads in the presence of surface functional groups.
Graphene and its related materials have unique physicochem-ical properties to support several electrochemical processesinvolving electro-catalysis [1], electrochemical sensing [2], super-capacitance [3] etc. It is well established that the heterogeneouselectron transfer, required for these processes, from/to a graphenesheet takes place on the edges and is affected by the attachedfunctional groups. The presence of functional groups also facilitatesthe adsorption/desorption of molecules on the graphene planes [4].The production method for graphene greatly influences theseproperties. A lot of work is going on with graphene as core interestand its production is being reported continuously via various
* Corresponding author. School of Engineering and Technology, Central MichiganUniversity, Mt. Pleasant, MI 48859, USA.
processes [5e7]. But the large-scale use of graphene is still hin-dered due to extensive and time-consuming production methods.
In this work, a relatively direct and efficient route is presented toproduce graphene from graphite and the effect of surface functionalgroups on electrochemical properties has been investigated.
2. Experimental
Graphite powder (Asbury Carbon Inc.) was oxidized to graphiteoxide (GO) using improved Hummer's method as explained inRef. [8]. After the reaction, GO was dried overnight at 80 �C. Toproduce graphene, GO was ultra-sonicated in DI water for 12 husing Branson® M2800H ultrasonic bath operating at the frequencyof 40 kHz. The powder from the suspension was obtained viacentrifugation with a g-force of ~2285 using the Rotofix 32ABenchtop Centrifuge by Helmer® Scientific.
Infrared (IR) spectra of graphite, GO and graphene was acquiredfrom FTIR-ATR spectrometer (NicoletTMiSTM 50) in Attenuated
Fig. 1. IR (a), XRD (b), and Raman (c) trends of graphite, graphite oxide, and graphene.
H. Asgar et al. / Materials Chemistry and Physics 206 (2018) 7e118
Total Reflection (ATR) mode. X-ray diffraction (XRD) patterns(Rigaku Mini Flex II) were obtained by using Cu Ka (l ¼ 1.54 Å)radiation source. For Raman spectra, the laser beam excitation of532 nm (Kaiser Optical Systems Inc.) was used. The XPS spectra(Thermo Scientific K-Alpha) of graphene were gathered by using AlKa irradiation source. The morphology and structure of graphenesheets were observed in transmission electron microscope (TEM)(HT7700).
For electrochemical studies, the working electrodes wereprepared by casting the paste made of an active material; GOand/or graphene (85 wt%), carbon black (5 wt%), binder; poly(-vinylidene fluoride) (10 wt%) and curing agent; 1-methyl-2-pyr-rolidinone into the electrode cavity. The as-cast paste was curedovernight at room temperature before electrochemical in-vestigations. Electrodes containing GO and graphene are termedas graphite oxide paste electrodes (GOPE) and graphene pasteelectrodes (GrPE), respectively in the following discussion. Elec-trochemical analyses of GOPE and GrPE were carried out by usingGamry-Potentiostat (R-3000) coupled with three-electrodes cellassembly in 0.5M KOH solution of pH 9.5 ± 0.5. GOPE/GrPE wereworking electrodes, saturated calomel electrode (SCE) was thereference, whereas, a platinum wire was used as a counterelectrode. Nitrogen gas was sparged for 30 min before each testto eliminate the effect of dissolved oxygen. Electrochemicalimpedance spectroscopy (EIS) was done with 5 mV AC potentialperturbation within 10 mHze100 kHz frequency range at 0 V DCbias potential versus OCP. Linear sweep voltammetry (LSV) scanswere obtained at sweep rates of 10, 5 and 2 mV/s in the reverse(cathodic) direction from 0 to �1.5 V vs. OCP.
Fig. 2. XPS high-resolution (inset; survey) spectra (a) and TEM
3. Results and discussion
IR, XRD and Raman spectra of graphite, GO and graphene areshown in Fig. 1. Graphite exhibited an insensitive behavior in theinfrared range (Fig. 1a). The transmittance peaks at 2325-1981 cm�1 could be associated with the diamond crystal used inATRmode [9]. A broad peak at 3405 cm�1 presented by GO could beattributed to the -OH bond stretching vibrations belonging to C-OHand/or adsorbed moisture. Similarly, the peak at 1173 cm�1 couldbe related to the stretching of C-O bond [10]. Graphene demon-strated similar behavior as graphite in the IR range; no signaturesfor functional groups which could be removed during ultra-soniccleavage of GO. The XRD patterns (Fig. 1b), represented the char-acteristic peak of graphite (2 <theta> ¼ 26.5�) corresponding to(002) plane. This peak was shifted to lower 2 <theta> values of10.8� and 23.9� in case of GO and graphene, respectively. In GO thiscorresponded to (001) plane which also suggests the successfuloxidation of graphite [11]. Whilst for graphene the peak at 23.9�
was related to the (002) plane of sp2 hybridized carbon atoms [12].The Raman spectra of graphite (Fig. 1c) showed a G band peak at1579 cm�1 which was affiliated with the stretching vibrations of in-plane carbon atoms. In GO and graphene, the G band vibrationsoriginated at relatively higher wavenumbers 1597 cm�1 and1601 cm�1, respectively, compared to graphite. Another band,known as D band, was also observed at 1359 cm�1 (GO) and1352 cm�1 (graphene) which may be related with the defects orirregularities in the plane of carbon chains and/ormay be due to theformation of grain boundaries [13]. XPS survey and high-resolutionspectra of graphene are presented in Fig. 2a. All the peaks were
micrograph (inset; diffraction pattern) (b) of graphene.
