This chapter is intended to provide an overview of the main results and implications of the work presented within this thesis. The low temperature growth of diamond using 50%CH4/50%CO2 gas mixtures will be dealt with first, followed by a discussion concerning sulfur doping of diamond films using H/C/S and H/C/O/S containing gas mixtures.
The use of 50%CH4/50%CO2 CVD gas mixtures has been shown to allow the growth of diamond films at lower Tsub than is possible using 1%CH4/H2 gas mixtures []. It was argued that the growth surface of the diamond film could be (at least partially) terminated by CO, in the case of growth from CO2/CH4 gas mixtures. This surface termination was predicted to be less stable than the H-termination that dominates in the case of growth using 1%CH4/H2 gas mixtures. Thus, a more dynamic surface was envisaged with CO (or HCO) species frequently attaching and detaching from the diamond surface. This has the effect of increasing the chance of a CH3 radical finding a hanging bond to attack and thus add a carbon to the lattice. In other words the CO (or HCO) diamond surface termination is less stable than H surface termination with the result that growth is possible at lower Tsub for 50%CH4/50%CO2 versus 1%CH4/H2 gas mixtures.
However, as the results in Chapter 5 demonstrate, there is still a fall observed in both the growth rate and quality of the deposited films with reduced Tsub. This can be attributed to two separate effects. Firstly, reducing Tsub has the effect of increasing the gradient in Tgas between the microwave plasma and the substrate, thus leading to increased quenching of CH3 (e.g. to CH4) as the species diffuses towards the substrate. Secondly, surface diffusion of species to a step edge is generally considered important for the regular crystal growth of diamond films []. The lowering of Tsub hampers such surface diffusion and therefore leads to a reduction in crystal and film quality.
In conclusion, although compared to 1%CH4/H2 systems CO2/CH4 gas mixtures provide a better chance for an incoming CH3 radical to find a hanging bond (and therefore incorporate into the lattice) reducing Tsub has the effect of both reducing the flux of CH3 to the surface and also impeding the surface diffusion of CH3 (or more probably CH2, as discussed in References 2 and ). It is therefore reasonable to conclude that the growth of well faceted, high quality, diamond films by CVD at reduced Tsub is probably an impossible goal.
This section will discuss the major differences observed during the investigation of hot filament (HF) and microwave (MW) activated H/C/S containing gas mixtures, as discussed in Chapter 6. A more general discussion of the major conclusions, and wider implications, of investigations concerning the attempted sulfur doping of diamond films using H2S addition to 1%CH4/H2 and 51%CH4/49%CO2 gas mixtures will follow.
The major contrast between the gas phase environment in the vicinity of a HF, as opposed to a MW plasma, is the spatial profile of Tgas. The temperature gradient in the region of the HF is relatively steep as Tgas reduces from ~ 1800 K within 1 mm of the filament, to ~1100 K at a distance of 5 mm from the filament []. On the other hand, the temperature gradient within a MW plasma is much more shallow, falling from ~ 2000 K in the plasma centre to ~ 1600 K at its visible edge [,].
The chemistry of a HF activated gas mixture is initiated by H atoms that are produced (from H2) on the filament surface. These H atoms then diffuse away from the filament and initiate gas phase reactions, thus activating the input gas mixture. However, the steep gradient in Tgas means that the volume of gas that is within the temperature range for which this activation occurs is relatively small. This has the effect (as discussed in Section 6.8) of making CS production unfavourable in the case of HF activated H2S/1%CH4/H2 and CS2/H2 gas mixtures. In contrast, a MW plasma consists of a much larger volume of gas, which is at a temperature high enough to allow activation of the input gas mixture. This has the effect of making CS production favourable for MW activated H2S/1%CH4/H2 gas mixtures.
Chapter 6 highlighted the possible role of CS in the incorporation of sulfur into CVD diamond films. Although sulfur was seen to be incorporated (with film S/C ratios of up to 0.22%) into the films deposited using H2S/1%CH4/H2 gas mixtures, little effect on film resistivities was observed. This suggests that sulfur was incorporated into the grain boundaries in preference to the diamond lattice. This has similarities to the findings of Carlisle et al. in which N2 addition to 1%CH4/Ar gas mixtures was found to yield conductive nanocrystalline diamond films []. They concluded that nitrogen incorporation was via CN, which they suggested moves to the film surface and then into the grain boundaries [].
These findings relating to CN could be analogous to the situation relating to sulfur incorporation into diamond films via CS. This means that the lack of success with the S-doping of homoepitaxial diamond films could be due to such a movement of CS to the surface where the S could be either retained in a thin layer, or etched away, thus resulting in a low incorporation of sulfur (i.e. the film S/C ratio versus the input gas S/C ratio). On the other hand, the situation is different during the growth of polycrystalline films. Now the CS can be trapped within grain boundaries, thus leading to the relatively high degree of sulfur incorporation into films, as shown in Chapter 6.
The use of H2S/51%CH4/49%CO2 gas mixtures was found to increase the degree of S incorporation into the resulting films by about an order of magnitude, when compared to growth using H2S/1%CH4/H2 gas mixtures. However, as with the investigation of H2S/1%CH4/H2 grown samples, it proved impossible to deconvolute the effects of surface termination and grain boundaries from the four point probe resistivity measurements. Thus, the effect of using oxygen-containing gases on the S-doping of diamond films was not clearly demonstrated.
It is reasonable to conclude that H2S addition to either 1%CH4/H2, or 51%CH4/49%CO2 gas mixtures does not provide a satisfactory route to the incorporation of sulfur into the diamond lattice (rather than the grain boundaries). A more profitable direction for future investigation may be that of co-doping using both S and B in the input gas mixture. Such an approach was first suggested by Angus et al. who showed that n‑type conductivity could be obtained for films grown with an input B/S ratio <0.23 []. They concluded that the formation of HBS, BS2 or BS in the gas phase could provide an alternative route to the incorporation of S into diamond films. Therefore, it is clear that there is scope for future investigations into both the solid state and gas phase chemistry of H/C/B/S containing gas mixtures in relation to diamond CVD.
 J. Stiegler, T. Lang, M. Nygard-Ferguson, Y. von Kaenel, E. Blank, Diamond Relat. Mater. 5 (1996) 226.
 M. Frenklach, S. Skokov, J. Phys. Chem. B, 101 (1997) 3025.
 K. Larsson, J.O. Carlsson, Phys. Rev. B, 59 (1999) 8315.
 J.A. Smith, M.A. Cook, S.R. Langford, S.A. Redman, M.N.R. Ashfold, Thin Solid Films, 368 (2000) 169.
 A. Gicquel, K. Hassouni, Y. Breton, M. Chenevier, J.C. Cubertafon, Diamond. Relat. Mater. 5 (1996) 366.
 S.M. Leeds, P.W. May, E. Bartlett, M.N.R. Ashfold, K.N. Rosser, Diamond. Relat. Mater. 8 (1999) 1377.
 S. Bhattacharyya, O. Auciello, J.A. Carlisle, L.A. Curtiss, A.N. Goyette, D.M. Gruen, A.R. Krauss, J. Schlueter, P. Zapol, J. Appl. Phys. 79 (2001) 1441.
 J.A. Carlisle, J. Birrell, J.E. Gerbi, O. Auciello, J.M. Gibson, D.M. Gruen, as presented at the 8th International Conference of New Diamond Science and Technology, Melbourne, 2002.
 S.C. Eaton, A.B. Anderson, J.C. Angus, Y.E. Evstefeeva, Y.V. Pleskov, Electrochem. Solid-state Lett. 5 (2002) G65.