Hydrothermal deoxygenation of graphene oxide in sub- and supercritical water

Abstract

Graphene oxide (GO), an oxidized form of graphene, exhibits immense potential for a wide range of applications owing to its rich chemistry. This work reports the controlled deoxygenation of GO under sub- and supercritical hydrothermal conditions, which are considered to be foremost green, environmentally-friendly and economically viable. The remarkable thermo-physical and chemical properties of water, monitored by temperature (373–653 K) and pressure (0.04–22.75 MPa), facilitate the deoxygenation of GO. The gradual chemical and structural changes in GO in hydrothermal reactions, over a wide range of temperature and pressure are elucidated using XPS, FTIR, Raman, XRD, and HRTEM analyses. Plausible deoxygenation mechanisms, particularly elimination of hydroxyl, epoxide, carboxyl, and carbonyl groups and repairing of the π-conjugated network are discussed on the basis of spectroscopic analyses. The addressed hydrothermal route not only avoids the use of toxic and hazardous chemicals as reducing agents but also regulates the deoxygenation events.

---

Introduction

In recent years, graphene oxide (GO) has emerged as a most promising precursor to reduced graphene oxide (rGO), chemically functionalized graphene and graphene based composites. The mass-scale production of graphene and graphene-based materials by wet chemical processing of GO is a widely acceptable method, among all known approaches like micromechanical exfoliation of graphite, chemical vapor deposition (CVD), epitaxial growth, unzipping of carbon nanotubes, etc.^1–3 The GO, a nonstoichiometric and hygroscopic carbon material, contains ample oxygen carrying functionalities viz. hydroxyl, epoxide, phenolic, carbonyl, carboxyl, ethers, quinones, lactols etc.^4,5 These oxygen functionalities are not only disrupt the π-conjugated network of two-dimensional graphitic lattice but also make this material hydrophilic and insulator. The elimination of oxygen functionalities and restoration of π-conjugated network are very important in order to attain the graphene characteristics.

The rGO, resemble to graphene, can be prepared via chemical reduction using strong reducing agents like hydrazine hydrate, borohydride, hydroquinone, lithium aluminum hydride etc.^6,7 Considering the toxic effects of these reducing agents and contamination of product by hetero-atoms; alternately, photo-catalytic, thermal and electrochemical reduction methodologies have been developed.^8–10 In last five years, there is rapid progress in this area and many wet chemical recipes have been developed to prepare the rGO using vitamin C, saccharides, green tea, alcohol, amino acids, alkaline conditions etc. as green reducing media.^11–13 However, graphene community still face several challenges such as (a) understanding the deoxygenation events during the reduction of GO, (b) preparation of good quality and dispersible graphene, (c) controlling the deoxygenation events during reduction process and (d) usage of environmentally friendly and cost-effective reducing agents etc.

Herein, we report the controlled deoxygenation of GO under sub- and supercritical conditions using water as a reaction media. The gradual chemical and structural changes in GO over the wide range of hydrothermal temperature (373–653 K) are elucidated and the plausible deoxygenation events are discussed. The use of aqueous chemistry is considered to be most green and economic media for deoxygenation of GO.

Experimental section

Preparation of graphene oxide

Graphite oxide, a precursor to graphene oxide (GO), was prepared by harsh oxidation of graphite powder using a mixture of NaNO₃, H₂SO₄ and KMnO₄ as strong oxidizing reagents. Then the oxidized product was treated with 30% solution of H₂O₂, in order to digest the un-reacted content of KMnO₄. Subsequently, the product was rinsed and washed with copious amounts of water to remove the excess content of soluble ions until an acidic pH was observed for the decanted water. The processed dark brown oxidized material, known as graphite oxide, was then exfoliated into the GO using an ultrasonic probe. The exfoliated product contains basically two types of materials on the basis of their chemical and structural features. Hence, the exfoliated product was centrifuged at 3000 rpm for 15 minutes, which leads to two distinct phases; upper phase containing dispersible fine sheets of GO and lower deposited semi-solid phase. The finer fraction (upper phase) of GO was used for the hydrothermal treatment.

