Journal of Colloid and Interface Science 430 (2014) 108–112

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

Short Communication

Dispersion behaviour of graphene oxide and reduced graphene oxide

http://dx.doi.org/10.1016/j.jcis.2014.05.0330021-9797/� 2014 Elsevier Inc. All rights reserved.

⇑ Corresponding author.E-mail address: kymakis@staff.teicrete.gr (E. Kymakis).

Dimitrios Konios a,b, Minas M. Stylianakis a,b, Emmanuel Stratakis c, Emmanuel Kymakis a,⇑a Center of Materials Technology and Photonics & Electrical Engineering Department, School of Engineering, Technological Educational Institute (TEI) of Crete, Heraklion 71004,Crete, Greeceb Department of Chemistry, University of Crete, P.O. Box 2208, Heraklion 71003, Crete, Greecec Institute of Electronic Structure and Laser (IESL), Foundation for Research and Technology-Hellas (FORTH), Heraklion 71110, Crete, Greece

a r t i c l e i n f o a b s t r a c t

Article history:Received 2 April 2014Accepted 22 May 2014Available online 2 June 2014

Keywords:GrapheneGraphene oxideReductionDispersionSolubility

The dispersion behaviour of graphene oxide (GO) and chemically reduced GO (rGO) has been investigatedin a wide range of organic solvents. The effect of the reduction process on the GO solubility in eighteendifferent solvents was examined and analysed, taking into consideration the solvent polarity, the surfacetension and the Hansen and Hildebrand solubility parameters. rGO concentrations up to �9 lg/mL inchlorinated solvents were achieved, demonstrating an efficient solubilization strategy, extending thescope for scalable liquid-phase processing of conductive rGO inks for the development of printed flexibleelectronics.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction

Graphene is an atomically thin layer of sp2-bonded carbonatoms, stacked in a two-dimensional (2D) honeycomb lattice,forming the basic building block for carbon allotropes of anydimensionality [1]. Since its isolation as a monolayer, graphenehas attracted an extraordinary amount of interest due to its poten-tial application in the fastest growing scientific fields, such assupercapacitors [2], biosensors [3], photovoltaics [4] and touchpanels [5].

Chemical vapour deposition (CVD) [6] and micromechanicalexfoliation of graphite are the most widely used fabricationmethods of less defective graphene films. However, the CVDdeposition of uniform large area graphene films on arbitrarysubstrates at low temperatures is not possible and furthermorethis method is incompatible with roll to roll mass productionprocesses. At the same time, the exfoliated graphene exhibits verylow solubility in common organic solvents [7], due to the essentialaddition of a stabilizer as the exfoliation liquid medium [8].

On the other hand, exfoliated graphene oxide (GO) is the idealalternative for the production of solution processable graphene,as it can be synthesized in large quantities from inexpensivegraphite powder and can readily yield stable dispersions in varioussolvents [9]. GO is an oxidized graphene sheet having its basal

planes decorated mostly with epoxide and hydroxyl groups, inaddition to carbonyl and carboxyl groups located at the edges [10].

The covalent character of C–O bonds disrupts the sp2 conjugationof the hexagonal graphene lattice, making GO an insulator. Never-theless, GO can be partially reduced to conductive graphene-likesheets by removing the oxygen-containing groups [11–13]. In thisway the conjugated structure of graphene can be recovered, result-ing in reduced graphene oxide (rGO) with important electrical prop-erties partially restored [14].

However, the preparation of dispersed form of graphene forapplications in printed flexible electronics is not a straightforwardprocess, since its stability in various solvents is a critical point. Inthis context, the solubility of GO in various solvents has beenrecently examined by several groups [9,15,16]. However, there isa gap in the literature on the direct comparison on solubility valueson GO and rGO, which in principle they are different. Therefore, theknowledge on how conductive rGO stable solution can be obtainedin common organic solvents is vital.

In this work, the dispersion behaviour of GO and chemicallyrGO is compared, aiming to get an insight into how the removalof oxygen containing groups during the reduction process affectsits dispersion quality. The solubility/dispersibility of rGO is inves-tigated in eighteen different solvents and directly compared withthe pristine GO. In this way, critical solubility values are recordedaiming at the application of conductive rGO inks on printed flexibleelectronic devices [17].

Fig. 1. UV–Vis spectra of GO dissolved in water at different concentrations. Theinset shows the linear relationship between the absorbance per unit path lengthand the concentration of GO.

