**4. Defects in graphene and graphite**

Studies have shown that defects can arise during the preparation process of graphene. The defects in graphene exist in the form of sp3 carbons with more or less than six carbon atoms in a ring [27]. Usually, one or two H atoms are attached to the sp3 carbons. The presence of defects in graphene can change its electrical, mechanical and magnetic properties as well as surface chemical reactivities, thus having a significant impact on the performance of graphene-based devices. On the contrary, the surface of HOPG is almost defect-free with only delocalized sp2 carbons and can be used as a standard model for the study of surface defects on graphene. XPS has been the preferred choice for determining the structures of graphene and graphite surfaces [28]. **Figure 4a** shows an XPS C1s spectrum of HOPG, revealing a narrow main carbon peak with the binding energy at 284.5 eV, accompanied by a broad and asymmetric tail toward higher binding energy [28]. The asymmetry is due to the low energy electron-hole pair excitation as the valence electrons respond to the presence of the core hole. Two empirical approaches have been used to curve fit this asymmetric line shape. The first approach uses an asymmetric Doniach-Sunjic function, which was originally developed for analyzing the asymmetric line shapes of XPS spectra of metals [29], while the other considers HOPG as a neutral alternant hydrocarbon and fits its C1s spectrum with five symmetric components [30]. Xie *et al.* obtained clean HOPG and graphene surfaces by annealing the samples at 500°C in an ultra-high vacuum chamber [28]. A combination of Doniach-Sunjic and Gaussian-Lorentzian functions was used to curve fit the asymmetric C1s spectrum of HOPG. An asymmetric parameter of 0.035 was determined after considering the left full-width-at-half maximum (FWHMleft) and the right FWHM (FWHMright) of the C1s peak. The C1s spectrum of clean HOPG was fitted with two components including the sp2 carbon peak and the π-π\* shake-up peak (**Figure 4a**). However, for the curve-fitting of C1s spectrum of graphene, a sp3 carbon peak representing the defects at the binding energy varying between 285.0 and 285.5 eV also appears (**Figure 4b**). To confirm the nature of the sp3 peak, defects were introduced on the surface of HOPG. An effective way of inducing defects on the surface of HOPG is ion bombardment, in which the defect density can be controlled by varying the ion dose density. Defects created on a HOPG surface can broaden the FWHM of its XPS C1s spectrum and make the line shape on the high-binding-energy side of the peak more asymmetric due to the disorder of its delocalized sp2 structure and the development of the sp3 component. **Figure 4c** shows that the FWHM of the C1s peak becomes broader and the sp3 carbon peak intensifies as more defects were created on the surface of HOPG. The C1s curve of the sputtered HOPG at the take-off angle of 20° (sampling depth about 2.6 nm) showed a higher sp3 peak intensity compared with that at 90° (sampling depth about 7.5 nm). This result indicates that sp3 defects which were created by sputtering a defect-free HOPG surface mainly concentrated on the top surface. The atomic ratio of sp3 carbons to sp2 carbons in the sputtered HOPG samples determined by XPS was used to estimate the amounts of sp3 defects. For the HOPG sample, the ratio was close to zero. As the sputtering dose increased, the ratio gradually increased. Note

*Surface Analysis of Graphene and Graphite DOI: http://dx.doi.org/10.5772/intechopen.108203*

#### **Figure 4.**

*XPS C1s normalized spectra of (a) HOPG after 500°C annealing, (b) graphene on a SiO2/Si wafer after washing with acetone and annealing at 500°C, and (c) HOPG after 500°C annealing followed by Ar+ sputtering with a sputtering dose of 5.0 × 1015 ions cm−2. The black ( ) and green ( ) lines represent the experimental and curvefitted spectra, respectively. All spectra were obtained at 500°C at a take-off angle of 20° [28].*

that sputtering creates defects but at the same time also removes the defects from the surface. As a result, the ratio eventually reached a constant value when the rate of defect generation equaled the rate of defect removal.