Fig. 3. Nyquist plots (a), bode plots (b), and linear sweep voltammograms for GOPE (c) and GrPE (d) in de-aerated 0.5M KOH solution.
Table 1Simulated electrochemical parameters calculated from the ECM fitting to theexperimental impedance spectra of GOPE and GrPE.
m 0.270 0.414Ceff,dl (mF/cm2) 44.27 41.56Ceff,p (mF/cm2) 1.43 1.11Goodness of Fit 114.9 � 10�6 138.9 � 10�6
H. Asgar et al. / Materials Chemistry and Physics 206 (2018) 7e11 9
calibrated in accordance with the C1s peak at 285.08 eV. The O1speak was observed at 533.08 eV. In the high-resolution spectra, thepeaks at 285.98 and 288.78 eV were attributed to the remaining/unremoved epoxide (-C-O-C-) and carboxyl (-COOH) groups [14],respectively. The spectra of C1s has depicted the same trend asreported in the literature for reduced-GO/graphene [14,15].
The morphology and diffraction pattern of graphene wereevaluated from TEM images. The sheets of graphenewere observedcontaining wrinkles (Fig. 2b), which could form due to theincreased surface energy and/or drying front of the suspensionduring sample preparation for microscopic analysis. The electrondiffraction pattern of graphene (inset), also validated the existenceof characteristic peak corresponding to (002) plane which was inconfirmation with the XRD pattern.
The impedance spectra of GOPE and GrPE are shown in Fig. 3aand b. The spectra were simulated with the equivalent circuitmodel (ECM) (inset; Fig. 3a). The kinetic performance of the sam-ples could be evaluated from the high-frequency depressed semi-circle and a slight decrease in phase angle (inset; Fig. 3b) at lowfrequency could be associated with the heterogeneous distributionof charge due to adsorption of ionic species. This adsorption can betermed as pseudocapacitance and is represented as constant phaseelement (Fp) in series with the parallel combination of the doublelayer (Fdl) and charge transfer resistance (Rct). The relatively lowerRct observed in case of GOPE (82.75 U-cm2) than GrPE (117.74 U-cm2) was attributed to the interaction of ionic species with thefunctional groups, present on GOPE as identified in the IR spectrum
[16]. The effective capacitance (Ceff) was calculated from the Hsu-Mansfeld model [17] which could be simulated with the normaldistribution of time constants and originated from the additiveeffect of double layer and pseudocapacitance. From Table 1, theCeff,dl for the double layer in case of GOPE and GrPE was almostsimilar, however, the effective pseudocapacitance (Ceff,p) relatedwith the adsorption of ionic species was relatively higher in GOPE.
To further confirm this behavior, the LSV curves (Fig. 3c and d)were obtained at three different sweep rates. The existence ofsweep rate dependent reduction peak and relatively larger currentresponse at large overpotential was evident in the case of GOPEcompared to GrPE which represented large polarization under thesame conditions without any reduction peak.
Fig. 4. The proposed reaction sequence for the electron transfer from the surface functional groups and adsorption of Hads species from the dissociation of water.
H. Asgar et al. / Materials Chemistry and Physics 206 (2018) 7e1110
H2O þ ee ¼ Hþ þ OHe (1)
> C¼O þ Hþ þ ee 4 ≡C e OH (2)
e COO� þ Hþ 4 e COOH (3)
The origin of this current peak at relatively lower overpotentialpredicts the electron transfer (pseudocapacitive response) reactionpossibly either due to the adsorption of Hads species by the disso-ciation of water molecules over the surface of GOPE or to thereduction of surface functional groups as given in reactions (1)e(3).The current peak was also shifted to larger overpotential (h) withan increase in sweep rate which corresponded to the delayedelectrochemical response or to the quasi-reversible character of theelectrode material. Fang et al. [18] and some other studies [19,20]report the preferential adsorption of Hþ ions on the active sites(present as surface functional groups) by the reduction of waterover graphitic materials. However, the possibility of the reductionof surface functional groups cannot be overruled under theseconditions. The presence of epoxide and carbonyl functional groupswas confirmed from the IR analysis of GOPE in our case as shown inFig. 1. Based on the experimental evidence, the possible reactionsequence of hydrogen adsorption has also been postulated asshown in Fig. 4.