D. Konios et al. / Journal of Colloid and Interface Science 430 (2014) 108–112 109

2. Results and discussion

For the preparation of GO and rGO dispersions, the productsprepared as described in the Experimental Section (SI), were firstgrounded with a mortar and pestle. In order to compare thedispersion behaviour in the different solvents, the same quantityof GO and rGO powder (�1 mg) was added to a given volume ofsolvent (�2 mL), with an initial concentration of 0.5 mg/mL. GOand rGO dispersions were tested in the following organic solvents:(DI) water, acetone, methanol, ethanol, 2-propanol, ethyleneglycol, tetrahydrofuran (THF), N,N-dimethylformamide (DMF),N-methyl-2-pyrrolidone (NMP), n-hexane, dichloromethane(DCM), chloroform, toluene, chlorobenzene (CB), o-dichloroben-zene (o-DCB), 1-chloronaphthalene (CN), acetylacetone, diethylether. The dispersions were sonicated in an ultrasound bath cleaner(Elmasonic S30H) for 1 h and then mildly centrifuged at 500 rpm for90 min (Alegra X-22) to remove the large aggregates. Afterwards,the supernatant was collected for analysis.

For the estimation of solubility values for GO and rGO in differentsolvents, UV–Vis spectroscopy was performed on a ShimadzuUV2401PC UV–Vis spectrometer. Using the 2 weeks left suspen-sions, the dispersibility of GO and rGO in each solvent was examinedfrom the linear relationship between the absorbance (A) and theconcentration (C) of a compound in a solution, given by the Lam-bert–Beer law (A = a l C). It is necessary to determine the absorptioncoefficient (a), which is related to the absorbance per unit pathlength A/l and it is an important parameter in characterizing any dis-persion. For this purpose, a calibration line was constructed by mea-suring the absorbance at 660 nm of four GO and rGO solutions withdifferent, low concentrations (Fig. 1). The procedure was repeatedfor each solvent. The observed values divided by the cuvette length(l = 1 cm) were plotted versus the known concentration values,allowing to estimate the absorption coefficient for its suspension.Using a values, the maximum solubility of GO and rGO in each sol-vent could be extracted (Table 1).

Treating GO with hydrazine causes an enormous structuralchange with the recovery of the conjugated system, through theremoval of oxygen containing groups. The morphology, structureand composition of GO and rGO were characterized by Ramanspectroscopy, Fourier transform infrared spectroscopy (FT-IR),X-ray diffraction (XRD) and Thermogravimetric Analysis (TGA).

Raman spectroscopy is a powerful tool, which can be used tocharacterize carbonaceous materials and particularly for distin-guishing the disorder in the crystal structures of carbon. In theRaman spectrum of GO and rGO (Fig. 2a), two prominent peaks

Table 1Dipole moments, surface tensions and Hildebrand parameters of solvents and GO, rGO so

Solvents Dipole moment Surface tension (mN/m

Di water 1.85 72.8Acetone 2.88 25.2Methanol 1.70 22.7Ethanol 1.69 22.12-propanol 1.66 21.66Ethylene glycol 2.31 47.7Tetrahydrofuran (THF) 1.75 26.4N,N-dimethylformamide (DMF) 3.82 37.1N-methyl-2-pyrrolidone (NMP) 3.75 40.1n-Hexane 0.085 18.43Dichloromethane (DCM) 1.60 26.5Chloroform 1.02 27.5Toluene 0.38 28.4Chlorobenzene (CB) 1.72 33.6o-Dichlorobenzene (o-DCB) 2.53 36.71-Chloronaphthalene (CN) 1.55 41.8Acetylaceton 3.03 31.2Diethyl ether 1.15 17

are clearly visible, corresponding to the so-called D and G bands.In particular the Raman spectrum of GO exhibited a D band peakat 1330 cm�1, that corresponds to the breathing mode of j-pointphonons of A1g symmetry and a G band peak at 1592 cm�1, dueto the first-order scattering of the E2g phonons [18]. The corre-sponding D and G bands in the Raman spectrum of rGO appearedat 1341 cm�1 and 1598 cm�1, respectively. The intensity of the Dband is related to the size of the in-plane sp2 domains and therelative intensity ratio (ID/IG) is a measure of the extent of disorder[18]. After the reduction of GO, the intensity ratio (ID/IG) wasincreased significantly and the higher intensity of D band suggeststhe presence of more isolated graphene domain in rGO comparedto GO and removal of oxygen groups from the latter [19].