Hydrothermal deoxygenation of graphene oxide

An aqueous dispersion of GO (5 mg mL⁻¹) was loaded in the Parr reactor. The reaction vessel was filled ∼60% of total capacity and then executed the hydrothermal treatment. The hydrothermal reactions was carried out at four different temperatures (373, 473, 573, and 653 K); simultaneously the autogenous pressure generated due to high temperature, was noted. The hydrothermal reactions were also carried out for a variable time (2 to 12 hours). After the reactions under sub- and supercritical conditions, the product was washed with copious amount of water and then separated by filtration. The colour of brown GO was found to change into black, which was an instant indication of deoxygenation of GO. The product was then dried in an oven and used for chemical and structural analyses.

Chemical and structural characterizations

Fourier transform infrared (FTIR) spectra of GO and hydrothermally treated samples (KBr pellets) were recorded using a Thermo-Nicolet 8700 Research spectrophotometer with a 4 cm⁻¹ resolution. X-ray photoelectron spectroscopy (XPS, JPS-9010TRX, JEOL Ltd.) measurements were carried out using thin films of GO and hydrothermally treated samples and monitored the chemical changes as a function of hydrothermal-temperature. All XPS measurements were executed using a MgKα line as the X-ray source. The Peak-fitting of the C1s spectra for all samples was carried out by using a Gaussian–Lorentzian function after performing a Shirley background correction. Raman spectra of all samples were collected using a Renishaw micro-spectrometer at an excitation wavelength of 514.5 nm. The fluorescence background for the Raman spectra was subtracted using the Wire 2.0 Renishaw software. The Raman shifts for G and D bands are based on three different measurements. The powder X-ray diffraction (XRD) analyses of all samples were carried out using a Bruker D8 Advance diffractometer at 40 kV and 40 mA with CuKα radiation (λ = 0.15418 nm). The diffraction data were recorded for 2θ angles between 2° and 60° (step size 0.02°, step time 1 s). The BET surface area, pore volume and pore diameter of hydrothermally treated samples were calculated by their N₂ adsorption/desorption isotherms at 77 K using a BelsorbMax, BEL, Japan.

Results and discussion

Interest in the use of water as reaction media at high temperatures is because of its unique thermo-physical and chemical properties, which significantly changes with temperature and pressure.¹⁴ As the temperature of water increases from 298 to 573 K, the ionic product (K_w) increases by three orders of magnitude (from 10⁻¹⁴ to 10⁻¹¹), which accelerates the ionic chemistry. Over the same temperature range, the density of water decreases from 0.997 to 0.713 g cm⁻³ and its dielectric constant (hydrogen bonding structure) decreases from 78.9 to 19.7. As a result, polar and hydrogen networked water at 298 K transforms to more typical of non-polar solvent at 573 K.¹⁵ Herein, GO was treated with sub- and supercritical water at four different temperatures (Table 1). In a typical hydrothermal reaction, the closed vessel reactor was filled ∼60% of the total volume to allow for maintaining a liquid water phase under autogenous pressure for expansion of water. Under these conditions, compressed water facilitate and execute the deoxygenation events by heterolytic bond cleavage.

