1. Introduction
GO can be obtained in large quantities as colloidal dispersions in water that are considered promising precursors for the mass production of chemically RGO [
1,
2,
3]. RGO is widely used in making conductive films, heat-dissipation films can be spun into high-performance graphene fibers, energy storage, supercapacitors, etc., [
4,
5,
6,
7,
8,
9,
10], due to its unique structure [
11]. Numerous chemical methods for the reduction of graphene oxide have been reported [
12,
13,
14,
15,
16,
17]. A majority of previous studies have used hydrazine, sodium borohydride, and vitamin C (Vc) as reducing agents in aqueous or single organic solvents [
18,
19,
20,
21], where vitamin C is a natural antioxidant, and the degree of reduction is comparable to that of hydrazine reducing agents [
22,
23,
24]. On the other hand, Hernandez et al. [
25] obtained high-quality unoxidized monolayer graphene at a yield of about 1 wt.% by physical–mechanical exfoliation with dispersion and exfoliation in NMP. The authors confirmed the presence of 11 wt.% NMP residues in graphene sheets by using photoelectron spectroscopy analysis of vacuum-dried graphene sheets at room temperature. The value remained unchanged after a subsequent vacuum thermal treatment at 400 °C. The nanomaterial graphene has very strong surface activity. The surface/interface properties of graphene were reviewed by Zhao et al. [
26] graphene, and its derivatives can be combined with polymers through non-covalent and covalent interactions, which had stronger interactions at the interface, tending to form extremely strong chemical bonds with polymers and even exceed chemical bonds. Wang et al. [
27] investigated the difference in organic pollutant adsorption on GO and RGO to explore the potential adsorption mechanism. It was found that solution chemistry parameters affected the surface chemistry and aggregation properties of graphene, which, in turn, affected its interaction with organic contaminants. Coexisting surfactants also had different effects on the adsorption of polar and nonpolar aromatics on graphene.
Vipul Agarwal [
28] reviewed the chemical reduction of graphene in solution within the last decade or so, with conductivity generally in the range of 10–7700 S m
−1 and a C/O ratio between 3.85 and 15.1. The conductivity of graphene is mainly determined by the long-range conjugate network of the graphite lattice [
29,
30]. It is obvious that the chemically RGO prepared in solution is not ideal in conductivity at present. Initially, chemical reduction of graphene oxide was mostly conducted in water, where GO could be well dispersed, but the reduction product RGO could easily precipitate, leading to low reduction [
31,
32]. The introduction of organic weakly polar solvents aspired to minimize the generation of precipitation, and the final RGO obtained was also difficult to disperse well, due to the reduction of oxygen-containing functional groups on the surface after reduction [
33]. Accordingly, we observed that the well-dispersed GO in water or a single polar organic solvent was transformed into hydrophobic RGO after reduction [
34]. The change in hydrophilicity before and after reduction can easily lead to the aggregation preventing the reduction of the functional groups in the base of the layers from the reducing agent, resulting in a further incomplete reduction. It is difficult to ensure the good dispersion of both with a single reaction medium. Su et al. [
35] further developed the reduction of graphene oxide in mixed solvents; however, the reduction reaction was a dynamic process, and the purely mixed solvents could not be perfectly matched with it, and some degree of aggregation might occur during the reduction process. The stable dispersion of RGO can be ensured by gradually changing the polarity of the solvent system, which facilitated a more complete reduction of graphene oxide, resulting in a good dispersion of RGO. Moreover, in terms of the unusual structure of graphene, its large π-delocalized sp
2-bonded carbon atoms are densely arranged in planar sheets in a honeycomb lattice, giving rise to a huge specific surface area [
36]. It is in parallel with the usual nanomaterials that are prone to expose obvious physical or chemical adsorption on this high-energy surface [
37], which subsequently has a huge impact on the conductivity of graphene.
To deal with the three major problems of aggregation, incomplete reduction, and surface adsorption contamination in aqueous or single solvent systems, we designed a strategy of gradient solvent system. The gradual solvent system matched the hydrophobic change of the entire reduction process and the intermediate state of the GO sheet reduction structure change, thus maintaining an agglomeration-free reduction process throughout. Secondly, based on the gradual solvent system, a high concentration of reducing agents and high temperature were used to try to improve the degree of reduction. Furthermore, we used different dispersants and solvents as representatives of different types of adsorption, and then successive solvent rinsing, thermal solvent extraction, and high-temperature thermal treatment were used for the desorption study. The results showed that the reduction of GO in the gradient solvent system was without aggregation. The obtained RGO can be redispersed into NMP, with the particle size as low as about 200 nm, and the conductivity of RGO can reach 18,000 S m−1 after thermal treatment.
