Green modification of graphene dispersion with high nanosheet content, good dispersibility, and long storage stability

In this work, an easy, green, noncovalent surface modification of pristine graphene (GR) was carried out using a single-step method between sodium carboxymethyl cellulose (CMC) and pristine GR, resulting in the formation of a modified CMC–GR (CGR) dispersion with 15% nanosheet content, the first reported in water. Results obtained from thermogravimetry analysis (TGA), Raman spectroscopy, and atomic force microscopy (AFM) confirm that the CMC modifier is successfully decorated on the pristine GR surface. Analyses of transmittance spectrum, zeta potential and transmittance electron microscopy (TEM) images reveal that the modified CGR has a high degree of dispersion. More importantly, the pristine GR is insoluble, while the modified CGR-3, mixed with 1.1 wt% CMC modifier, is easily well dispersed in water and has good flowability, and no sedimentation is observed after more than 3 months.


Introduction
Waterborne coatings have gained much attention for metal protection on account of their reduced emission of volatile organic compounds (VOCs), avoiding serious workplace accidents. 1 However, waterborne coatings easily form polar channels that accelerate water permeation due to their hydrophilic groups. Thus, waterborne coatings with good corrosion resistance are urgently needed to be developed.
Polymer nanocomposites, including nanollers such as silica, 2,3 montmorillonite, 4,5 boron nitride, [6][7][8] and graphene, 9,10 are considered more efficient materials that enhance the corrosion resistance of coatings by improving their barrier properties. Graphene, a promising two-dimensional (2D) nanosheet with only one-atom thickness, has a series of preeminent characteristics, such as high barrier performance, good mechanical properties, outstanding thermal stability and high chemical stability. [11][12][13][14] At present, it has received increasing attention in the eld of anticorrosion coatings for ultralight metals, and it has been used in waterborne anticorrosion coatings to enhance anticorrosion performance. 15,16 However, graphene is not easy to disperse in water due to its high surface area, strong van der Waals forces, and intrinsic hydrophobicity. One effective method is to gain hydrophilic graphene oxide (GO) using chemical oxidation modication. Unfortunately, the defect sites introduced during the process can affect the long-term anticorrosion performance. Furthermore, many studies have been done to improve the wettability of graphene in matrix through facile noncovalent modication. Chang et al. 17 constructed polyaniline/graphene waterborne epoxy coatings to enhance the corrosion resistance of steel against O 2 and H 2 O. He et al. 18 used tannic acid (TA) as intercalator to disperse graphene in water via p-p noncovalent bonds, forming graphene-TA hybrids and obtaining highly efficient anticorrosive epoxy coatings. Yu et al. 19 prepared waterborne epoxy coatings combined with dispersed sandwiched polyvinyl butyral@graphene@polyvinyl butyral composites to improve the corrosion resistance of commercial aluminum alloys, and the effect can last 120 days even when exposed to simulated seawater.
In recent years, most works have only focused on graphene dispersion through covalent and noncovalent modication using a low graphene content, between 0.1% and 5%. However, it is difficult to use this kind of graphene dispersion as a corrosion resistance additive for industrial waterborne coatings. Based on the demand for corrosion resistance and mass production requirements, graphene dispersions with high solid content and good dispersibility in water need to be designed and developed.
In this paper, we report an easy, green modication of graphene by using a noncovalent strategy between CMC and graphene, as shown in Scheme 1, to prepare a modied graphene CGR dispersion with 15% nanosheet content, good dispersibility, and long storage stability, the rst reported in water.
This method for green surface modication of pristine GR can be used to prepare GR/waterborne polymer nanocomposites. The entanglements between the long chains of CMC and the polymer matrix substantially enhance the interactions between them and accordingly improve their compatibility. Importantly, this strategy is available and economical due to its easy operation in industry. We also believe that this work will accelerate the development of industrial anticorrosion coatings for ultralight metals.