Table 1 Hydrothermal reaction parameters

No.	Hydrothermal reaction conditions			Sample code
	Temp., K	Pressure, MPa	Time, hours
Subcritical conditions
I	373	0.04	2	GO3732
	373	0.04	6	GO3736
	373	0.04	12	GO37312
II	473	1.38	2	GO4732
	473	1.38	6	GO4736
	473	1.38	12	GO47312
III	573	8.46	2	GO5732
	573	8.46	6	GO5736
	573	8.46	12	GO57312
Supercritical conditions
IV	653	22.75	2	GO6532

---

The exposure of GO to sub- and supercritical water leads to significant chemical changes as revealed by FTIR and XPS analyses. The freshly prepared GO exhibits a strong and broad vibration peak centred at ∼3415 cm⁻¹, is attributed to O–H stretch of hydroxyl and phenolic groups in the materials. The intensity of O–H stretch is reduced significantly with increasing the hydrothermal temperature as shown in Fig. 1a. The residual intensity of O–H peak in the GO5732 and GO6532 samples is associated to the thermally stable phenolic groups. Hamilton et al. have addressed that phenol groups are thermally stable, when GO is subjected to the thermal reduction process.^9a A broad vibrational signature at 1725 cm⁻¹, attributed to the C=O stretch, representing the both carboxyl and carbonyl groups.^5,16–18 Recently, it has been noted that carbonyl groups can survive at higher temperature, hence the residual signature in GO5732 and GO6532 samples is associated to the thermally stable carbonyl groups.^9a,17 Further, a vibrational peak at 1620 cm⁻¹ is attributed to an overlapped signature of bending modes of trapped water molecules in the material and C=C stretch of unoxidized sp² carbon domains.^5,8,16,17 In general, the identification of sp²-hybridizied C=C groups at 1580–1600 cm⁻¹ is greatly complicated by the scissor mode of water at 1610–1640 cm⁻¹. Hence, the presence of sp² carbon domains was further corroborated by ¹³C SSNMR spectra (Fig. S1, ESI†), which shows a strong and broad nuclear resonance shifts at ∼131 ppm, assignable to the sp² carbon in the GO skeleton.⁴ The shifts of FTIR peak at 1620 cm⁻¹ towards the low wavenumber illustrating the elimination of trapped water molecules and an increase of peak intensity reveals the restoration of sp² carbon in the hydrothermally treated samples. The O–H deformations of the C–OH groups are appeared at 1355 cm⁻¹. An unresolved broad peak in the range of 1270–1210 cm⁻¹ is associated to C–O–C stretch owing to epoxy and ether groups.^5,17 The epoxy groups are known to be thermally unstable and eliminated during the hydrothermal treatment, whereas ether groups are found be hydrothermally stable. Shenoy et al. have revealed the formation of thermally stable ether groups during the thermal reduction process, which hindered the its complete deoxygenation.^5,9a The vibrational peak at 1213 cm⁻¹ revealed the ethers groups, which are not only thermally stable, but also emerges during the reduction process.^5,9a Furthermore, Li et al., who carried out the dehydration of polysaccharides, have also shown that ether linkage are thermally stable and emerged during the hydrothermal exercise.¹⁹ The FTIR results revealed that degree of deoxygenation events are increased with increasing the hydrothermal temperature and pressure, attributed to the thermo-chemical effect of pressurized water. Further, the effect of hydrothermal treatment time on deoxygenation of GO was also evaluated. Under superheated water (473 and 573 K), the degree of deoxygenation found to be very high at initial hours, and within six hours most of oxygenated functionalities were eliminated, except the thermally stable ether and phenolic moieties; simultaneously the intensity owing to C=C stretch increased significantly. However, under supercritical condition (653 K), most of oxygen functionalities were eliminated within two hours (ESI Fig. S2†).

Fig. 1 FTIR spectra of GO and its deoxygenated samples, prepared by hydrothermal treatment in sub- and supercritical water. (a) Effect of hydrothermal temperatures (373, 473, 573, and 653 K) for two hours, and (b) effect of hydrothermal time (2, 6, and 12 hours) at temperature of 473 K.

