Abstract: Particle-in-cell/Monte Carlo collision (PIC/MCC) simulations are performed to investigate the asymmetric secondary electron emission (SEE) effects when electrons strike two different material electrodes in low pressure capacitively coupled plasmas (CCPs). To describe the electron-surface interactions, a realistic model, considering the primary electron impact energy and angle, as well as the corresponding surface property-dependent secondary electron yields, is employed in PIC/MCC simulations. In this model, three kinds of electrons emitted from the surface are considered: (i) elastically reflected electrons, (ii) inelastically backscattered electrons, and (iii) electron induced secondary electrons (SEs, i.e., delta-electrons). Here, we examined the effects of electron-surface interactions on the ionization dynamics and plasma characteristics of an argon discharge. The discharge is driven by a voltage source of 13.56MHz with amplitudes in the range of 200-2000V. The grounded electrode material is copper (Cu) for all cases, while the powered electrode material is either Cu or silicon dioxide (SiO2). The simulations reveal that the electron impact-induced SEE is an essential process at low pressures, especially at high voltages. Different electrode materials result in an asymmetric response of SEE. Depending on the instantaneous local sheath potential and the phase of the SEE, these SEs either are reflected by the opposite sheath or strike the electrode surface, where they can induce delta-electrons upon their residual energies. It is shown that highly energetic delta-electrons contribute significantly to the ionization rate and a self-bias forms when the powered electrode material is assumed to be made of SiO2. Complex dynamics is observed due to the multiple electron-surface interaction processes and asymmetric yields of SEs in CCPs.

Abstract: Superimposition of dual frequencies (DFs) is one of the methods used for controlling plasma distribution in an inductively coupled plasma (ICP) source. The effects of a superimposed DF on the argon plasma characteristics have been investigated using a two-dimensional self-consistent fluid model. When both currents are fixed at 6A, the plasma density drops with decrease in one of the source frequencies due to less efficient heating and the plasma uniformity improves significantly. Moreover, for ICP operated with superimposed DFs (i.e., 4.52MHz/13.56MHz and 2.26MHz/13.56MHz), the current source exhibits the same period as the low frequency (LF) component, and the plasma density is higher than that obtained at a single frequency (i.e., 4.52 and 2.26MHz) with the same total current of 12A. However, at superimposed current frequencies of 6.78MHz/13.56MHz, the plasma density is lower than that obtained at a single frequency of 6.78MHz due to the weaker negative azimuthal electric field between two positive maxima during one period of 6.78MHz. When the superimposed DF ICP operates at 2.26 and 13.56MHz, the rapid oscillations of the induced electric field become weaker during one period of 2.26MHz as the current ratio of 2.26MHz/13.56MHz rises from 24A/7 A to 30A/1 A, and the plasma density drops with the current ratio due to weakened electron heating. The uniformity of plasma increases due to sufficient diffusion under the low-density condition.

Abstract: A hybrid model is used to investigate the fragmentation of C4F8 inductive discharges. Indeed, the resulting reactive species are crucial for the optimization of the Si-based etching process, since they determine the mechanisms of fluorination, polymerization, and sputtering. In this paper, we present the dissociation degree, the density ratio of F vs. CxFy (i.e., fluorocarbon (fc) neutrals), the neutral vs. positive ion density ratio, details on the neutral and ion components, and fractions of various fc neutrals (or ions) in the total fc neutral (or ion) density in a C4F8 inductively coupled plasma source, as well as the effect of pressure and power on these results. To analyze the fragmentation behavior, the electron density and temperature and electron energy probability function (EEPF) are investigated. Moreover, the main electron-impact generation sources for all considered neutrals and ions are determined from the complicated C4F8 reaction set used in the model. The C4F8 plasma fragmentation is explained, taking into account many factors, such as the EEPF characteristics, the dominance of primary and secondary processes, and the thresholds of dissociation and ionization. The simulation results are compared with experiments from literature, and reasonable agreement is obtained. Some discrepancies are observed, which can probably be attributed to the simplified polymer surface kinetics assumed in the model.

Abstract: We perform first-principles calculations to investigate the preferred positions of B and P dopants, both neutral and in their preferred charge state, in Si and Si/SiO2 core-shell nanowires (NWs). In order to understand the observed trends in the formation energy, we isolate the different effects that determine these formation energies. By making the distinction between the unrelaxed and the relaxed formation energy, we separate the impact of the relaxation from that of the chemical environment. The unrelaxed formation energies are determined by three effects: (i) the effect of strain caused by size mismatch between the dopant and the host atoms, (ii) the local position of the band edges, and (iii) a screening effect. In the case of the SiNW (Si/SiO2 NW), these effects result in an increase of the formation energy away from the center (interface). The effect of relaxation depends on the relative size mismatch between the dopant and host atoms. A large size mismatch causes substantial relaxation that reduces the formation energy considerably, with the relaxation being more pronounced towards the edge of the wires. These effects explain the surface segregation of the B dopants in a SiNW, since the atomic relaxation induces a continuous drop of the formation energy towards the edge. However, for the P dopants, the formation energy starts to rise when moving from the center but drops to a minimum just next to the surface, indicating a different type of behavior. It also explains that the preferential location for B dopants in Si/SiO2 core-shell NWs is inside the oxide shell just next to the interface, whereas the P dopants prefer the positions next to the interface inside the Si core, which is in agreement with recent experiments. These preferred locations have an important impact on the electronic properties of these core-shell NWs. Our simulations indicate the possibility of hole gas formation when B segregates into the oxide shell.

