Role of substrate temperature during diamond film growth using the newly developed time-modulated chemical vapor deposition process

Abstract

Diamond has many excellent properties that make it useful for many industrial and consumer applications [1]. Researchers have successfully deposited diamond films onto a variety of substrate materials using a range of techniques [2–4]. The most successful and the most widely used method for depositing diamond coatings is chemical vapor deposition (CVD). It is generally accepted that the morphology, crystal size/orientation, surface roughness and coating adhesion, are critical in determining the suitability of the coating being used for a particular application. Films grown using conventional CVD processes display rough surfaces, which become pronounced with increasing film thickness. Therefore, this limits their potential for use in microelectronic, optical and biomedical applications. Researchers have employed numerous techniques to produce smoother looking diamond films [5–12]. In addition, smooth nano-crystalline diamond films have also been deposited using a range of methods [13–20]. Recently, we proposed a new time-modulated CVD (TMCVD) for depositing smooth diamond films [21]. The key feature of this process is that it involves CH4 pulse cycles during film growth. The process of dynamic film growth using CH4 modulations constitutes an interesting concept in CVD diamond technology. However, the TMCVD process is not yet fully understood, for example, the mechanisms involved during film growth remain un-investigated. In this communication, results relating to the effects of substrate temperature, during CH4 pulses, on the properties of diamond films grown using TMCVD have been presented. Diamond films were deposited onto silicon (100) substrates (5 mm × 5 mm × 0.5 mm). The substrates were abraded with diamond powder prior to film deposition in order to enhance the nucleation density. A conventional hot-filament CVD system [22] was used to deposit the diamond films. Prior to the depositions the filament was pre-carburized in order to prevent filament poisoning. In this investigation, two types of samples were prepared, namely, sample A and sample B. The conditions employed during film growth were: hydrogen flow, 150 sccm; pressure, 30 Torr; growth time, 126 mins; filament-substrate distance, 4 mm. The TMCVD process pulsed CH4 throughout the growth process. The substrate temperature was measured using a K-type thermocouple. A Hitachi 4100 scanning electron microscope (SEM) was used to characterize the films for morphology. A surface profiler (Hommelwerke, T1000) was used to measure the surface roughness of the films. In addition, a Renishaw 2000 micro-Raman system with a 514 nm He-Ne laser was used to characterize the diamond films for diamond-carbon phase purity. Fig. 1 shows the variations in substrate temperature and CH4 flow during sample A (a) and sample B (b) preparation using the TMCVD process. It is important to note that during sample B preparation, although the CH4 flow was modulated, the substrate temperature was maintained constant throughout the growth process. It may be of interest to note that the substrate temperature was kept constant during the high/low CH4 pulse cycles merely by adjusting the filament power accordingly. The temperature of the silicon substrates at 3 and 4.5 sccm CH4 flow was noted to be 802 ◦C and 776 ◦C, respectively. The fluctuations in substrate temperature, resulting from the CH4 pulses during TMCVD, can significantly alter the diamond film growth process. Fig. 2a displays the graph relating substrate temperature to CH4 concentration under typical diamond CVD growth conditions. Generally, the substrate temperature decreased with increasing CH4 concentration. It should be noted that the observed temperature trend was obtained in the presence of two gases, namely, CH4 and hydrogen, present in the reactor chamber during film growth. Since the TMCVD process pulses CH4 at 3% and 2% concentrations, the fluctuation in the substrate temperature, from 776 to 802 ◦C, respectively, needs to be considered, as it is a critical factor influencing film growth during diamond CVD. In order to investigate the independent affect of CH4 on substrate temperature, only CH4, without any hydrogen, was introduced into the vacuum chamber. It is clear from Fig. 2b that the substrate temperature remains almost constant with varying CH4 flow. Therefore, it can be stated that CH4 flow has no significant affect on the substrate temperature. However, once excess of hydrogen ions are present in the CVD reactor then the resulting hydrogen/CH4 plasma plays a critical role in governing the substrate temperature. There are three main factors that can contribute to the substrate temperature changes observed during CH4 pulses, namely, (i) gas species thermal motion, (ii) chemical reactions between carbon-containing species and hydrogen, and (iii) the carbon-phase transitions. It is important to realize that generally during diamond CVD, these three factors come into play