Furthermore, high-resolution XPS spectra of C1s signature of GO and hydrothermally-treated samples revealed a significant reduction of oxygenated carbon as illustrated in Fig. 2. The C1s spectra of GO exhibited an overlapped double peak structure with a small tail towards higher binding energy. The C1s spectra could deconvoluted into four different chemically shifted components at 284.5, 286.6, 288.2, and 289.3 eV and are assigned to the predominantly C–C/C=C skeleton, hydroxyl/epoxide/phenolic/ether (C–OH/C–O–C), carbonyl/quinone (C=O) and carboxyl (COOH) groups, respectively.^9a,20 The double-peak structure of C1s in GO gradually changes as a function of hydrothermal temperature. An apparent indication of deoxygenation of the GO was deduced by a shift of the C1s peak maxima towards lower binding energy. The improved peak intensity due to C=C component at 284.5 eV in hydrothermally treated samples revealed the regeneration of sp² carbon. Whereas, C–OH/C–O–C peak component at 286.6 eV in hydrothermally treated samples revealed the presence of thermally stable phenolic and ether groups, which is further corroborated by their FTIR spectra. The higher residual intensity owing to C=O peak component at 288.2 eV in hydrothermally treated samples show that carbonyl groups are thermally stable and couldn't eliminated during hydrothermal treatment. The C1s spectra of supercritical water-treated (653 K) GO shows a broadened tail associated to π–π* component at 290.2 eV. This is well-recognized shakeup satellite peak and indicates that the delocalized π conjugation, a characteristic of graphitic carbon, is restored in supercritical water-treated GO (GO6532).^9a Further, the relative percent intensity associated with C–C component, increases with increasing hydrothermal temperature, whereas the relative percent intensities of carbon bonded to the oxygen-carrying functionalities, decreases substantially (Fig. 2f). These results illustrate that deoxygenation of GO can be carried out under sub- and supercritical conditions using water as a green medium and degree of such deoxygenation events could be regulated by monitoring the hydrothermal temperature.

Fig. 2 High-resolution C1s XPS spectra of (a) GO and its deoxygenated samples, prepared by hydrothermal treatment at temperature of (b) 373, (c) 473, (d) 573, and (e) 653 K for two hours. The deconvoluted components illustrating various types of oxygen functionalities in the materials. (f) Changes in the relative% intensities of the C–C carbon and carbon bonded with oxygen functionalities over the hydrothermal-treatment temperature. The relative% intensity for each component was calculated by dividing the peak area of respective component by the whole C1s peak area.

During the hydrothermal process, significant structural changes occurred in GO due to the elimination of oxygen functionalities. As a result, basal plane carbon structure also rearranges, which is monitored by Raman spectroscopy taken after each hydrothermal step. In general, graphite exhibits two characteristics Raman-active bands; graphitic lattice (G) band at 1575 cm⁻¹ and disorder band (D band) owing to defects and edges carbon at 1355 cm⁻¹.²¹ Raman spectrum of GO exhibit these two characteristics bands at ∼1604 and ∼1354 cm⁻¹ correspond to G and D modes, respectively (Fig. 3).²² A shift in the G band towards higher wave numbers is attributed to the presence of isolated sp² network separated by oxygen functionalities, defects on the carbon network, and the reduced number of layers in the GO. The hydrothermal deoxygenation of GO gradually shifts the G band towards low wave number revealing the recovery of hexagonal network (sp²-carbon domains) and such phenomenon increases with increasing the hydrothermal temperature as shown in Fig. 3b. Furthermore, the structural changes during hydrothermal treatment of GO were elucidated by XRD analysis. The oxygen functionalities within the basal planes of GO along with the trapped water molecules afforded a high d-spacing, as revealed by diffraction peak at 2θ = 10.8° (Fig. 4). The gradually shifts of XRD characteristic peak towards the 2θ = 24–26° and broadening of peak with diffused structure indicates the elimination of oxygen functionalities during the hydrothermal treatment. As a result, interlayer distance in hydrothermally treated samples reduced significantly. The broad and diffused nature of the diffraction peak for GO5732 and GO6532 samples illustrate the poor packing of sheets and the limited number of layers within the sheet.^6b This is further supported by the BET surface area of hydrothermally treated samples (Fig. 5).