Fig. 2b shows the FTIR spectra of GO and rGO. The peaks at�3400 cm�1 (OAH stretching vibrations), at �1700 cm�1 (C@Ostretching vibrations), at �1600 cm�1 (skeletal vibrations fromunoxidized graphitic domains), at �1200 cm�1 (CAOAC stretchingvibrations), at �1050 cm�1 (CAO stretching vibrations) are charac-teristic for the GO. The removal of oxygen-containing groups dur-ing the reduction is confirmed from the decrease (almostdisappearance) of the bands of C@O stretching, CAOAC stretching,CAO stretching. The relative decrease in the intensity of OAHstretching band indicates that CAOH still exist, but in lowerproportion.

lubility values for all solvents studied.

) dT (MPa1/2) GO Solubility (lg/mL) rGO Solubility (lg/mL)

47.8 6.6 4.7419.9 0.8 0.929.6 0.16 0.5226.5 0.25 0.9123.6 1.82 1.233 5.5 4.919.5 2.15 1.4424.9 1.96 1.7323 8.7 9.414.9 0.1 0.6120.2 0.21 1.1618.9 1.3 4.618.2 1.57 4.1419.6 1.62 3.420.5 1.91 8.9420.6 1.8 8.120.6 1.5 1.0215.6 0.72 0.4

Fig. 2. Characterization of rGO (a) Raman spectra, (b) FTIR spectra, (c) XRD patterns and (d) decomposition behaviour of GO and rGO.

110 D. Konios et al. / Journal of Colloid and Interface Science 430 (2014) 108–112

The XRD pattern (Fig. 2c) indicates a larger interlayer spacing inGO than in rGO. Water molecules, as well as the formation of oxy-gen-containing groups between the layers during the preparationof GO result in a lower angle reflection peak 2h = 9.32� (d-spacing�9.52 Å). The decrease in the interlayer spacing in rGO and theshift of the low peak at higher 2h angles (23.56�, d-spacing�3.77 Å) verify the efficient reduction by hydrazine method, dueto the more thorough removal of surface functional group.

TGA was used to further assess the level of reduction. Fig. 2ddisplays the TGA thermograms that show weight loss as a functionof temperature for dried-down GO and rGO. The evaporation of theabsorbed water molecules from room temperature to 200 �Ccauses a slight loss in the weight of GO, which is further decreaseddue to decomposition of oxygen-containing functional groups,losing in total approximately 40% of its mass up to 600 �C. Onthe contrary, rGO displays higher thermal stability, stemming fromthe deoxygenation during the reduction process.

Fig. 3. Digital picture of GO dispersions in 18 different solven

Digital pictures were taken to display the dispersion quality ofGO and rGO in different solvents, immediately after the sonication(Figs. 3 and 4). To identify the degree of sedimentation pictureswere again taken a day after. Just after sonication, GO showed verygood dispersion in NMP, DMF, ethylene glycol, THF and water.These five solvents exhibit significant dipole moment values,although o-DCB, which has similar dipole moment (2.53 D) failedto give a stable GO dispersion (Table 1). This suggests that solventpolarity is not the only factor for obtaining good dispersibility [20].

In previous reports, it has been shown that surface tension is animportant factor for choosing an effective solvent for graphene andits derivatives [8,21]. The existence of oxygen containing groups inthe GO results in higher surface energy, compared with the rGO inwhich the loss of surface polarity increases its hydrophobicity. Byperforming wettability and contact angle measurements, the sur-face energies of GO and rGO have been estimated to be �62 mN/m and �46 mN/m respectively [22]. Solvents with surface tension

ts. Top: immediately after sonication. Bottom: after 24 h.

Fig. 4. Digital picture of rGO dispersions in 18 different solvents. Top: immediately after sonication. Bottom: after 24 h.

Fig. 5. Digital picture of GO and rGO dispersions after 2 weeks, showing the long-term stability of different solutions.

D. Konios et al. / Journal of Colloid and Interface Science 430 (2014) 108–112 111

similar to the previous values are the most efficient solvents for thedispersion of GO and rGO. Our results (Figs. 3 and 4) confirmed thistheory, demonstrating improved dispersion behaviour of rGO in o-DCB, CN and CB compared with GO.

