The recombination of O (3 P) atoms on the surface of a Pyrex tube containing a DC glow discharge in pure O 2 was studied over a wide range of pressure (0.2-10 Torr) and discharge current (10-40 mA) for two fixed surface temperatures (+50°C and +5°C). The recombination probability, , was deduced from the observed atom loss rate (dominated by surface recombination) determined by time-resolved optical emission actinometry in partially-modulated (amplitude ~15-17%) discharges. The value of  increased with discharge current at all pressures studied. As a function of pressure it passes through a minimum at ~0.75 Torr. At pressures above this minimum  is well-correlated with the gas temperature, T g , (determined from the rotational structure of the O 2 (b 1  g + ,v=0)  O 2 (X 3  g-,v=0) emission spectrum) which increases with pressure and current. The temperature of the atoms incident at the surface was deduced from a model, calibrated by measurements of the spatially-averaged gas temperature and validated by radial temperature profile measurements. The value of  follows an Arrhenius law depending on the incident atom temperature, with an activation energy in the range 0.13-0.16 eV. At the higher surface temperature the activation energy is the same, but the pre-exponential factor is smaller. Under conditions where the O flux to the surface is low  falls below this Arrhenius law. These results are well explained by an Eley-Rideal (ER) mechanism with incident O atoms recombining with both chemisorbed and more weakly bonded physisorbed atoms on the surface, with the kinetic energy of the incident atoms providing the energy to overcome the activation energy barrier. A phenomenological Eley-Rideal model is proposed that explains both the decrease in recombination probability with surface temperature as well as the deviations from the Arrhenius law when the O flux is low. At pressures below 0.75 Torr  increases significantly, and also increases strongly with the discharge current. We attribute this effect to incident ions and fast neutrals arriving with sufficient energy to clean or chemically modify the surface, generating new adsorption sites. Discharge modeling confirms that at pressures below ~0.3 Torr a noticeable fraction of the ions arriving at the surface have adequate kinetic energy to break surface chemical bonds (> 3-5 eV).

In this paper we review and provide an overview to the understanding of the chemical vapor deposition (CVD) of diamond materials with a particular focus on the commonly used microwave plasma-activated chemical vapor deposition (MPCVD). The major topics covered are experimental measurements in situ to diamond CVD reactors, and MPCVD in particular, coupled with models of the gas phase chemical and plasma kinetics to provide insight into the distribution of critical chemical species throughout the reactor, followed by a discussion of the surface chemical process involved in diamond growth.

CVD Diamond can now be deposited either in the form of single crystal homoepitaxial layers, or as polycrystalline films with crystal sizes ranging from mm, μm or nm, and with a variety of growth rates up to 100s of μm h−1 depending upon deposition conditions. We previously developed a model which provides a coherent and unified picture that accounts for the observed growth rate, morphology, and crystal sizes, of all of these types of diamond. The model is based on competition between H atoms, CH3 radicals and other C1 radical species reacting with dangling bonds on the diamond surface. The approach leads to formulas for the diamond growth rate G via mono and biradical dimer sites and for the average crystallite size , that use as parameters, the substrate temperature and the concentrations of H and CH x (0 ≤ x ≤ 3) near the growing diamond surface. We recently added a correction factor to the equation for and we now test the predictions of this new equation for diamond crystallite sizes ranging from 10 nm (ultrananocrystalline diamond) to several mm (for single crystal diamond). We find that our model predicts the growth rates of all the forms of diamond to within a factor of 2, and predicts crystal sizes for the growth from CH3 that are consistent with those seen experimentally. We deduce that growth of diamond is a sliding scale, with different types of diamond arising from a smoothly changing ratio of atomic H to hydrocarbon radical concentrations [H]:∑CH x ] at the growing surface. The different growth conditions, gas mixtures, temperatures and pressures reported in the literature for diamond growth, simply serve to fix the value of this ratio, and with it, the resulting film morphology and growth rate. In general, for conditions of high [H] at the surface, diamond growth is predominantly from CH3 addition to monoradical sites, leading to large crystals (or even single crystal growth). With decreasing [H]/[CH3], a competing growth channel emerges whereby CH3 adds to biradical sites and the average crystallite size is reduced simultaneously to μm or even nm for very low [H]/[CH3] ratio. In a third growth channel involving atomic C adding to either mono or biradical sites, the spare ‘dangling bond’ can promote renucleation events and increase possibilities for cross-linking, leading to even smaller nm-sized crystallites. This channel can be dominant in high temperature reactors (e.g., MW plasma-enhanced CVD in 1%CH4/(0−2%)H2/Ar mixtures) where high hydrogen dissociation degree shifts the population distribution in CH x group in favor of C atoms.

O2(a 1Δg) production in a non-self-sustained discharge (ND) in pure oxygen and oxygen mixtures with inert gases (Ar and He) has been studied. A self-consistent model of ND in pure oxygen is developed, allowing us to simulate all the obtained experimental data. Agreement between the experimental and simulated results for pure oxygen over a wide range of reduced electric fields was reached only after taking into account the ion component of the discharge current. It is shown that the correct estimation of the energetic efficiency of O2(a 1Δg) excitation by discharge using the EEDF calculation is possible only with the correct description of the energy deposit into the plasma on the basis of an adequate discharge model. The testing of an O2(a 1Δg) excitation cross-section by direct electron impact, as well as a kinetic scheme of processes involving singlet oxygen, has been carried out by the comparison of experimental and simulated data. The tested model was then used for simulating O2(a 1Δg) production in ND in oxygen mixtures with inert gases. The study of O2(a 1Δg) production in Ar : O2 mixtures with small oxygen content has shown that the ND in these mixtures is spatially non-uniform, which essentially decreases the energetic efficiency of singlet oxygen generation. While simulating the singlet oxygen density dynamics, the process of three-body deactivation of O2(a 1Δg) by O(3P) atoms was for the first time taken into account. The maximal achievable concentration of singlet oxygen in ND can be limited by this quenching. On the basis of the results obtained and the model developed, the influence of hydrogen additives on singlet oxygen kinetics in argon–oxygen–hydrogen mixtures has been analysed. The simulation has shown that fast quenching of O2(a 1Δg) by atomic hydrogen is possible due to significant gas heating in the discharge that can significantly limit the yield of singlet oxygen in hydrogen-containing mixtures.