Fig. 3 (a) Raman spectra of GO and its deoxygenated samples, prepared by hydrothermal treatment at temperature of 373, 473, 573, and 653 K for two hours. (b) Raman shifts of the G and D bands as a function of hydrothermal temperature.

Fig. 4 XRD patterns of GO and its deoxygenated samples, prepared by hydrothermal treatment at temperature of 373, 473, 573, and 653 K for two hours.

The nitrogen adsorption/desorption isotherm, BET surface area and pore volume of hydrothermally treated samples are shown in Fig. 5. The BET surface area of GO was found to gradually increase with increasing the hydrothermal temperature. The GO sheets, which are tightly packed by hydrogen bonding network between the oxygenated functionalities showed very low BET surface area i.e. 30 m² g⁻¹. The elimination of these oxygen functionalities in GO, facilitated by hydrothermal treatment, leads to agglomeration in the partial overlapping and coalescing of the deoxygenated sheets. Simultaneously, the high temperature deoxygenation events create enormous pressure within the stacked layers of GO. Evaluation of the Hamaker constant predicts that a pressure of only 2.5 MPa is necessary to separate two stacked graphene oxide platelets.²³ Herein, the high pressure generated owing to increasing temperature in the closed reaction vessel not only facilitate the deoxygenation events but also exfoliate these sheets with lot of crumpled structural features (HRTEM images, Fig. 6), which increases with increasing the hydrothermal temperature. Therefore, the BET surfaces area and pore volume of hydrothermally treated samples increase with increasing the hydrothermal treatment temperature (Fig. 5b). The pore size distribution as calculated by DFT method revealed the presence of both mesopores and microspores in hydrothermally treated samples. Most likely, the mesopores are developed in the hydrothermally treated samples owing to incompact stacking, fluffy nature of the material, entanglement and overlaying of the materials, while the microspores come from the defects and holes, generated during the hydrothermal treatment.^24,25 The significant mass loss as a function of increasing hydrothermal temperature, further revealed the biting of carbon skeleton through formation of CO₂, CO etc.

Fig. 5 (a) Nitrogen adsorption/desorption isotherms of deoxygenated samples of GO, prepared by hydrothermal treatment at temperature of 373, 473, 573, and 653 K for two hours. (b) BET surface area and pore volume of deoxygenated samples of GO as a function of hydrothermal temperature.

Fig. 6 HRTEM images of (a) GO and its deoxygenated samples, prepared by hydrothermal treatment at temperature of (b) 373, (c) 473, (d) 573, and (e) 653 K for two hours.

Both GO and hydrothermally treated samples are efficiently exfoliated into very thin sheets. The microstructural images of GO (Fig. 6a), as examined by HRTEM, show wrinkled and partially folded sheets with limited number of layers within the sheet. The sp³-carbon centres associated with the oxygen functionalities and various structural defects in the basal plane of GO, disturbed the two-dimensional structure, thereby resulting in a roughened surface with many wrinkles and folded regions. The significant structural changes are observed when GO is subjected to the hydrothermal treatment. The microstructural images of hydrothermally treated samples illustrate that sheets are highly aggregated with a fluffy appearance. This could be due to (a) presence of residual oxygen functionalities along with defects, (b) ether linkages between the sheets, and (c) weakening of hydrogen bonding network between these sheets. However, HRTEM images of hydrothermally treated samples show an interlayer distance of ∼0.35 nm, which is a characteristic interlayer distance of reduced graphene oxide.²⁶