Following the Ruoff’s et al. approach, the Hansen solubilityparameters were used to investigate the dispersion mechanismof GO and rGO [15]. The theory takes into account the dispersioncohesion parameter (dD), the polarity cohesion parameter (dP),and the hydrogen bonding cohesion parameter (dH), which arecombined into the equation:

d2T ¼ d2

D þ d2P þ d2

H

to give the Hildebrand solubility parameter (dT) [23].To estimate the three Hansen parameters of GO and rGO, the

following equation was used:

hdii ¼P

solvCdi;solvPsolvC

where i = D, P, H or T, C is the GO and rGO solubility and di,solv is theith Hansen parameter in a given solvent [16]. For the studiedsolvents, the Hansen and Hildebrand parameters for GO were esti-mated to be hdDi �17.1 MPa1/2, hdPi �10 MPa1/2, hdHi �15.7 MPa1/2

and hdTi �25.4 MPa1/2. Our results are in good agreement with thepreviously reported Hansen parameters of GO [15]. The same modelwas used to estimate the respective parameters of rGO, which weremeasured to be hdDi �17.9 MPa1/2, hdPi �7.9 MPa1/2, hdHi �10.1MPa1/2 and hdTi �22 MPa1/2.

Owing to the presence of oxygen containing groups, the GO val-ues for polar and H-bonding components are higher than in the rGO.Similar values of the Hildebrand solubility parameter of solvent andsolute is an important criterion for choosing an efficient solvent.This explains the higher solubility values of rGO in chlorinated

solvents (DCM, CB, chloroform, o-DCB, CN) in contrast to the GO(Table 1).

The long-term stability was examined by leaving the suspen-sions undisturbed for three weeks (Fig. 5). The results clearly dis-played that GO retained its excellent solubility in NMP, whilethere was a slight increase in precipitation of GO in DMF, waterand ethylene glycol. It is worth mentioning that the GO showedlow but stable dispersibility in non-polar solvents, like toluene,chlorobenzene and o-DCB. Similar to GO, rGO gave very good dis-persions in NMP, water and ethylene glycol, which implies thatoxygen-containing functional groups are still present at defectsites. Thus, the relatively stable aqueous solutions of GO and rGOcan be attributed to the electrostatic repulsion due to thenegatively charged GO and rGO sheets, when dispersed in water.Furthermore, rGO presented greater interaction with non-polarsolvents (chloroform, toluene, chlorobenzene) than GO, but onlyin o-DCB and CN retained its solubility solutions of GO and rGOcan be attributed to the electrostatic repulsion due to the nega-tively charged GO and rGO sheets, when dispersed in water [24].Furthermore, rGO presented greater interaction with non-polarsolvents (chloroform, toluene, chlorobenzene) than GO, but onlyin o-DCB and CN retained its solubility.

3. Conclusions

The dispersion behaviour of GO and rGO in eighteen solventswas compared. The Hansen and Hildebrand parameters of GOand rGO were estimated verifying that the reduction process hasa strong effect on the solubility and stability. Solutions of GO inNMP, ethylene glycol and water presented significant long-termstability with solubility values reaching �8.7 lg/mL for NMP.While, the dispersion behaviour of GO changed after its reduction,

112 D. Konios et al. / Journal of Colloid and Interface Science 430 (2014) 108–112

presenting better interaction with solvents like o-DCB (�9 lg/mL)and CN (�8.1 lg/mL).

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.jcis.2014.05.033.

References

[1] A.K. Geim, K.S. Novoselov, Nat. Mater. 6 (2007) 183–191, http://dx.doi.org/10.1038/nmat1849.

[2] M.D. Stoller, S.J. Park, Y.W. Zhu, J.H. An, R.S. Ruoff, Nano Lett. 8 (2008) 3498–3502, http://dx.doi.org/10.1021/nl802558y.

[3] C.H. Lu, H.H. Yang, C.L. Zhu, X. Chen, G.N. Chen, Angew. Chem. Int. Ed. 48 (2009)4785–4787, http://dx.doi.org/10.1002/anie.200901479.

[4] E. Kymakis, K. Savva, M.M. Stylianakis, C. Fotakis, E. Stratakis, Adv. Funct.Mater. 23 (2013) 2742–2749, http://dx.doi.org/10.1002/adfm.201202713.

[5] S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H.R. Kim,Y.I. Song, Y.-J. Kim, K.S. Kim, B. Özyilmaz, J.-H. Ahn, B.H. Hong, S. Iijima, Nat.Nanotech. 5 (2010) 574–578, http://dx.doi.org/10.1038/nnano.2010.132.

