Abstract

It is well known that the nucleation density during diamond film growth can be enhanced by simply scratching the substrates. The abrading particles need not be diamond, as high nucleation densities have been reported on substrates treated with particles of SiC [1], BN [2] etc. However, preferential growth of diamond on diamond has also been reported when diamond is present on the substrate surface as a residue from the polishing [3, 4]. Providing diamond or other nucleation sites is essential as the diamond nuclei formed in the gas phase as a result of the homogeneous nucleation remain unstable until they have reached a certain critical size. This is because the addition of carbon atoms to the nucleus decreases the bulk contribution to the free energy but increases the surface contribution. Therefore, the nuclei formed from the gas phase under normal conditions are unlikely to survive long enough to have an adequately low surface to volume ratio [5]. Thus nucleation on untreated surfaces is slow and leads to a very slow growth rate of the film [6]. Different techniques have been utilized to enhance the nucleation and growth rate of diamond thin films during chemical vapour deposition (CVD). Nonabrasive techniques have been sought since the polishing process is generally manual and seldom exactly reproducible. Stoner et al. [7] used a combination of negative bias and high methane concentration at the beginning of the deposition run in order to produce an SiC layer followed by carbon cluster formation which eventually led to diamond nucleation. A nucleation density three to five orders of magnitude higher than with conventional polishing was obtainable. Several investigators have also successfully used bucky balls [8] and bucky tubes [9] for diamond nucleation. Hartnett et al. [10] used laser ablated amorphic carbon layer for diamond nucleation and reported a nucleation density of 106 cmy2. In this letter we report the growth of diamond films on Si substrates using a nanodiamond monolayer as the nucleation enhancement layer. Enhancements of almost an order of magnitude in the nucleation density and more than a factor of three in the deposition rate over conventional polishing techniques have been realized. The substrate treatment involved making a slurry of commercially available du Pont nanodiamond powder and methanol. The nanoparticles in the nanodiamond powder were typically 5–10 nm in size. The slurry was formed by mixing approximately 4 g nanodiamond powder in 1 l methanol. The substrates were then placed in this slurry such that the slurry covered the surface of the substrates, and the methanol was allowed to evaporate. By controlling the amount of nanodiamond powder in methanol it is possible to deposit a monolayer of the nanodiamond particles. The substrate was never mechanically abraded in this experiment. Diamond films were deposited in an ASTEX microwave plasma assisted CVD system, which has been described elsewhere [11]. The plasma contained 0.6% CH4 and H2 excited by a 2.45 GHz source with 1500 W power. Deposition was carried out at 850 8C substrate temperature and 500 sccm flow rate. For comparison, films were also deposited on Si substrates that had been mechanically polished with diamond paste. Films were deposited for 15, 30 and 240 min. Fig. 1b shows a SEM micrograph of a sample grown with the nanodiamond layer for 15 min. The nucleation density calculated from this micrograph was 1:5 3 108 cmy2. Parallel samples grown on the Si substrates that were mechanically polished using diamond paste gave a nucleation density of 1:7 3 107 cmy2 after 15 min growth, as shown in Fig. 1a. After 30 min growth the nucleation densities were 2:7 3 109 and 1:8 3 108 cmy2 on the substrates with a nanodiamond layer and on mechanically polished substrates, respectively. Films grown for longer times were used to measure the deposition rate. The deposition rate for the diamond film with the nanodiamond interlayer was measured to be 1.3 μm hy1 as compared to 0.4 μm hy1 obtained for the deposition on the polished substrate under identical deposition conditions. Fig. 2 shows Raman spectra obtained from diamond films on the substrates with nanodiamond interlayer and with mechanically polished substrates. The spectra are virtually identical, indicating similar quality film growth in both cases, despite a very high deposition rate for the sample with the nanodiamond nucleation layer. In both cases, in addition to the sharp peak at 1332 cmy1 attributed to the crystalline sp3 fraction of the film, there is a broad peak

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call