Structural characterization of diamond at 200, 250, and 300W

Further the AC/Si substrates were placed in the MPCVD reactor for diamond growth at various microwave power (200 to 350W at fixed pressure of 20 torr). After uploading the AC/Si substrates into the CVD chamber, then gas was flowed (0.6% CH₄ in H₂ plasma) into the chamber. When the pressure reached 10 torr (became constant) then the microwave power was increased to 380W (in general plasma formed between 370 to 400W). When the plasma formed, the power was reduced to 200-350W) while the pressure was maintained at 10 torr. After reducing the microwave power, the pressure was increased up to 20 torr for diamond growth. After diamond growth, the specimen was cooled down at ambient pressure in the presence of hydrogen gas (10 torr). The structure, surface morphology, and size of the synthesized diamond on the Si substrate were examined in a field-emission scanning electron microscope (SEM, JEOL JSM-6700F).

Figure 4.9 shows the SEM image of diamond synthesized at 200W. From the SEM image, it is clear that the unusual large crystal sizes ~ 4μm along with a few small (~100nm) particles are formed. The density of small particles is < 10² cm^-2. The diamond growth on AC/Si substrate at 250W is shown in Figure 4.10 (a), from which the diamond density is

~10³ cm^-2, relatively higher than that at 200W. Figure 4.10 (b) shows the size of diamond is ~ 650nm. It is clearly indicated that not only the diamond density but also their sizes are increased with microwave power. Though the exact substrate temperatures could not be measured at 200 and 250W, the temperature could be below 450°C because the measured temperature at 350W was~ 530°C (qualitatively).

Figure 4.9: SEM image of diamond on AC/Si at 200W after 270 min growth.

Figure 4.10: (a) low-magnification and (b) high-magnification SEM image of diamond on

Further both Si substrates coated with adamantane and without adamantane were placed side by side in the MPCVD reactor for diamond deposition at 300W for 270 min. Figures 4.11 (a) and (b) show typical plan-view and cross-sectional SEM images of synthesized diamonds on the AC/Si substrate. It is shown that a contnuous diamond film can form on the substrate. The average size and the thickness of synthesized diamond are ~ 2 and ~ 1.8 μm respectively, as shown in Figures 4.11(a) and (b).

Figure 4.11: SEM images of diamond: (a) plan-view of AC, (b) cross-section view of AC, (c) plan-view of WAC and (d) cross-section view of WAC.

To know the effect of adamantane on diamond nucleation, we have grown the diamond on the WAC/Si substrate under the same experimental condition as AC. The

morphology of synthesized diamonds of WAC is shown in Figures 4.11 (c) and (d).

No continous diamond film can be seen on WAC/Si. Figures 4.11 (c) and (d) show a local region where the diamond density is the highest. Most of the areas on WAC are rarely observed diamond particles. The average size of synthesized diamond is ~ 1 μm. In comparison, the density of diamond is ~10⁶ and ~10² cm^-2 for AC and WAC respectively, clearly indicating that the density of AC diamond is higher than that of WAC diamond, by four orders. Therefore, it is suggested that the adamantane strongly assists in the nucleation and growth of diamond at low temperature and pressure.

Furthermore, the synthesized diamond films on the AC/Si substrate at 300W were evaluated by Raman spectroscopy, X-ray diffraction, and X-ray photoelectron spectroscopy (XPS, Thermo VG 350, MgKα X-ray source). After diamond growth on AC/Si, all adamantane peaks disappeared and only an extremely sharp Raman peak appeared at the position of 1332 cm^-1, as shown in Figure 4.12. The sharp peak at 1332 cm^-1 in the Raman spectrum is attributable to the first–order phonon mode for diamond, and the width of the peak is 5.6 cm^-1. The sharp Raman peak at 1332 cm^-1 and the absence of graphitic (D and G bands) peaks suggest that the good-quality diamond is obtained. In addition, the XRD pattern is shown in Figure 4.13. In the XRD pattern, the 2θ peak at 28.3° is the Si{111} peak while the sharp peak at 43.9° is diamond{111}.