[6] A. Dato, V. Radmilovic, Z. Lee, J. Phillips, M. Frenklach, Nano Lett. 8 (2008)2012–2016, http://dx.doi.org/10.1021/nl8011566.

[7] L. Rodríguez-Pérez, M.A. Herranz, N. Martín, Chem. Commun. 49 (2013) 3721–3735, http://dx.doi.org/10.1039/C3CC38950B.

[8] U. Khan, H. Porwal, A. O’Neill, K. Nawaz, P. May, J.N. Coleman, Langmuir 27(2011) 9077–9082, http://dx.doi.org/10.1021/la201797h.

[9] J.I. Paredes, S.V. Rodil, A.M. Alonso, J.M.D. Tascon, Langmuir 24 (2008) 10560–10564, http://dx.doi.org/10.1021/la801744a.

[10] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, Y.Wu, S.T. Nguyen, R.S. Ruoff, Carbon 45 (2007) 1558–1565, http://dx.doi.org/10.1016/j.carbon.2007.02.034.

[11] C. Petridis, K. Savva, Y. Lin, G. Eda, E. Kymakis, T.D. Anthopoulos, E. Stratakis,Appl. Phys. Lett. 102 (2013) 093115, http://dx.doi.org/10.1063/1.4794901.

[12] I.K. Moon, J. Lee, R.S. Ruoff, H. Lee, Nat. Commun. 1 (2010) 73–79, http://dx.doi.org/10.1038/srep01929.

[13] H. Feng, R. Cheng, X. Zhao, X. Duan, J. Li, Nat. Commun. 4 (2013) 1539–1546,http://dx.doi.org/10.1038/ncomms2555.

[14] S. Eigler, C. Dotzer, A. Hirsch, Carbon 50 (2012) 3666–3673, http://dx.doi.org/10.1016/j.carbon.2012.03.039.

[15] S. Park, J.H. An, I.W. Jung, R.D. Piner, S.J. An, X.S. Li, A. Velamakanni, R.S. Ruoff,Nano Lett. 9 (2009) 1593–1597, http://dx.doi.org/10.1021/nl803798y.

[16] Y. Hernandez, M. Lotya, D.R. Shane, D. Bergin, J.N. Coleman, Langmuir 26(2010) 3208–3213, http://dx.doi.org/10.1021/la903188a.

[17] F. Torrisi, T. Hasan, W. Wu, Z. Sun, A. Lombardo, T.S. Kulmala, G.-W. Hsieh, S.Jung, F. Bonaccorso, P.J. Paul, D. Chu, A.C. Ferrari, ACS Nano 6 (2012) 2992–3006, http://dx.doi.org/10.1021/nn2044609.

[18] Y.Y. Wang, Z.H. Ni, T. Yu, Z.X. Shen, H.M. Wang, J. Phys. Chem. C 112 (2008)10637–10640, http://dx.doi.org/10.1021/jp8008404.

[19] S. Thakur, N. Karak, Carbon 50 (2012) 5331–5339, http://dx.doi.org/10.1016/j.carbon.2012.07.023.

[20] S. Villar-Rodil, J.I. Paredes, A. Martínez-Alonso, J.M.D. Tascón, J. Mater. Chem.19 (2009) 3591–3593, http://dx.doi.org/10.1039/B904935E.

[21] Y. Hernandez, V. Nicolosi, M. Lotya, F.M. Blighe, Z.Y. Sun, S. De, I.T. McGovern,B. Holland, M. Byrne, Y.K. Gun’ko, J.J. Boland, P. Niraj, G. Duesberg, S.Krishnamurthy, R. Goodhue, J. Hutchison, V. Scardaci, A.C. Ferrari, J.N.Coleman, Nat. Nanotechnol. 3 (2008) 563–568, http://dx.doi.org/10.1038/nnano.2008.215.

[22] S.R. Wang, Y. Zhang, N. Abidi, L. Cabrales, Langmuir 25 (2009) 11078–11081,http://dx.doi.org/10.1021/la901402f.

[23] C.M. Hansen, Hansen Solubility Parameters: A User’s Handbook, second ed.,CRC Press, Hoboken, 2007.

[24] D. Li, M.B. Muller, S. Gilje, R.B. Kaner, G.G. Wallace, Nat. Nanotechnol. 3 (2008)101–105, http://dx.doi.org/10.1038/nnano.2007.451.