I. INTRODUCTION
Graphite-based nanoplatelets have garnered considerable attention because of their unique mechanical, thermal and electrical properties (Novoselov et al., Reference Novoselov, Geim, Morozov, Jiang, Zhang, Dubonos, Grigorieva and Firsov2004). Exfoliated graphite-based platelets when reduced to a single layer of graphene have the potential to revolutionize the flexible electronics market (Geim and Novoselov, Reference Geim and Novoselov2007). However, the exfoliation process is often cumbersome and has not been reduced to a robust manufacturable practice. Some of the commercially marketed graphene powders or inks are basically micronized graphite, which often does not have the transparency and conductivity required for display applications. In an interesting approach, several groups have focused on creating graphite oxide (GO), sometimes erroneously referred to as graphene oxide, which is relatively easy to exfoliate in aqueous media and can then be subsequently reduced by chemical means or heat treatment. This approach is attractive, since the process chemistry for creating GO in large quantities is well established and GO is easily dispersed in water (Dreyer et al., Reference Dreyer, Park, Bielawski and Ruoff2010). Various methods of oxidation of graphite can be found in the literature and the chemical structures created by these methods may vary (Dreyer et al., Reference Dreyer, Park, Bielawski and Ruoff2010). A commonly used oxidation method is based on Hummers method (Hummers and Offeman, Reference Hummers and Offeman1958) or modifications thereof. A proposed molecular structure for GO is based on the Lerf–Klinowski model (Lerf et al., Reference Lerf, He, Forster and Klinowski1998). As per this model, the GO structure includes regions with unoxidized benzene rings and regions with aliphatic six-membered rings; the oxygen-containing species include epoxides (1,2-ether) and C–OH groups with a sprinkling of COOH groups at the edges. Reduction of GO creates a partially reduced version of graphite oxide (r-GO) which is significantly more conductive than GO (Gilje et al., Reference Gilje, Han, Wang, Wang and Kaner2007; Yang et al., Reference Yang, Velamakanni, Bozoklu, Park, Stoller, Piner, Stankovich, Jung, Field, Ventrice and Ruoff2009). Nanocomposites of GO–polymer have also been studied (Tung et al., Reference Tung, Kim, Cote and Huang2011; Vuluga et al., Reference Vuluga, Thomassin, Molenberg, Huynen, Gilbert, Jerome, Alexandre and Detrembleur2011). The hydrophilic nature of GO and its layered structure make it a good host for water-soluble or water-dispersible polymers to intercalate inside the GO lattice.
XRD is an important analytical technique for the characterization of intercalation and exfoliation in composites comprising polymer and layered materials (Blanton et al, Reference Blanton, Majumdar and Melpolder2000). XRD can accurately measure the interlayer or basal plane d-spacing, for example, of GO and monitor intercalation of any species in the gallery of the GO lattice. Whereas the interlayer spacing of graphite is 3.35 Å, conversion to GO results in an increase in this basal plane spacing due to functionalization of graphite with oxygen-containing groups (Kou et al., Reference Kou, He and Gao2010). To determine whether intercalation or exfoliation due to polymers in the GO lattice has occurred in GO–polymer composites, a baseline d(001) basal plane spacing for neat GO is required. A literature review of XRD results for the analysis of GO reveals a range of d(001) values. Selected publications indicate that this range is from d(001) = 7.8 to 12 Å (Liu et al., Reference Liu, Wang, Yang and Ooi2002; Bissessur et al., Reference Bissessur, Liu and Scully2006; Stankovich et al., Reference Stankovich, Dikin, Piner, Kohlhass, Kleinhammes, Jia, Wu, Nguyen and Ruoff2007; Cote et al., Reference Cote, Cruz—Silva and Huang2009; Kou et al., Reference Kou, He and Gao2010; Zhao et al., Reference Zhao, Zhang, Hao, Li, Fang and Chen2010). In an extreme drying study, Jeong et al. (Reference Jeong, Lee, Jin, Kim, Bae and Lee2009) annealed GO at 200 °C for 2 h resulting in d(001) = 6.2 Å. However, further annealing resulted in degradation of GO, indicating that thermal processing at this temperature is chemically altering GO. In a modeling study (Boukhvalov and Katsnelson, Reference Boukhvalov and Katsnelson2008) of GO with 25% coverage by hydroxyl groups, the interlayer distance was found to be about 7 Å. Since a study by Hofmann et al. in 1932 (Hofmann et al., Reference Hofmann, Frenzel, Wilm and Csalan1932), it is known that GO can be intercalated by water. The range of d(001) spacings mentioned earlier suggests that XRD data were collected with varying amounts of water intercalated in between GO layers, likely the result of data collection at different relative humidity, RH. In this study, the effect of RH on the GO interlayer spacing is evaluated. Upon establishing baseline parameters, changes in the GO interlayer spacing due to interaction of GO in a polymer matrix can be determined.
II. EXPERIMENTAL
A. Sample preparation
A specimen of 0.5 wt% GO dispersed in water (N002PS) was obtained from Angstron Materials. Neat GO coatings were formed by depositing a few drops of GO solution onto Teflon or glass microscope slides followed by drying in ambient air. Solutions coated on Teflon could be removed after drying, resulting in self-supporting films. Solutions coated on glass were not removed after drying. Aqueous GO–polymer composite solutions were made by adding GO solutions to 0.5 wt% polymer solutions to obtain the desired final weight ratio of GO to polymer. Films were prepared on glass microscope slides using the same method as described for neat GO coatings.