2. Experimental
2.1. Materials
Natural crystalline flake graphite (NG) (99.85% purity, 325 mesh), sulfuric acid (H2SO4, 98%), hydrochloric acid (HCl, 36%), anhydrous ethanol (≥99.7%), 2-methoxy ethanol (EGM, 99%), ascorbic acid (Vc ≥ 99%), hydrogen peroxide (H2O2, 30%), potassium permanganate (KMnO4), xylene, N-methyl-pyrrolidone (NMP), cyclohexane, polyvinyl pyrrolidone 30 (PVP30), polystyrene (PS), glycol phthalate (DEP), and anhydrous ethanol (C2H6O) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Alkyl-phenol polyoxyethylene (7) ether (OP-7) was purchased from Wen Hua Chemical Reagent Factory (Tianjin, China), and fatty alcohol polyoxyethylene ether (AEO-3) was purchased from Jinan Maifeng Chemical Co. (Jinan, China).
2.2. Preparation of Graphite Oxide
GO was prepared by a modified Hummers method. The specific steps were as follows: 1 g of graphite powder was mixed with 78 mL of concentrated sulfuric acid and 12 mL of deionized water and stirred for 30 min. Then 2.5 g of potassium permanganate (KMnO4) was added very carefully below 5 °C. The stirring was maintained at 50 °C for 12 h. Then 100 mL of deionized water was slowly added dropwise, with vigorous stirring, below 70 °C. The mixture gradually turned into a paste and light brown solution. Subsequently, 30% H2O2 was added slowly dropwise to the mixture until no bubbles appeared. The mixture was washed by rinsing with 10% HCl and centrifugation, followed by several washes with deionized (DI) water. After filtration and vacuum drying, graphite oxide was obtained.
2.3. Preparation of Reduced Graphene Oxide
GO (100 mg) dissolved in EGM was mixed with different volume ratios of H2O, EGM or NMP, polymer, and surfactant as the initial solvent system, which was sonicated at 50 W for 2 h. The reducing agent, Vc, was added, while the third solvent was added dropwise to the above mixture. After a 4 h reaction at 80 °C, the obtained dispersion was filtered in vacuum and then rinsed with NMP 2 or 3 times. The final RGO was dried in a vacuum drying oven at 150 °C for 48 h to obtain the RGO film. RGOx changes for different conditions are indicated by the subscript x.
2.4. Thermal Annealing
Heat-treatment experiments were carried out in a graphite resistance furnace under a nitrogen (N2) atmosphere at a 2 L/min flow and heating rate of 5 °C/min. The annealing temperatures were maintained at 300 and 500 °C, respectively, for 6 h.
2.5. Characterization
Fourier-transform infrared (FTIR) spectra were obtained with an FTIR spectrometer (Thermo Scientific Nicolet 6700, Waltham, MA, USA). Raman spectra of graphene were obtained with Raman spectroscopy (Renishaw InVia Raman microscope, Great Britain), with a laser excitation energy of 530 nm. Particle size was obtained by laser particle sizing (Mastersizer 2000, Great Britain, UK). The morphologies and structures of GO and RGO were investigated by SEM (FESEM, Zeiss Ultra Plus, Jena, Germany), XRD (D8 Focus 3 KW, Bruker ATX, Ettlingen, Germany), and XPS (Thermo Scientific ESCALAB 250 Xi, Waltham, MA, USA). Moreover, four-point probe measurements of resistivity of RGO film were conducted by multifunction digital four-probe tester (ROOKO FT-341, Zhejiang, China).
4. Conclusions
The chemical reduction of graphene oxide is a hot topic of current research in obtaining high-quality monolayers or few layers and high dispersion. Based on the analysis of a large number of reported graphene preparations, we systematically investigated the reduction-related aggregation, incomplete reduction, and surface adsorption problems to design a gradient solvent strategy. In this gradient solvent system, the incremental dropwise addition of xylene ensures the continuous dispersion of the reduction process. The prepared RGO dispersions are suitable for long-term storage and can still be redispersed to form dispersions with an average particle size as low as 200 nm, even in the dry powder state; such an excellent redispersal ability is of great advantage in the functionalization and miniaturization of graphene devices. Conventional multiple solvent washing methods, supplemented by appropriate low-temperature heat treatment, can effectively remove some of the adsorbates on the surface of RGO, resulting in a marked increase in conductivity to 18,000 S m−1. This has provided a feasible direction for the preparation of highly conductive chemically reduced graphene, with the choice of lower-boiling-point solvents and dispersants of less conjugated structure contributing more readily to the high conductivity properties during reduction.