Abstract: Computer simulations based on the particle-in-cell/Monte Carlo collision method are performed to study the plasma characteristics and especially the transition in electron heating mechanisms in a hybrid direct current (dc)/dual-frequency (DF) capacitively coupled CF 4 discharge. When applying a superposed dc voltage, the plasma density first increases, then decreases, and finally increases again, which is in good agreement with experiments. This trend can be explained by the transition between the four main heating modes, i.e., DF coupling, dc and DF coupling, dc source dominant heating, and secondary electron dominant heating.

Abstract: A theoretical study has been conducted of ultrashort pulsed laser induced electron emission from an aluminum surface. Electron emission fluxes retrieved from the commonly employed Fowler-DuBridge theory were compared to fluxes based on a laser-induced non-equilibrium electron distribution. As a result, the two-and three-photon photoelectron emission parameters for the Fowler-DuBridge theory have been approximated. We observe that at regimes where photoemission is important, laser-induced electron emission evolves in a more smooth manner than predicted by the Fowler-DuBridge theory. The importance of the actual electron distribution decreases at higher laser fluences, whereas the contribution of thermionic emission increases. Furthermore, the influence of a space charge effect on electron emission was evaluated by a one dimensional particle-in-cell model. Depending on the fluences, the space charge reduces the electron emission by several orders of magnitude. The influence of the electron emission flux profiles on the effective electron emission was found to be negligible. However, a non-equilibrium electron velocity distribution increases the effective electron emission significantly. Our results show that it is essential to consider the non-equilibrium electron distribution as well as the space charge effect for the description of laser-induced photoemission. (C) 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4729071]

Abstract: Molecular dynamics simulations using improved TersoffBrenner potential parameters were performed to investigate Cl+ etching of a {2×1} reconstructed Si(100) surface. Steady-state Si etching accompanying the Cl coverage of the surface is observed. Furthermore, a steady-state chlorinated reaction layer is formed. The thickness of this reaction layer is found to increase with increasing energy. The stoichiometry of SiClx species in the reaction layer is found to be SiCl:SiCl2:SiCl3 = 1.0:0.14:0.008 at 50 eV. These results are in excellent agreement with available experimental data. While elemental Si products are created by physical sputtering, most SiClx (0<x<4) etch products are produced by chemical-enhanced physical sputtering.

Abstract: We study the properties of a capacitive 13.56 MHz discharge with a mixture of Ar/C₂H₂ taking into account the plasmochemistry and growth of heavy hydrocarbons. A hybrid model was developed to combine the kinetic description for electron motion and the fluid approach for negative and positive ion transports and plasmochemical processes. A significant change in plasma parameters related to injection of 5.8% portion of acetylene in argon was observed and analyzed. We found that the electronegativity of the mixture is about 30%. The densities of negatively and positively charged heavy hydrocarbons are sufficiently large to be precursors for the formation of nanoparticles in the discharge volume.

Abstract: The role of the Cu atoms sputtered from the cathode material in a cylindrical hollow cathode discharge (HCD) and the corresponding Cu+ ions are studied with a self-consistent model based on the principle of Monte Carlo (MC) and fluid simulations. In order to obtain a more realistic view of the discharge processes, this model is coupled with other submodels, which describe the behavior of electrons, fast Ar atoms, Ar+ ions, and Ar metastable atoms, also based on the principles of MC and fluid simulations. Typical results are, among others, the thermalization profile of the Cu atoms, the fast Cu atom, the thermal Cu atom and Cu+ ion fluxes and densities, and the energy distribution of the Cu+ ions. It was found that the contribution of the Ar+ ions to the sputtering was the most significant, followed by the fast Ar atoms. At the cathode bottom, there was no net sputtered flux but a net amount of redeposition. Throughout the discharge volume, at all the conditions investigated, the largest concentration of Cu atoms was found in the lower half of the HCD, close to the bottom. Penning ionization was found the main ionization mechanism for the Cu atoms. The ionization degree of copper atoms was found to be in the same order as for the argon atoms (10-4). (c) 2005 American Institute of Physics.