B. X-ray diffraction (XRD)
XRD data were collected using a Rigaku D2000 diffractometer, equipped with a copper rotating anode, a diffracted-beam monochromator tuned to CuKa radiation and a scintillation detector. Coated films were analyzed in ambient air or under controlled humidity conditions using an environmental cell with a nitrogen–water bubbler. The RH inside the environmental cell was monitored using a VWR digital hygrometer.
III. RESULTS AND DISCUSSION
Reflection and transmission mode XRD patterns collected at 30% RH for a self-supporting film of GO are shown in Figure 1. In reflection mode, the XRD pattern shows a strong (001) peak, indicating preferred orientation of graphene oxide basal planes parallel to the sample plane. The (001) d-spacing is 9.1 Å. Based on the (001) peak width, the crystallite size was determined using the Scherrer technique and was found to be 41 Å in the [00l]-direction. All films coated with commercial GO solutions were found to have a small amount of graphite, observed in Figure 1(a). In transmission mode, the XRD pattern shows a weak (100) peak, indicating preferred orientation of (100) planes perpendicular to the sample plane. The (100) d-spacing is 2.12 Å. Based on the (100) peak width, the crystallite size was determined using the Scherrer technique and was found to be 71 Å in the [h00]-direction.
Initial studies of a GO film on glass found that this sample had different (001) d-spacings when analyzed on different days. Humidity was believed to be the likely cause. XRD patterns (Figure 2) collected at variable RH confirmed this hypothesis. The GO d(001) spacing increases (smaller 2θ) with increasing humidity, due to water insertion between (00l) planes. Note that the (001) GO peak intensity is observed to increase as more water intercalates the GO lattice. The (001) graphite diffraction peak, due to residual graphite in commercial GO solutions, serves as an internal standard for measuring GO d(001) spacing. A detailed evaluation of the effect of RH on the GO basal plane spacing was performed using a GO on glass sample in the RH-controlled environmental cell. Diffraction patterns were collected at a specified RH and d (001) spacings determined (Figure 3). From these data, it is apparent that the range of reported d(001) spacings for neat GO in the literature can be attributed to the RH during data collection, demonstrating the importance of documenting (and reporting) RH when presenting XRD results for GO materials.
Having established RH vs. d(001) for GO films, attention is now placed on GO–polymer composites. In Figure 4, XRD patterns for GO–polyethylene oxide (PEO) films on glass collected at 30% RH are shown. Based on the shift in the d(001) diffraction peak to lower 2θ, PEO is observed to intercalate the GO lattice. At 50 wt% PEO d(001) = 15.2 Å compared to d(001) = 9.0 Å for neat GO at 30% RH. An increase in wt% PEO (70 wt%) did not result in further intercalation as there was no observed increase in the GO d(001) spacing. With PEO polymer intercalated in the GO lattice, the effect of humidity on this composite was evaluated. In Figure 5, the d(001) spacing changes from 15.5 Å at 50% RH to 14.7 Å at 0% RH for the GO:PEO 50:50 composite film coated on glass. Although this change is smaller than what was observed in neat GO, it is a reminder that the RH must be noted when collecting XRD data for GO or GO–polymer samples.
In Figure 6, XRD patterns for GO–polyvinyl pyrrolidone (PVP) films on glass collected at 35% RH are shown. Similar to what was observed with GO–PEO, PVP is observed to intercalate the GO lattice at 50 wt% PVP with the d(001) = 20.0 Å, at 70 wt% PVP d(001) = 31.5 Å and at 90 wt% PVP it appears GO is exfoliated based on the absence of any d(001) diffraction peak. Before defining a GO composite sample as exfoliated, further analysis should be considered. The GO–PVP XRD patterns in Figure 6 all start at 2°2θ (44 Å using CuKα). It is possible to have layered material composites with d(001) > 44 Å (Majumdar et al., Reference Majumdar, Blanton, Dontula, Thomas, Zaikov and Valsaraj2009). Therefore diffraction patterns collected at lower 2θ are required for the instrument conditions used in this study to observe larger d(001) spacings. Low-angle XRD patterns for GO–PVP at 90 wt% PVP are shown in Figure 7. The initial XRD pattern from Figure 6 started at 2°2θ with no (001) GO–PVP diffraction peak observed. A scan starting at 1°2θ using the same diffractometer conditions including slits does show evidence of a diffraction peak; however, significant scatter from the direct beam is present. Using smaller divergence, scatter and receiving slits allows for collection of XRD data started at 1°2θ that shows a resolved d(001) GO–PVP 90 wt% diffraction peak with d(001) = 53.2 Å indicating the GO–PVP composite in this case is intercalated, not exfoliated. Another option to improve observation of large (d(001) d-spacing data would be to change the X-ray source to a longer wavelength, though attenuation of intensity by air needs to be considered. In the situation where collecting low-angle diffraction data using optimized diffractometer conditions still results in the absence of a d(001) diffraction peak, conclusive evidence of exfoliation in layered material composites can be achieved using transmission electron microscopy (TEM) to examine cross-sections of the composite film.
IV. SUMMARY
-
• Graphene oxide, GO, has a hexagonal lattice with a = 2.12 Å and c variable depending on ambient humidity.
-
• Since d(001) varies with humidity, the %RH during XRD data collection must be recorded to accurately assess any real changes in the GO basal plane spacing. Collection of data at 0% RH would be the easiest experimental condition to replicate.
-
• GO can be intercalated with water-soluble polymers such as PEO and PVP.
-
• The amount of intercalation, based on the GO–polymer d(001) spacing, is dependent on polymer type and the amount of polymer present, and to a lesser extent the %RH during data collection.
-
• Low angle XRD patterns should be collected when GO–polymer composites appear exfoliated based on typical XRD data collection conditions. Further evidence of exfoliation should be confirmed using TEM.