Study of ultraviolet-visible light absorbance of exfoliated graphite forms

As our technology evolves, the need for new materials that possess superior qualities is very clear. One of these materials is graphene. Graphene and its oxides are regarded as next generation materials that transcend any existing ones due to their excellent electrical, chemical and biological properties.^1,2 The reason for that is its potential in many different areas such as battery/supercapacitor technology,³ transistor development,⁴ transparent display technology and replacing rare earth metals. Graphene is the strongest material with very high Young’s modulus (∼1.0 TPa).⁵ It also possesses very high thermal conductivity (∼ 4.84-5.30 kW m^-1K^-1),⁶ high carrier mobility (200,000 cm² V^-1 s^-1) and electron density (2x10¹¹ cm²).⁷ There have been many techniques, i.e., Raman Spectroscopy, Atomic Force Microscopy, and Transmission Electron Microscopy in determining layer thickness.^8–10 However, these techniques are not cheap and may not reveal the same information as the UV-Vis spectroscopy, where a recent study shows the percentage of few-layer graphene oxide (GO) versus those of multi-layer GO or thick-layer GO.¹¹ Our study is aimed at determining the rate of exfoliation of various forms of graphite and its oxide using Top-down method, UV-Vis spectroscopy and gaining deeper understanding about the electronic transitions in these forms. Graphite contains Carbon 12 (C¹²) atom with six protons, six electrons and six neutrons. Carbon atoms form covalent bonds with the other carbon, hydrogen or oxygen atoms depending upon the chemical composition of the exfoliating agent. The carbon atom ground state electron configuration is given by 1s²2s²2p². The 1s² orbital contains two electrons tightly bound to the nucleus. The 2s orbital contains two electrons (one is spin up and the other is spin down) in the ground state. The final two electrons are located in the 2p orbital. In the excited state, one electron from 2s state moves up to 2p level by absorbing a photon with energy about 4 eV. With an additional electron in the 2p orbital, carbon atom can form covalent bond with other carbon, hydrogen or oxygen atoms. The actual energy gain from the covalent bond is larger than 4 eV spent on the electronic excitation.

Graphene structure contains six carbon atoms attached to each other in a hexagonal structure as shown in Figure 1. Each carbon atom establishes three covalent σ bonds and one π bond with immediate neighbors. Within the hexagonal structure there are three π bonds between the C atoms and the π electrons are delocalized over the whole structure, resulting in excellent electronic conduction properties. Each carbon atom is connected to three other C atoms on the two dimensional plane via covalent σ bonds, the remaining single electron (π electron) is very mobile and located above or below the two dimensional plane. These π orbitals overlap with σ bonds, thereby enhancing the bonds (Figure 2). The σ bonds are strong bonds that require larger energy to break. The π bonds, on the other hand, are weaker, and form the Van der Waals bonds between the consecutive graphite layers.

There are two known methods to produce graphene. First method is called the top-down method, where mainly graphite is exfoliated, i.e., mechanical or liquid exfoliation, to form graphene. Second method is called the bottom-up method where graphene is formed by chemical deposition. The top-down method produces much better quality graphene flakes, but with small size. The bottom-up technique produces large graphene sheets, but with lower quality.

Graphite has a very low solubility in water or other solutions. Graphite oxide, however, is more soluble in water or other solutions. The reason for that is because, having oxygen atoms inserted between graphene layers somewhat reduces Van der Waals forces by increasing the distance between the layers. Essentially, Van der Waals forces are caused by the charge density fluctuations resulting in dipole-dipole interactions between layers. The top-down technique tries to break down the Van der Walls bonds in graphite. Traditionally, graphite is oxidized via strong acids to form graphite oxide.¹² However, this process, can introduce atomic defects into the graphene sheets, thereby greatly reducing the superb properties of graphene. Graphite oxide can then be dissolved in solutions like water or other chemical solutions to result in graphene oxide.¹³ Finally, graphene oxide can be reduced to graphene by removing oxygen atoms. Since the strong acids introduce defects into the final product of graphene, in our study we tried to exfoliate different form of graphite directly with acetonitrile. By changing the form of graphite, i.e., powder, solid or sooth, we would like to test if there are any differences in exfoliation characterized by UV-Vis spectrometer. We also tested graphite oxide exfoliation in acetonitrile to compare it with different exfoliated graphite forms. Lastly, we tested if microwave treatment of sonicated graphite powder in acetonitrile affect the absorption spectrum. Microwaves deliver energy to the bonds between graphene layers, thereby reducing the energy needed to break the bonds by the exfoliating agent. Scanning Electron Microscopy (SEM) and optical microscopy are used to investigate the morphology. These techniques helped us to visualize the extent of exfoliations for each sample. The samples had to be filtered via filter papers and dried before SEM and optical microscopy used.