Modication of GR with CMC modier
First, deionized water ((85-n) g) and CMC modier (n g) water were added to a three-necked ask and stirred with a high-speed stirrer at 900 rpm for 15 minutes. Then, crude GR (15 g) was added to the preprepared mixture and sonicated for 1 hour to obtain a uniform CGR dispersion. For the sake of comparing different CGR dispersions, four different contents of CMC modier (n ¼ 0.7, 0.9, 1.1 and 1.3 wt%) were designed to prepared CGR-x (x ¼ 1, 2, 3 and 4), correspondingly. Then, the products were puried aer each modication by ultraltration. The resulting CGR solid powders were dried overnight at 80 C in vacuum.

Characterization
Thermogravimetry analysis (TGA) of pristine GR and CGR was performed under nitrogen on a STA409PC thermogravimetric analyzer (Netzsch instruments) with a temperature range of 25-1000 C at a heating rate of 10 C min À1 . Raman spectroscopy of pristine GR and CGR was carried out using a Kaiser Holo-Lab 5000 series spectrometer furnished with a 514 nm excitation laser. The particle size distribution and zeta potential of the graphene dispersion were characterized by a 90Plus PALS Zeta Potential (Brookhaven Instruments, USA). The transmittance of the graphene dispersion was investigated with an ultraviolet spectrophotometer (UV2600). Morphology of the graphene dispersion was investigated by eld emission scanning electron microscopy (FE-SEM, SU8010) and transmittance electron microscopy (TEM, JEOL JEM2011). Atomic force microscopy (AFM, Park XE7) was used to characterize the thickness of the GR nanosheets.

TGA measurements
Supporting evidence for the noncovalent attachment of CMC modier on the pristine GR surfaces comes from the thermogravimetry analysis (TGA). Fig. 1 shows TGA curves for the analysis of pristine GR and modied CGR with different contents of CMC at a temperature ramp rate of 10 C min À1 , respectively. The rst stage of mass loss terminates at approximately 180 C, which is due to the rejection of the adsorbed water from the interlayers of the materials. The TGA plot of CMC indicates a gradual mass loss of around 58.4% as the temperature reached 600 C. Furthermore, we noted that CGR showed a signicant weight loss in the range of 260-600 C, corresponding to pyrolysis of the CMC modier. The results indicate that the CMC modier can be adsorbed on the surface of graphene.

Raman spectroscopy
Raman spectroscopy is usually employed to distinguish the ordered and disordered carbon structures of graphene. 18 The D band and G band are represented in the in-plane vibration of sp 2 carbon atoms and the vibration of sp 3 carbon atoms from the functional groups, respectively. It can be seen from Fig. 2 that pristine GR exhibited two characteristic peaks at 1342 cm À1 (D band) and 1586 cm À1 (G band), I D /I G ¼ 1.03. The I D /I G ratio presents a slight increase from 1.03 for pristine GR to 1.06 for CGR-4, indicating that no more defects are introduced aer the modication of pristine GR with CMC modier, and the graphene preserves its basic structural properties.

Size and distribution
The diluted dispersions of pristine GR and CGR with obvious Tyndall effect in water were prepared through mechanical dispersion for 15 min. On the basis of the dynamic light scattering (DLS) principle, the average diameter and size distribution of the graphene dispersions were analyzed using a zeta potential analyzer, as shown in Fig. 3.
According to the result, the pristine GR was dispersed poorly in water, and serious agglomeration was formed because of the high surface area, strong van der Waals forces, and intrinsic hydrophobicity of GR. Its average diameter (D 50 ) is about 4560 nm, as shown in Fig. 3B. Compared with pristine GR, the CGR modied with CMC showed good dispersibility in water, especially CGR-2 (920 nm), which is mixed with 0.9 wt% modier. It was suggested that the dispersion effect of graphene using ultrasonic exfoliation in water can be improved aer modication with a macromolecular modier. Because of the new noncovalent modication between CMC and GR nanosheets, the agglomeration was controlled effectively. With the increasing content of CMC modier, the average diameter of graphene decreased gradually, but aer the content of CMC modier exceed 0.9 wt%, the average diameter increased again. This may be because when the content of CMC is insufficient, the uncoated graphene will be likely to agglomerate, while if the CMC content is superabundant, the average diameter of the modied CGR may increase again because of thickening of the CMC coating layer and entanglement among the long and exible chains of the macromolecular modier. Fig. 1 TGA curves of pristine GR, modified CGR, and CMC modifier. Fig. 2 Raman spectroscopy of pristine GR and modified CGR.