The cleavage of heteroatom linkage (C–O) occurred most readily under hydrothermal conditions, when organic materials such as alcohols, glucose, cellulose etc. are exposed to the superheated water.¹³ In general, deoxygenation reactions in sub- and supercritical water are accelerated by compressed water owing to its unique thermo-physical and chemical properties. The presence of multiple functional groups on both side of GO nanosheets provide structural strained, whose structure and chemistry evolve during hydrothermal exercise by various chemical pathways as shown in Scheme 1. The deoxygenation process, particularly for hydroxyl groups is believed to be analogous to the dehydration of alcohol, where the compressed water act as a source of hydrogen ion owing to its high ionic product (K_w).¹⁴ The hydrogen ions initiate the dehydration process by abstracting adjacent OH groups via either inter- or intramolecular reactions.²⁷ The intramolecular dehydration generates ether linkage between two different nanosheets and aggregates them (Scheme 1, position 4). The closely placed hydroxyl groups also generate ether and epoxide linkage, depending on their position in the GO skeleton, and eliminates water molecules (Scheme 1, positions 1 and 3). Further these plausible pathways are supported by presence of strong vibrational signature of C–O–C stretch at 1265–1220 cm⁻¹ (attributed to ether linkage) in FTIR spectra of GO4732, GO5732, and GO6532 samples. Recently, Riedo et al. have revealed that both epoxide and hydroxyl groups can abstract by consuming the hydrogen atom from adjacent C–H linkage.²⁸ The epoxide groups placed in the basal plane of GO are highly strained. The presence of nearby hydrogen atom to epoxide groups facilitates their reduction to hydroxyl groups under the hydrothermal conditions and generates the π bond (Scheme 1, position 7). The hydroxyl groups having the adjacent hydrogen atom could generate the olefinic bond by eliminating the water molecule (Scheme 1, position 2). The intensity of carboxyl groups in GO were decreased with increasing the hydrothermal temperature as deduced by their FTIR and XPS analyses. These results revealed that, probably, the carboxyl groups might have decomposed at high temperature by eliminating the CO₂ molecules (Scheme 1, position 5). Recent study based on HRTEM and scanning tunnelling microscope imaging have shown the porous nature of GO.^25,29 The generation of these holes and biting of edges in the GO skeleton indicates the consumption of carbon atom (Scheme 1, position 8). The elimination of these oxygen functionalities in GO not only makes this material thermodynamically stable but also restores the conjugated π-network. The addressed reaction pathways for deoxygenation of GO are supported by their spectroscopic results. However, considering the wide spectrum of oxygen functionalities and their distribution in GO, there could be more plausible pathways, which need to be addressed for understanding the full deoxygenation mechanism.

Scheme 1 A plausible mechanism for the chemical changes in the GO during hydrothermal treatment using sub- and supercritical water. (a) Dehydration to cyclic ether (position 1) by eliminating the water molecule, (b) dehydration to olefinic bond (position 2) by consuming hydroxyl group and adjacent hydrogen atom, (c) dehydration to epoxide by (position 3) consuming two adjacent hydroxyl group, (d) intermolecular dehydration (position 4) to ether between two different GO nanosheets by eliminating the water molecule, (e) decarboxylation (position 5) by consuming the carboxyl group, (f) reduction of carbonyl group into the hydroxyl group (position 6) with generation of olefinic bond, (g) epoxide-to-hydroxyl conversion (position 7) by consuming adjacent hydrogen atom, and (h) decarboxylation (position 8) by consuming carbon from the GO skeleton.