Briefly, under applied negative potential the activation energyrequired for the electron tunneling through the surface functionalgroups e.g. carbonyl and epoxides over GOPE would decrease,leading to the adsorption of (Hþ) species from the water. This couldalso increase the active site for hydrogen adsorption and hencewould enhance the pseudo-capacitive response in the alkalineaqueous solution as observed in Fig. 3c. Therefore, the overpotentialprior to the current peak could be related to the activation energyrequired for the electron transfer reaction which was found to bepotential sweep rate dependent.
The limited but larger current response provided by GOPEcompared to GrPE even at very large h (~�1.0 V) and the existenceof broad current peak was assigned to the adsorption of Hads spe-cies by the reduction of water. Also, the continuous increase in
current density with increase in the overpotential (h) without anyobservable reduction current peak by GrPE was related to thestructural characteristics of the graphene.
4. Conclusions
Graphene was synthesized via a direct route and its electro-chemical response in de-aerated 0.5M KOH solutionwas evaluated.Conversion of graphite to graphite oxide and then to graphene wasvalidated by XRD, IR and Raman spectroscopy. XPS spectra pro-vided the insight into the surface chemistry of the graphene.Moreover, the formation of few layer graphene ((002)-plane) wasvalidated by TEM micrograph and diffraction pattern. Finally, theeffect of functional groups on the electrochemical properties ofGOPE and GrPE was evaluated. The lower value for Rct (82.75 U-cm2) and larger Ceff,p (1.43 mF/cm2) exhibited by GOPE wasattributed to the presence of surface functional groups. It wasconcluded that the presence of surface functional groups i.ecarbonyl and epoxide etc. on the graphene oxide could improve thepseudocapacitive electrochemical response in de-aerated 0.5MKOH solution.
References
[1] P. Fernandez, E. Castro, S. Real, M.E. Martins, Inter. J. Hydrogen Ener 34 (2009)8115e8126.
[2] M. Pumera, A. Ambrosi, A. Bonanni, E. Chng, H. Poh, Trends Anal. Chem. 29 (9)(2010) 954e965.
[3] C. Liu, Z. Yu, D. Neff, A. Zhamu, B. Jang, Nano Lett. 10 (2010) 4863e4868.[4] M. Pumera, Chem. Rec. 9 (2009) 211e223.[5] G. Nandamuri, S. Roumimov, R. Solanki, Nanotechnology 21 (14) (2010).[6] S. Stonkovich, D. Dikin, R. Piner, K. Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu,
S. Nguyen, R. Ruoff, Carbon 45 (7) (2007) 1558e1565.[7] F. Johra, J. Lee, W. Jung, J. Ind. Eng. Chem. 20 (2014) 2883e2887.[8] D. Marcano, D. Kosynkln, J. Berlin, A. Sinitskll, Z. Sun, A. Slesarev, L. Alemany,
W. Lu, J. Tour, ACS Nano 4 (2010) 4806e4814.[9] H. Lu, Y. Peng, M. Lin, S. Chou, J. Lo, B. Cheng, Opt. Photonics J. 3 (6A) (2013).
[10] C. Uthaisar, V. Barone, B. Fahlman, Carbon 61 (2013) 558e567.[11] C. Willemse, K. Tlhomelang, N. Jahed, P. Baker, E. Iwuoha, Sensors 11 (2011)
3970e3987.[12] D. Wang, W. Niu, M. Tan, M. Wu, X. Zheng, Y. Li, N. Tsubaki, ChemSusChem 7
H. Asgar et al. / Materials Chemistry and Physics 206 (2018) 7e11 11
[13] K. Krishnamoorthy, M. Veerapandian, K. Yun, S. Kim, Carbon 53 (2013) 38e49.[14] J. Zhang, H. Yang, G. Shen, P. Chneg, J. Zhang, S. Guo, Chem. Commun. 46
(2010) 1112e1114.[15] D. Dreyer, S. Park, C. Bielawski, R. Ruoff, Chem. Soc. Rev. 39 (2010) 228e240.[16] D. Bhattacharjya, J.S. Yu, J. Power Sources 262 (2014) 224e231.[17] C. Hsu, F. Mansfeld, Corrosion 57 (9) (2001) 747e748.
[18] Y. Fang, B. Luo, Y. Jia, X. Li, B. Wang, Q. Song, F. Kang, L. Zhi, Adv. Mater 24(2012) 6348e6355.
[19] Y.J. Oh, J.J. Yoo, Y.I. Kim, J.K. Yoon, H.N. Yoon, J.H. Kim, S.B. Park, Electrochim.Acta 116 (2014) 118e128.
[20] K.M. Deen, E. Asselin, Electrochim. Acta 212 (2016) 979e991.