Transmittance spectrum
For quickly evaluating the dispersion effect of graphene, it is necessary to compare the optical transmittance. Fig. 4 shows the transmittance spectra of dispersions of pristine GR and modied CGR, with incident light wavelength ranging from 400 nm to 800 nm. The number in the gure indicates the optical transmittance at 550 nm incident light. We extracted the experimental data at 550 nm from pristine and modied CGR in Fig. 4. As seen from the enlarged inset, we can conclude that the four modied CGR dispersions have similar optical transmittance values, at around 39.2%. Compared to 64.4% at 550 nm for pristine GR dispersion, the optical transmittance of the modied CGR is lower, indicating that the dispersibility of GR has been improved by the CMC modier.

Dispersibility and stability
The most important parameter dening surface properties of electrostatically stabilized nanomaterials in aqueous solutions is the zeta potential value. The relation between zeta potential and CMC modier content in the GR nanosheets is shown in Fig. 5. This plot shows that the zeta potential decreases with increasing CMC until the content of CMC modier reached 1.1 wt%. The plot seems to keep balance when the content of CMC modier was increased; then, the zeta potential reaches the maximal absolute value at 29.3 mV. Therefore, it can be presumed that there are more carboxylate groups of the CMC modier on the surface of the GR nanosheets when the absolute value of the zeta potential is high. In other words, the stability of GR dispersions has been improved due to high zeta potential. The inset photo in Fig. 5 shows the dispersibility of pristine GR and modied CGR in water. The dispersibility of the CMC-modied GR is much better than that of their physical mixture.   The sample of physical mixture has obvious black sediments in water. In contrast to CGR modied with different contents of CMC coated on the surface of graphene, the pristine GR is insoluble, while the CGR-3 dispersion mixed with 1.1 wt% CMC modier was easily well dispersed in water, and no sedimentation was observed aer more than 3 months. This suggests that CGR-3 has excellent solubility in water. However, 0.7-0.9 wt% CMC-modied GR dispersions had a little sediment, indicating that the content of CMC modier was not enough to bring mutual exclusion and steric hindrance effects. When the content of CMC modier is 1.3 wt%, the GR dispersion has bad owability due to the higher viscosity. The above results demonstrate that the CGR dispersion is a physically stable system.

Surface morphology
The surface morphologies of pristine GR and modied GR dispersed in water were examined by SEM, as shown in Fig. 6.
The pristine GR nanosheets dispersed in water at the Si substrate show tight agglomeration, 18,20 which is attributed to the poor dispersion of GR nanosheets (Fig. 6A). In comparison, the aggregation of modied CGR is reduced, which is due to the modication of CMC.
Furthermore, TEM results conrmed the dispersibility of pristine GR and modied CGR-3 with the best dispersion stability in Fig. 7. The pristine GR nanosheets dispersed in water at the Cu grid also show tight agglomeration and exist as large sheets, which is due to the poor dispersion (marked as red circles in Fig. 7A) of GR nanosheets. Aer modication with CMC, the CGR-3 nanosheets are well dispersed in water, show good dispersibility, and exist as small nanosheets (Fig. 7B), indicating that CGR dispersion is easy to disperse in waterborne coatings and builds good barrier properties. The thickness of pristine GR and modied CGR-3 was further examined by AFM. As shown in Fig. 7C, the thickness of the GR layers reaches  about 16-20 nm. The thickness of CGR-3 nanosheets with the thick layer is 40-60 nm, indicating that the increase in thickness is due to the presence of CMC modier on the GR nanosheets (Fig. 7D). Thus, it can be inferred from the increased thickness that the GR surface was successfully modied with CMC.

Conclusion
In summary, a CMC-modied CGR dispersion was successfully prepared in aqueous solution by single-step noncovalent functionalization technique. The results show that the optimum content of CMC modier coated on the surface of graphene is 1.1 wt%. The modied CGR possesses good dispersibility and good owability in water. The prepared CGR-3 dispersion, with no sedimentation aer more than 3 months, has potential application as a functional corrosion resistance additive for waterborne coatings, providing a new strategy for the high-value-added use of graphene.

Conflicts of interest
There are no conicts to declare.