Conclusion

In conclusion, the controlled deoxygenation of GO under sub- and supercritical conditions using water as a reaction media was thoroughly investigated based on spectroscopic measurements. At high temperatures (373–653 K) and pressure (0.04–22.75 MPa), the GO experiences the deoxygenation by eliminating the various oxygen functionalities. The hydroxyl and epoxide groups are observed to be thermally unstable and eliminated during the hydrothermal treatment, where as the ether and phenols groups are comparatively stable and also emerges during the hydrothermal treatment as deduced by FTIR and XPS results. At the high temperature (473–653 K), the degree of deoxygenation is found to be higher compared to that of at moderate temperature (373 K). This is associated to the unique thermo-physical and chemical properties of water at high temperature and pressure, where the cleavage of heteroatom linkage (C–O) occurred most readily. The partial restoration of π-conjugated network in the hydrothermally treated samples was confirmed by gradual shifts of G band toward the lower wavenumber with increasing the hydrothermal treatment temperature. On the basis of spectroscopic results, the plausible mechanism for deoxygenation of GO under hydrothermal conditions could be the hydrogen ion initiated dehydration by inter- or intramolecular reactions, reduction of highly strained epoxide groups, decarboxylation and generation of conjugated π-network. The developed hydrothermal approach is very important to prepare reduced graphene oxide, having controlled degree of oxygen functionalities, for their use in various applications including polymeric composite formation, chemically functionalized GO preparation, thin film preparation and so forth. The additional positive aspects of the use of aqueous chemistry are its simplicity, low cost, and favourable environmental impact.

Acknowledgements

This work was generously supported by the CSIR and DST-JSPS. Authors kindly acknowledge ASD of IIP Dehradun, IISc Bangalore, and JNCASR Bangalore for providing help in the analyses of the samples. Author HPM is thankful to UGC for financial support.