Materials used are listed below:

• ➢

  Graphite rod (0.2 g, 99.99% purity)

• ➢

  Graphite powder (0.2 g, 99.99% purity)

• ➢

  Graphite sooth (0.2 g, O₂/C₂H₂)

• ➢

  Graphite oxide (0.2g, GO)

• ➢

  Acetonitrile C₂H₃N, (20 ml, anhydrous 99.8%).

Tip sonicator, which delivers more power to the mixtures in shorter times is used instead of bath sonicator. Sonication is done for 3, 5, 8, 10, 20, 30 minutes using Ultrasonic Processor, 300W - 20KHz, with 100% amplitude modulation.

Experimental details are explained in Table I and Figures 3–8 above exhibit the setup and exfoliation products.

Data is taken by using a UV-Vis-NIR spectrometer that covers wavelengths 200 – 1100 nm. The exfoliating liquid is chosen to be acetonitrile due to the lowest wavelengths (about 190 nm) of measurement afforded by this solvent. Figure 9 shows graphite powder in acetonitrile without sonication. The absorption increases from near infrared to ultraviolet. There is no distinguishable peak observed, even though there is a shoulder around 270 nm. The absorbance spectrum of graphite rod sonicated in acetonitrile for 5, 8, and 10 min is presented in Figure 10. There is one absorption peak observed at 273 nm and it grew in intensity with increasing sonication time. The 5 min sonicated sample exhibits a shoulder at around 223 nm, which disappears with the 8 and 10 min sonicated samples due to increased absorption. Sonication times are adjusted as much as possible to avoid overexposure of the UV detector. The absorbance spectrum of graphite powder sonicated in acetonitrile 3, 5 and 10 min is plotted in Figure 11. The 3 and 5 min sonicated samples show two distinct peaks at 223 nm and 273 nm respectively, whereas the 10 min sonicated sample only shows the 273 nm peak due to increased absorption. The absorbance spectrum of graphite sooth sonicated in acetonitrile 3, 5 and 10 min is plotted in Figure 12. The graphite sooth was obtained (from Chris Sorensen at Kansas State University) by detonation of hydrocarbon (C₂H₂) in the presence of O₂.¹⁴ The graph shows that the 3 min sonicated sample shows two absorbance peaks at 223 and 273 nm, the 5 min sonicated sample only shows the 273 nm peak, and the 10 min sonicated sample does not show any peak due to large absorbance. The absorbance spectrum of graphite oxide sonicated in acetonitrile 3, 5 and 10 min is plotted in Figure 13. There are four peaks observed for the 3 and 5 min sonicated samples at 237 nm, 273 nm, 312 nm and 361 nm. The 10 min sonicated sample has too much absorbance and saturates the detector. Figure 14 shows the absorbance spectrum of graphite oxide sonicated in acetonitrile for 3 minutes more clearly. There may be an additional peak at about 200 nm, shown with a star in the spectrum. The absorbance spectra of graphite rod, graphite powder, graphite sooth and graphite oxide sonicated in acetonitrile for 5 minutes is plotted in Figure 15. Variations exist not only in peak positions but also in relative intensities. The absorbance spectrum of microwaved and un-microwaved, 5-min sonicated graphite powder in acetonitrile is plotted in Figure 16. The spectrum shows increased absorbance below ∼270 nm, but decreased absorbance above that wavelength. Figure 17 shows Scanning Electron Microscope (SEM) images of flakes obtained from graphite powder sonicated in acetonitrile. The bright images indicates that flakes are conductive and vary in size and shape. The optical image of graphite oxide sonicated in acetonitrile is shown in Figure 18. The transparent graphene oxide/graphene dispersions are clearly visible with varying sizes.

The observation of two peaks at 223 nm and 273 nm indicates, our samples contain a mixture of graphene oxide (223 nm peak) and graphene (273 nm peak). Graphene oxide absorbance peak is shifted to 237 nm with graphite oxide sample (Figure 13, 14). We think this is because additional oxygen between layers of graphite oxide making it easier to exfoliate layers, thereby resulting in more graphene-like flakes with reduced oxygen. This is justified by the absorbance peak to shift from 223 nm to 237 nm towards graphene peak at 273 nm. Longer sonication times increase both graphene oxide and graphene concentration in the mixture. Figure 11 indicates that relative intensities of the peaks change with the increased sonication time from 3 min to 5 min. The 5-min-sonicated sample exhibit higher amount of graphene (judging by the peak intensity) compared to the 3 min one. This suggests that increasing sonication time converts more graphene oxide to graphene by breaking the Van der Waals bond between graphene oxide layers. In reference 11, few layer (1-3 layers), multi-layer (4-10 layers) and thick layer (>10 layers) graphene oxide were studied using UV-Vis spectrometer. Their data show that few layer graphene oxide dispersions can be distinguished from multi- and thick layer dispersions by a more intense peak at 230 nm. In our case, this peak location is 223 nm. When 5-min sonicated samples analyzed in Figure 10–14 with respect to their intensity for 223 nm peak, it is seen that graphite sooth and graphite oxide samples have absorbance intensities higher than 4.0, whereas graphite rod has about ∼3.3 and graphite powder has about ∼2.0. Thus, it appears that few layer (1-3 layers) graphene oxide dispersions exist more in graphite sooth and graphite oxide samples. Multi or thick layer dispersions exist in graphite powder sample. It is also interesting to note that graphite rod sample has thinner graphene oxide dispersions than graphite powder sample. With the same analogy, we can analyze the 273 nm peak for the same samples. From Figure 10–14, we see that the highest peak intensity for 273 nm corresponds to graphite oxide sample with peak intensity about ∼3.0. Graphite sooth sample has the second highest peak intensity with about ∼2.5. The third highest peak intensity with about ∼0.8 belongs to graphite powder sample, and the fourth highest peak intensity with about ∼0.6 belongs to graphite rod sample. Thus, we can conclude that graphite oxide sample has the thinnest graphene dispersions. The next thinnest graphene dispersions exist in graphite sooth sample. The thickest graphene dispersions belongs to graphite rod sample.

