Paul S. Weiser

Growth-sector dependence of fine structure in the first-order Raman diamond line from large isolated chemical-vapor-deposited diamond crystals

The growth‐sector dependence of fine structure in the first‐order Raman diamond line is investigated for the first time in high resolution spectra taken from large isolated diamond single crystals grown using microwave plasma chemical vapor deposition on tungsten wire tips. The volumes of crystal beneath the (100) and (111) surfaces of these crystals were sampled using a high resolution Raman microprobe from which the line shape of the 1332 cm−1 diamond line was found to be distinctly different. A splitting of the diamond line into two components of up to 7 cm−1 was observed for (100) growth sectors. This splitting may be caused by the buildup of directional strain fields caused by the different fundamental growth processes occurring on the (100) and (111) growth surfaces. An additional third peak near 1326 cm−1 was only observed in (111) growth sectors and may be attributable to the presence of stacking faults.

Carbon diffusion in uncoated and titanium nitride coated iron substrates during microwave plasma assisted chemical vapor deposition of diamond

Auger electron spectroscopy has been employed to investigate the effectiveness of thin films of TiN as barriers to carbon diffusion during chemical vapor deposition (CVD) of diamond onto Fe substrates. Auger depth profiling was used to monitor the C concentration in the TiN layer, through the interface and into the substrate both before and after CVD diamond deposition. The results show that a layer of TiN only 250 A thick is sufficient to inhibit soot formation on the Fe surface and C diffusion into the Fe bulk.

Variation of the raman diamond line shape with crystallographic orientation of isolated chemical-vapour-deposited diamond crystals

Abstract The growth faces of large isolated chemical-vapour-deposited (CVD) diamond crystals have been studied using Raman microprobe spectroscopy. The crystals were grown on wire tips using microwave plasma chemical vapour deposition. This enables an effective study to be made of fundamental growth mechanisms by eliminating those defects and stresses associated with particle intergrowth in polycrystalline diamond thin films. A variation of the diamond line shape was observed from different growth directions of the same crystal. The [100] and [111] growth directions were sampled by focusing the laser spot onto the adjacent (100) and (111) growth faces of the same crystal. The high-resolution Raman spectra from these two growth directions exhibit different, non-symmetric 1332 cm−1 diamond line shapes. Not only is there a shift in the diamond line, but a splitting is also observed indicative of the presence of large anisotropic stresses in these crystals. Furthermore, an additional peak at about 1326 cm−1 is observed, but only when sampling the volume of the crystal underneath a (111) growth face. This 1326 cm−1 peak displays an enhanced depolarization ratio for the incident electric field vector parallel to the [112] direction. The results suggest that the origin of this peak probably stems from the presence of stacking faults on 111 planes.

CHEMICALLY VAPOR DEPOSITED DIAMOND FILMS GROWN ON TITANIUM NITRIDE COATED AND UNCOATED IRON SUBSTRATES

The nature of the interfaces of chemically vapor deposited diamond films on Fe substrates with and without a protective TiN coating is investigated. For unprotected Fe substrates a thick graphitic soot containing 6.5% Fe grows upon the Fe in the first few minutes of exposure to the plasma and, once this soot completely covers the substrate, diamond can nucleate and grow upon it into an average quality unfaceted continuous diamond film. However, adhesion is poor, the weak link being the lack of structural integrity of the soot layer itself. A TiN coating is found to prevent soot formation, C diffusion into the Fe bulk, and Fe diffusion into the diamond films. In the initial stages of growth the TiN is covered with a thin layer of amorphous carbon (a‐C), and it is on this layer that diamond nucleates and grows. Here, again, adhesion is not strong, with delamination occurring at the TiN/a‐C interface.

Chemical vapour deposition of diamond onto steel: the effect of a Ti implant layer

Abstract The efficacy of high dose Ti implantation into steel to aid in the deposition diamond onto Fe based substrates is explored. Auger electron spectroscopy (AES) and Rutherford backscattering spectroscopy (RBS) measurements show that a Ti implanted (3 × 1016 ions cm−2) layer prevents the diffusion of C into Fe (6 h at 900 °C). However, no diamond (or other carbons) is deposited under these conditions. When conditions more favourable for the synthesis of diamond are employed (i.e. 1000 °C deposition) the Ti implant is only effective at preventing C diffusion and soot formation for a limited period of exposure to the plasma (5 min). For longer deposition times, soot formation is observed, presumably preceded by diffusion of the Ti layer into the bulk Fe.

