22nd International Symposium on Plasma Chemistry July 5-10, 2015; Antwerp, Belgium Global (volume-averaged) model of oxygen plasmas E.H. Kemaneci1, J. van Dijk2 and R.P. Brinkmann1 1 2 Ruhr University Bochum, Theoretical Electrical Engineering, Bochum, Germany Department of Applied Physics, Eindhoven University of Technology, Eindhoven, the Netherlands Abstract: In this contribution, we investigate oxygen plasmas using a global (volumeaveraged) model. The investigation mainly focuses on the population of the vibrational quantum levels of oxygen molecule in radio-frequency plasmas, where a ladder-like dissociation is included in addition to the various e-V, V-V and V-T transitions. The plasma interactions with the wall, such as recombination of oxygen atom and ozone formation, are also investigated. Furthermore, we cover both the continuous and pulsed power input modes, where we observe agreement with various experimental data of oxygen plasmas in literature. Keywords: oxygen plasma, global model, vibrational quantum levels 1. General Oxygen plasmas are widely used for technological purposes and scientific investigations. These plasmas include variety of species, such as, different types of ions and electronic energy levels of atomic and molecular species. Additionally, the vibrational degree of freedom of the molecules contributes to this variety with the corresponding quantum energy levels. Although most of these species are covered in previous studies [1-5], a detailed and up-to-date study on the population of the vibrational levels as well as their role in the plasmas are still missing. In this study we address this issue by providing a detailed investigation. The chamber wall is also an important factor in the plasma behaviour. For example, it is shown that the recombination of oxygen atoms at the chamber wall plays major role on the plasma parameters [5]. Recent studies also reveal ozone formation at certain types of wall materials contacting the plasma. The role of these wall reactions are also investigated within the framework of this contribution. Pulse-modulated power input of the oxygen plasmas is often preferred mostly due to the better control of the ion wall fluxes and reduced thermal load. We also analyse the temporal plasma chemical kinetics in the pulse-on and pulse-off regions of the modulation period. 2. Global (volume-averaged) model Global (volume-averaged) models are widely used in literature to investigate oxygen plasmas [1-3, 5]. These models are often preferred over detailed spatially-resolved models due to the low cost in the computational load and smaller simulation durations. Although, the main focus of their usage is the inductively (or capacitively) coupled plasmas, they are also employed to analyse plasmas induced by the microwaves [6, 7]. The global models include the particle balance equations and the electron energy balance equation, that are both spatially averaged over the plasma volume with the assumption of spatial P-I-2-77 homogeneity [8]. These equations describe the particle densities and the electron temperature, whereas the heavyparticle temperature and the pressure are externally provided in the model. The volume-averaged particle balance equation of a particle i can be written as 𝒅𝒏𝒊 𝒅𝒅 = 𝑺𝒊 , where n i represents the density and S i represents the source of the particle. The source term includes the volume-averaged chemical reactions in the plasma bulk, the wall induced reactions, as well as the transport losses at the plasma boundary. The volume-averaged electron energy balance equation is given by the relation 𝟑 𝒅(𝟐𝒏𝒆 𝑻𝒆 ) 𝒅𝒅 = 𝑸𝒆 , where T e is the electron temperature and Q e is the electron energy source. The source term includes the absorbed power, the energy losses in the plasma bulk due to various elastic and inelastic collisions, as well as the energy losses at the plasma boundary. The simulations are carried on the modular plasma simulation platform Plasimo developed in Technical University of Eindhoven. The set of ordinary differential equations are iteratively solved for a given set of initial conditions. In the iterative process Livermore Solver for Ordinary Differential Equations (LSODA) is used. 3. Experimental comparison The simulation results are compared with various experimental data of oxygen plasmas in literature and we show only two example cases here. The first case is the electron and negative ion densities measured by Stoffels et al. [9] on a capacitively coupled plasma. In Fig. 1, the measured and calculated densities are shown with respect to the pressure. The second case is the electron, negative 1 ion and atomic oxygen densities measured by Baeva et al. [4] on a microwave waveguide plasma, sustained by pulse-modulated power input. The time-resolved densities and the simulation data of this setup are shown in Fig. 2. In addition to the direct dissociation of the molecular oxygen due to the electron impact, Cacciatore et al. suggests an additional mechanism to dissociate molecular oxygen. This mechanism excites the vibrational levels ladder-like manner mostly due to the e-V processes. After a pseudo level has been reached the dissociation occurs via various vibrational processes. 5. Acknowledgements The authors gratefully acknowledges the support by BmBF via PluTO+. Fig. 1. Electron and negative ion densities measured by Stoffels et al. [9] (points) on a capacitively coupled plasmas and the corresponding simulation data (lines). Fig. 2. Electron, negative ion and atomic oxygen densities measured by Baeva et al. [4] (points) on a pulse-modulated microwave waveguide plasma and the corresponding simulation data (lines). 4. Vibrational chemical kinetics The vibrational quantum levels of the molecular oxygen is created via vibrational excitation by electron impact. The direct vibrational excitation is in general an inefficient process. On the other hand, the resonant vibrational excitation, in which the incident electron firstly attaches to the molecule and forms a temporary resonant anionic state, may effectively populate these levels. These levels are heavily depopulated due to the V-T processes with oxygen atoms [4]. This fact is solely responsible in their absence in the earlier oxygen plasma models [1-5] by reducing their importance in the chemical kinetics. 2 6. References [1] E.G. Thorsteinsson and J.T. Gudmundsson. “The low pressure Cl2/O2 discharge and the role of ClO”. Plasma Sources Sci. Technol., 19, 055008 (2010) [2] S. Panda, D.J. Economou and M. Meyyappan. “Effect of metastable oxygen molecules in high density power-modulated oxygen discharges”. J. Appl. Phys., 87, 8323-8333 (2000) [3] T.H. Chung, H.J. Yoon and D.C. Seo. “Global model and scaling laws for inductively coupled oxygen discharge plasmas”. J. Appl. Phys., 86, 3536 (1999) [4] M. Baeva, X. Luo, B. Pfelzer, T. Repsilber and J. Uhlenbusch. „Experimental investigation and modelling of a low-pressure pulsed microwave discharge in oxygen”. Plasma Sources Sci. Technol., 9, 128 (2000) [5] C. Lee and M.A. Lieberman. “Global model of Ar, O2, Cl2, and Ar/O2 high-density plasma discharges”. J. Vacuum Sci. Technol. A, 13, 368-380 (1995) [6] A. Yanguas-Gil, J. Cotrino and A.R. GonzalezElipe. “Global model of a low pressure ECR microwave plasma applied to the PECVD of SiO2 thin films”. J. Phys. D: Appl. Phys., 40, 3411-3422 (2007) [7] E.H. Kemaneci. “Modelling of Plasmas with Complex Chemistry: Application to Microwave Deposition Reactors”. PhD thesis. (Eindoven, the Netherlands: Eindhoven University of Technology) (2014) [8] P. Chabert and N. Braithwaite. Physics of RadioFrequency Plasmas. (Cambridge: Cambridge University Press) (2011) [9] E. Stoffels, W.W. Stoffels, D. Vender, M. Kando, G.M.W. Kroesen and F.J. de Hoog. “Negative ions in radio-frequency plasma”. Phys. Rev. E, 51, 2425-2435 (1995) P-I-2-77
© Copyright 2026 Paperzz