The absorbance spectra of graphite oxide sonicated in acetonitrile showed four distinct peaks at about 237, 273, 312 and 361 nm (See Figure 14). The 237 nm peak is associated with graphene oxide dispersions. The 273 nm peak is associated with graphene dispersions. The remaining 312 nm and 361 nm peaks, to our knowledge, has not been observed before. We think that these new peaks are formed due to conjugated polyenes CH₃ – (CH – CH)_n – CH₃(n = 4 for 312 nm peak and n ≥ 6 for 361 nm peak respectively). We also think that the absorbance peak at 223 nm is related to the π → π^* electron transition of C-C ring (see Figure 2) and the absorbance peak at 273 nm is related to $n→$ π^* transition of C-O bonds. Exact location of the graphene oxide and graphene peaks appears to be different from one study to another. Reference 15 shows the two peak positions to be 230 nm and 270 nm, respectively.¹⁵ We believe this shift may be due to the amount of oxide in the graphene oxide.

Our study resulted in many interesting and useful conclusions about graphene and its oxides. The UV-Vis technique is proven to be very effective in studying the number of layers exfoliated and detecting the difference in conjugated polyenes that affect π → π^* plasmon peak. Detecting the number of layers cheap and non-invasive way is very crucial in producing next generation materials. Studies of graphene indicate that the thinner the exfoliated graphite (ideally one monolayer) the better the material quality. In our study, we observed that increasing sonication time converts more graphene oxide to graphene by breaking the Van der Waals bonds between graphene oxide layers. The function of sonication is to provide the additional energy to break these bonds. Our study can also reveal information about the relative amount of graphene dispersions versus graphene oxide dispersion in our samples, which is a very important information to have in producing mass amount of graphene. Graphene oxide does not have the superior qualities that graphene has. Graphite oxide sample has the thinnest graphene dispersions among the compared samples. The thinnest to thickest graphene dispersions exist in this order: graphite oxide, graphite sooth, graphite powder and graphite rod. We believe the reason for this order is because graphite oxide is processed with strong acids to increase the separation of graphene layers thereby making it easy for exfoliating agent to break the bonds. Graphite sooth exhibited remarkable results considering there was no acid involved in its production, thereby better quality layers. Although graphite powder and graphite rod samples did not result in the thinnest graphene flakes, with longer sonication times, they can still produce thinner and defect free graphene layers. It appears that few-layer graphene oxide dispersions exist more in graphite sooth and graphite oxide samples. This makes sense since graphene oxide eventually results in graphene layers after the oxide removal. Multi or thick layer dispersions of graphene oxide exist more in graphite powder sample. It is also interesting to note that graphite rod sample has thinner graphene oxide dispersions than graphite powder sample. This may be because the process of powdering graphite may be causing oxidation to the sample.

The absorbance spectra of graphite oxide sonicated in acetonitrile showed four distinct peaks at 237 nm, 273 nm, 312 nm and 361 nm. The first two peaks (237 nm and 273 nm) are originating from graphene oxide and graphene dispersions, respectively. To our knowledge the last two peaks (312 nm and 361 nm) have not been observed previously. We think that these new peaks are formed due to conjugated polyenes CH₃ – (CH – CH)_n – CH₃ (n = 4 for 312 nm peak and n ≥ 6 for 361 nm peak respectively). We believe these peaks may be coming from the chemicals added to graphite powder to oxidize. These chemicals interact with carbon bonds attaching hydrogens and forming conjugated polyenes.

The absorbance spectrum of microwaved and un-microwaved, 5-min sonicated graphite powder in acetonitrile indicates increased absorbance below ∼270 nm, but decreased absorbance above that wavelength. We think that this increased absorbance could mean that thicker graphene oxide dispersions are created from the microwave treatment.

We would like to thank Chris Sorensen for providing the Graphite sooth sample.