Polarized Raman spectroscopy of chemically vapor deposited diamond films

Polarized micro‐Raman spectra of chemically vapor deposited diamond films are presented. It is shown that important parameters often extracted from the Raman spectra such as the ratio of the diamond to nondiamond component of the films and the estimation of the level of residual stress depend on the orientation of the diamond crystallites with respect to the polarization of the incident laser beam. The dependence originates from the fact that the Raman scattering from the nondiamond components in the films is almost completely depolarized while the scattering from the diamond components is strongly polarized. The results demonstrate the importance of taking polarization into account when attempting to use Raman spectroscopy in even a semiquantitative fashion for the assessment of the purity, perfection, and stress in chemical vapor deposition diamond films.

Chemical vapour deposition of diamond onto iron based substrates—The use of barrier layers

Abstract When Fe is exposed to the plasma environment suitable for the chemical vapour deposition of diamond, the surface is rapidly covered with a thick layer of graphitic soot and C swiftly diffuses into the Fe substrate. Once the soot reaches a critical thickness, diamond films nucleate and grow on top of it. However, adhesion of the film to the substrate is poor owing to the lack of structural integrity of the soot layer. A thin coating of TiN on the Fe can act to prevent diffusion and soot formation. Diamond readily grows upon the TiN via an a-C interface layer, but the a-C-TiN interface is weak and delamination occurs at this interface. In order to try and improve the adhesion, the use of a high dose Ti implant was investigated to replace the TiN coating.

Correlation between crystalline perfection and film purity for chemically vapor deposited diamond thin films grown on fused quartz substrates

Chemically vapor deposited (CVD) diamond films have been deposited on quartz substrates using a configuration in which the substrate is placed parallel to the direction of the gas flow in the deposition system. Spatially resolved Raman spectroscopy and optical microscopy of the resultant films revealed that (a) as the diamond component of the films increases, the defect density (as measured by the FWHM of the Raman 1332 cm−1 line) decreases, (b) there is a decrease of the quality and perfection of the CVD diamond particles as they overgrow to form a continuous film, and (c) the best quality diamond particles (FWHM) of the 1332 cm−1 line=2.7 cm−1) are produced downstream at the bottom of the plasma ball. It is suggested that the limitations on the continuous film quality appear to be governed not so much by the details of the growth chemistry, but rather by the effects of particle overgrowth.

Homo-epitaxial diamond film growth on ion implanted diamond substrates

Abstract The effect of strain and defects, within a diamond substrate, on the growth of homo-epitaxial CVD diamond has been investigated. The strategy employed is to create laterally confined regions of strain in the substrates by focused MeV implantation of light ions. Raman microscopy has been employed to obtain spatially resolved maps of the strain in these implanted regions. A homo-epitaxial CVD diamond film was grown on top of both the implanted and unimplanted regions of the substrate. Raman analysis of the film grown on top of the implanted region revealed it to be under slightly tensile strain, compared with that grown on the unimplanted diamond substrate. The film deposited on the implanted portion of the diamond showed a lower fluorescence background. The results indicate that defects and strain in the diamond substrate can effect the nature of homo-epitaxial diamond growth.

Enhanced diffusion of C in Fe under CVD diamond deposition conditions

Abstract It has been reported, by other authors [1–3], that under DC plasma conditions the diffusion of C into Fe is much greater than that obtained under thermal equilibrium conditions. They postulated that the exceptionally high C uptake is due to rapid infusion (mass transfer) of C into the surface from the plasma, and not to an anomalously high diffusion rate in the bulk. The aim of the current investigation was to determine whether the rapid infusion of C and classical C diffusion rates obtained from high CH 4 concentration plasmas are also observed in the lower C content (1% CH 4 ) plasma employed in CVD diamond deposition. A series of experiments were conducted for short deposition times (up to 5 min) under plasma-CVD diamond deposition conditions at a substrate temperature of 950°C. The C profile was analysed using Auger electron spectroscopy (AES) and Rutherford backscattering spectroscopy (RBS). The diffusion coefficient (D o ) calculated from thermodynamic data for 950 ± 20° is (3.0 ± 0.6) × 10 −11 m 2 s −1 , whereas that estimated from the AES depth profile using a constant surface concentration solution to Ficks second law was found to be 4.2 ± 1.3 × 10 −10 m 2 s −1 . Hence, even though the carbon content in the plasma is very low the diffusion occurring under plasma conditions at a sample temperature of 950 ± 20°C is equivalent to that expected from thermal diffusion at 1244 ± 42°C. Possible mechanisms for this enhancement are discussed.