Since Si{100} substrate was used for diamond growth, the appearance of the Si{111}

peak may imply that the during the diamond growth the Si might be etched by the hydrogen plasma. The two sharp peaks at 37.8° and 64.38° are identified to be cubic silicon carbide (SiC) {111} and {220}, respectively [109], consistent with the standard JCPDS values (JCPDS file number: 49-1623). The sharp XRD peak of diamond shows that the highly crystalline diamond film has been deposited as supported by our SEM observations in Figure 4.11.

Figure 4.12: Raman spectrum of diamond film on AC/Si at 300W.

Figure 4.13: XRD pattern of diamond (D) film on AC/Si at 300W. The symbol D in this pattern represents diamond.

Moreover, structural characterization of diamond was done by TEM using a cross-sectional TEM specimen prepared by the focused ion beam (FEI Nova 200 Dual beam

energy ion beam (30 keV Ga⁺) in FIB, the specimen was coated with platinum. A cross-sectional bright-field TEM image of diamond/Si is shown in Figure 4.14 (a). The thickness of diamond is ~ 2μm, as shown in Figure 4.14 (a). The interface between Si and diamond is shown in Figure 4.14 (b), and the enlarged-view with the EDX spectrum is shown in Figure 4.14 (c). It is seen that there is an amorphous interlayer between diamond and Si. EDX taken from the interface shows the presence of oxygen, carbon, and Si.

Figure 4.14: (a) Cross-sectional BF-TEM image of diamond/Si, (b) BF-TEM image of interface, (c) enlarged BF image of interface; insert EDX spectrum from interface, and (d) SAED pattern of interfacial region indicating the presence of Si and diamond spots {111} plane along <011> direction; Inset diffraction spots (i) and (ii) showing the of diamond and Si along <011>

zone axis.

The Si signal comes from substrate and C from diamond, while the oxygen signal suggests that the presence of native oxide layer on Si substrate. The selected area electron diffraction pattern (SAED) of the interfacial region is shown in Figure 4.14 (d). From the SAED pattern only noticed diffraction spots of Si and diamond as shown in Figure 4.14 (d). Inset images (i) and (ii) in Figure 4.14 (d) shows the diffraction spots of diamond and Si along <011> zone axis. As no SiC diffraction spots can be observed, it might be that the SiC is not a continuous film.

Figure 4.15: The XPS survey spectrum of diamond on AC/Si, insert: high-resolution spectrum of the C 1s region.

In addition, to get a better insight into the above mechanisms we have used XPS to characterize the surface chemical composition of the AC/Si surface after diamond growth.

The XPS survey spectrum of the synthesized diamond on the Si substrate is shown in Figure 4.15, which allows measure of the carbon 1s core-level signal considered as an

expected peaks from elements on the diamond surface (C), surface contamination with oxygen containing molecules can be detected. An XPS survey shows strong signals of carbon C 1s at 285.2 eV, attributed to the C-C bond in diamond films. In fact, insert XPS high-resolution spectrum of carbon in Figure 4.15 showed that the C (1s) signal consists of three peaks: the first two peaks at 285.2 and 283.7eV is the characteristics of C-C (sp³) bonding and the other peak at 283.2eV is the characteristics of Si-C bonding [110]. Many authors have shown that silicon carbide (SiC) forms readily after diamond growth starts [111, 112]. In our case, most fractions of adamantane particles might have been pumped out. However, some embedded hydrocarbon species could react with Si surface and form SiC. Probably, the etched Si (rough) surface would provide more active sites to react with carbon species to form SiC. In general, it is well known that SiC forms at high temperature. However, according to our XPS and XRD results we can say that SiC may form at low temperature. At high temperature, hydrocarbon species come from the plasma and react with Si to form SiC interlayer between diamond and Si. In our case, we have used adamantane therefore either adamantane or other carbon species assist to the formation of SiC. The distribution of carbon species is not homogeneous. The oxygen signal at 532.7 eV is associated with oxide formation with Si surface during the thermal reaction, which can attribute to the O-Si bond in SiO₂ [94, 95]. It seems that the adamantane compact density is low, allowing for high diffusion of oxygen. Therefore, during adamantane coating by hotplate method the oxygen atoms may have diffused through adamantane to Si surface to form native oxide layer. It is also possible that after diamond growth, when we have taken out specimens from the CVD chamber the oxygen would have reacted with Si surface which was uncovered with diamond.