The defects in graphene can also be characterized using ToF-SIMS [28]. A comparison between the ToF-SIMS spectra of graphene and HOPG shows similar ions, including Cx + , CxH+ , Cx − , and CxH− (**Figure 5**). The typical fragment ions of graphene and HOPG can be categorized into several types. The Cx + and Cx − ion series are the type that mainly comes from the direct breaking up of the sp2 areas of graphene and HOPG surfaces. The H-containing fragment ions (CxH+ and CxH− ) form another type of typical fragment ions of graphene and HOPG. The CxH+ and CxH− ions contain at least one hydrogen which would normally be absent from a defect-free graphene or HOPG surface. Their presence in the spectra suggests that they were created *via* direct ejection of the carbons at or near the defect areas of the graphene and HOPG surfaces. Since graphene has many more defects on its surface, the probability of formation of the CxH+ and CxH− ions in graphene is higher, leading to their higher normalized intensities. Another major difference between the spectra of graphene and HOPG is that CxH2 +· ions are present in the positive ion spectrum of graphene but absent in the spectrum of clean HOPG. Therefore, it was suspected that the CxH2 +· ions might be generated at or near a defect in graphene. To further determine the origin of CxH2 +· ions, ion bombardment was carried out to create defects on a HOPG surface.

**Figure 5.** *ToF-SIMS (a) positive and (b) negative ion spectra of HOPG after annealing at 500°C and graphene on a SiO2/ Si wafer after washing with acetone and annealing at 500°C. All spectra were obtained at 500°C [28].*

A detailed examination of the positive ion spectra of ion-bombarded HOPG surfaces reveals the presence of the CxH2 +· ions, confirming that these ions originated from the defects created on the sputtered HOPG surfaces. Therefore, the CxH2 +· ions can be used as an indicator for the existence of defects in graphene and graphite surfaces.

Studies have shown that defects on graphite and graphene surfaces can be repaired under a carbon atmosphere, a noble gas atmosphere, or in a vacuum [31–33]. The repairs can improve their thermal and electrical conductivities and enhance their mechanical strength. Xie *et al.* revealed that the defects created on the surface of HOPG by ion bombarding can be repaired through high-temperature annealing [34]. Both XPS and ToF-SIMS were applied to monitor the creation and repair of defects in a HOPG surface. The ratio of sp3 to sp2 carbons was calculated from XPS results for the HOPG surface before and after sputtering as well as the sputtered HOPG surface after annealing under Ar at different temperatures. These results are shown in **Figure 6a**. This ratio is close to zero for the HOPG and increases to about 0.09 after a dose of 5.8×1015 ions cm−2 sputtering. After annealing at either 500 or 650°C, the ratio remains the same, but annealing at 800°C reduces the ratio to about zero. Therefore, defects on the surface of

**Figure 6.**

*(a) The ratio of sp3 to sp2 carbons in the fresh HOPG sample, HOPG after Ar<sup>+</sup> sputtering at a dose of 5.8 × 1015 ions cm−2 and after annealing in Ar at different temperatures. (b) C KLL spectra of the fresh, sputtered, and annealed HOPGs and (c) their first derivatives obtained at 500°C [34].*

the sputtered HOPG can be repaired by annealing at 800°C in Ar. The ratio of the sp3 to sp2 carbons can be estimated using the distance between the most positive maximum and the most negative minimum of the first derivative of an XPS C KLL spectrum which is referred to as the D parameter [35, 36]. For HOPG, this value varies from 21.2 to 23.1 eV [35, 36]. As shown in **Figure 6b** and **c**, the calculated D parameters are 21.3 and 20.1 eV for the fresh and sputtered HOPG, respectively. The decrease of the D value implies an increase in the number of sp3 carbons on the surface. After annealing at 800°C in Ar, the D value is 21.6 eV, which is similar to that of the fresh HOPG. The recovery of the D value for sputtered HOPG after annealing again implies the conversion of the sp3 carbons which are present as defects on the sputtered sample to sp2 carbons through annealing.

ToF-SIMS spectra were obtained from the surfaces of HOPG before and after Ar<sup>+</sup> sputtering and the sputtered HOPG sample after annealing under flowing Ar at different temperatures [34]. The normalized intensity of the CxH2 +· ions was used as an indicator to reveal the concentration of the defects on the HOPG surfaces. It is obvious that the intensity of these ions was close to zero for the fresh HOPG because it is a defect-free surface. Interestingly, the intensity of these ions for the sputtered HOPG increased, indicating the creation of defects; while the intensity of these ions for sputtered HOPG after annealing at 800°C was close to zero except for the CH2 +· and C2H2 +· ions, suggesting the healing of these defects created by sputtering.