Notes and references

1. (a) M. J. Allen, V. C. Tung and R. B. Kaner, Chem. Rev., 2010, 110, 132–145; (b) W. Choi, I. Lahiri, R. Seelaboyina and Y. S. Kang, Critical Rev. Solid State Mater. Sci., 2010, 35, 52–71.
2. D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. Soc. Rev., 2010, 39, 228–240.
3. C. K. Chua and M. Puram, Chem. Soc. Rev., 2014, 43, 291–312.
4. W. Gao, L. B. Alemany, L. Ci and P. M. Ajyan, Nat. Chem., 2009, 1, 403–408.
5. (a) A. Bagri, C. Mattevi, M. Acik, Y. J. Chabal, M. Chhowalla and V. B. Shenoy, Nat. Chem., 2010, 2, 581–587; (b) S. Chudhary, H. P. Mungse and O. P. Khatri, J. Mater. Chem., 2012, 22, 21032–21039.
6. (a) S. Stankovich, D. A. Dikin, R. D. Piner, K. A.Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu, S. T. Nguyen and R. S. Ruoff, Carbon, 2007, 45, 1558–1565; (b) H. J. Shin, K. K. Kim, A. Benayad, S. M. Yoon, H. K. Park, I. S. Jung, M. H. Jin, H. K. Jeong, J. M. Kim, J. Y. Choi and Y. H. Lee, Adv. Funct. Mater., 2009, 19, 1987–1992.
7. (a) A. Ambrosi, C. K. Chua, A. Bonanni and M. Pumera, Chem. Mater., 2012, 24, 2292–2298; (b) G. Wang, J. Yang, J. Park, X. Gou, B. Wang, H. Liu and J. Yao, J. Phys. Chem. C, 2008, 112, 8192–8195.
8. V. Abdelsayed, S. Moussa, H. M. Hassan, H. S. Aluri, M. M. Collinson and M. S. El-Shall, J. Phys. Chem. Lett., 2010, 1, 2804–2809.
9. (a) A. Ganguly, S. Sharma, P. Papakonstantinou and J. Hamilton, J. Phys. Chem. C, 2011, 115, 17009–17019; (b) R. Larciprete, S. Fabris, T. Sun, P. Lacovig, A. Baraldi and S. Lizzit, J. Am. Chem. Soc., 2011, 133, 17315–17321.
10. H. L. Guo, X. F. Wang, Q. Y. Qian, F. B. Wang and X. H. Xia, ACS Nano, 2009, 3, 2653–2659.
11. J. I. Paredes, S. Villar-Rodil, M. J. Fernandez-Merino, L. Guardia, A. Martinez-Alonso and J. M. D. Tascon, J. Mater. Chem., 2011, 21, 298–306.
12. (a) Y. Wang, Z. Shi and J. Yin, ACS Appl. Mater. Interfaces, 2011, 3, 1127–1133; (b) D. R. Dreyer, S. Murali, Y. Zhu, R. S. Ruoff and C. W. Bielawski, J. Mater. Chem., 2011, 21, 3443–3447; (c) S. Choudhary, H. P. Mungse and O. P. Khatri, Chem.–Asian J., 2013, 8, 2070–2078.
13. X. Fan, W. Peng, Y. Li, X. Li, S. Wang, G. Zhang and F. Zhang, Adv. Mater., 2008, 20, 4490–4493.
14. (a) M. Siskin and A. R. Katritzky, Chem. Rev., 2001, 101, 825–835; (b) A. A. Peterson, F. Vogel, R. P. Lachance, M. Froling, M. J. Antal and J. W. Tester, Energy Environ. Sci., 2008, 1, 32–65.
15. A. Kruse and E. Dinjus, J. Supercrit. Fluids, 2007, 39, 362–380.
16. C. Su, M. Acik, K. Takai, J. Lu, S. Hao, Y. Zheng, P. Wu, Q. Bao, T. Enoki, Y. J. Chabal and K. P. Loh, Nat. Commun., 2012, 3, 1298.
17. M. Acik, G. Lee, C. Mattevi, A. Pirkle, R. M. Wallace, M. Chhowalla, K. Cho and Y. Chabal, J. Phys. Chem. C, 2011, 115, 19761–19781.
18. (a) S. Verma, H. P. Mungse, N. Kumar, S. Choudhary, S. L. Jain, B. Sain and O. P. Khatri, Chem. Commun., 2011, 47, 12673–12675; (b) H. P. Mungse, S. Verma, N. Kumar, B. Sain and O. P. Khatri, J. Mater. Chem., 2012, 22, 5427–5433.
19. X. Sun and Y. Li, Angew. Chem., Int. Ed., 2004, 43, 597–601.
20. C. Mattevi, G. Eda, S. Agnoli, S. Miller, K. A. Mkhoyan, O. Celik, D. Mastrogiovanni, G. Granozzi, E. Garfunkel and M. Chhowalla, Adv. Funct. Mater., 2009, 19, 2577–2583.
21. F. Tuinstra and J. L. Koenig, J. Chem. Phys., 1970, 53, 1126.
22. K. N. Kudin, B. Ozbas, H. C. Schniepp, R. K. Prud'homme, I. A. Aksay and R. Car, Nano Lett., 2008, 8, 36–41.
23. M. J. McAllister, J. L. Li, D. H. Adamson, H. C. Schniepp, A. A. Abdala, J. Liu, M. Herrera-Alonso, D. L. Milius, R. Car, R. K. Prud'homme and I. A. Aksay, Chem. Mater., 2007, 19, 4396–4404.
24. B. Xu, S. Yue, Z. Sui, X. Zhang, S. Hou, G. Cao and Y. Yang, Energy Environ. Sci., 2011, 4, 2826–2830.
25. K. P. Loh, Q. Bao, G. Eda and M. Chhowalla, Nat. Chem., 2010, 2, 1015–1024.
26. K. Moom, J. Lee, R. S. Ruoff and H. Lee, Nat. Chem., 2010, 1, 73.
27. Y. Zhou, Q. Bao, L. A. L. Tang, Y. Zhong and K. P. Loh, Chem. Mater., 2009, 21, 2950–2956.
28. S. Kim, S. Zhou, Y. Hu, M. Acik, Y. J. Chabal, C. Berger, W. Heer, A. Bongiorno and E. Riedo, Nat. Mater., 2012, 11, 544–549.
29. K. Erickson, R. Erni, Z. Lee, N. Alem, W. Gannett and A. Zettl, Adv. Mater., 2010, 22, 4467–4472.

---

Footnote

† Electronic supplementary information (ESI) available: XPS, FTIR, and TGA results. See DOI: 10.1039/c4ra01085j

---