Modeling of a hemispherical inductively coupled plasma using neural network

The temporal evolution of plasma density and electron temperature in high power impulse magnetron sputtering discharges has been measured using the Langmuir probe and laser Thomson scattering techniques. Measurements were performed (nonsimultaneously) at two positions within the plasma, in the low magnetic field strength region on the discharge axis and in the high magnetic field strength region of the magnetic trap, for peak power densities of 450 W cm–2 and 900 W cm−2, respectively. The maximum plasma densities and temperatures were 6.9 × 1019 m−3 and 3.7 eV in the pulse-on time, and values decayed to 4.5 × 1017 m−3 and 0.1 eV at times up to 250 μs into the afterglow. The results indicate that although intrusive, the Langmuir probe can provide a good indication of electron properties in regions of different electron magnetization in the discharge.

The electron temperature and plasma density at the lateral distance of 125 /spl mu/m from the center of sustaining electrode gap have been investigated by a Langmuir probe along with the high-speed discharge image in coplanar alternating current plasma display panels. The plasma density at the lateral distance of 125 /spl mu/m from the center of sustaining electrode gap is shown to be maximum value of 3.7/spl times/10/sup 11/ cm/sup -3/, whereas the electron temperature is measured to be decreased from 1.8 to 0.8 eV as the gas pressure increases from 150 to 400 torr in this experiment. It is noted that the electron temperatures measured by the Langmuir probe and high-speed image camera are in good agreement with each other within 5% error limit.

Electron temperature and density in pure He, Ar, and Xe plasmas are estimated by zero-dimensional particle and power balance equations and measured by a Langmuir probe. Both of the modeling and experimental results show that the He (Xe) plasma has the highest (lowest) electron temperature and lowest (highest) electron density for a given fill pressure and source power. We find that the electron temperature is weakly dependent on the rf power, and thus the electron density can be controlled using the rf power. The electron temperature and density are also modeled and measured in mixtures of two noble gas species. We find that the electron temperature can be controlled by altering the composition of the noble gas mixture. Thus modulation of noble gas admixture ratios and rf power allows the electron density and temperature to be controlled independently. This independent control is shown to maintained with the addition of up to 20% partial pressure of oxygen, suggesting binary noble gas admixtures may provide additional control of dissociation kinetics.

An electrical method is proposed to measure the dielectric film thickness, electron temperature, and plasma density for deposition plasmas. In this method, a square voltage is applied to a flat probe coated with a dielectric film, and the transient voltage of a series capacitor connected to the probe is measured. The thickness of the dielectric film is obtained from the transient voltage because the applied voltage is divided across the dielectric film and the series capacitor. The deposited dielectric film causes a change in transient voltage of the series capacitor. A circuit model with a nonlinear sheath, a series capacitor, and the dielectric films is suggested to obtain plasma density and electron temperature. The time response of transient voltage is related to plasma density, electron temperature and total capacitance of the series capacitor and the dielectric film. Two square voltages with different amplitudes were applied to measure plasma density and electron temperature. Experiments were conducted in inductively coupled plasma. Various capacitors were used to replace the dielectric films, and a flat probe coated with Al2O3 was used to verify the measurement of dielectric film thickness, showing the high accuracy of our method. In addition, the electron temperature from our method is in good agreement with that from electron energy distribution functions, and the plasma density is in good agreement with that from orbital motion limited theory. Therefore, this method would be useful for monitoring plasma parameters and deposited film thickness in industrial plasma processing.

Effect of neutral pressure on the He plasma flow measurement in a linear device

An algorithm for diagnosing the electron density and temperature of helium plasma at atmospheric pressure has been developed based on a revised helium collisional-radiative (CR) model. Atomic collision processes are included, and part of the atomic data of electron collision processes in the conventional CR model has been updated to expand its valid pressure. The algorithm uses eight emission lines in the visible-wavelength range as inputs to determine the electron density, electron temperature, and number density of the two metastable states by fitting the number density of the states corresponding to the emission lines. The algorithm has a considerably small theoretical error. In the microwave-discharged low-pressure helium plasma experiment, the results obtained with the algorithm agreed well with the results obtained with the probe method. The electron density and temperature of the atmospheric-pressure helium plasma obtained with the algorithm agreed well with the results of the continuum spectrum analysis.

The plasma of a radio frequency (rf) magnetron discharge used for deposition of TiOx thin films was separated by a mesh grid into source (negative glow) and diffusion (deposition) plasmas. Langmuir probes were used to measure the potential, electron density, and electron temperature of both plasmas. While the change of the grid bias affected the diffusion plasma parameters, it did not change the source plasma parameters (except the potential). The use of the biased grid resulted in a decrease of temperature and an increase of density of plasma electrons in the film deposition region. With an increase of the grid biasing potential, the plasma potential increased in order to maintain a negative bias voltage across the grid sheath, the electron temperature increased, and the electron density decreased. The biased-grid-induced change of the plasma potential and the bias voltage across the wall sheath is explained based on the model of the asymmetric double probe, which is formed by the discharge chamber wall and the grid. The effect of the biased grid on the diffusion plasma parameters is discussed accounting for the particle transport phenomena in the system.

Summary form only given. Low pressure microwave plasma sources used for materials processing generally operate as overdense plasmas with the plasma density greater than the critical density. These sources can be operated with a static magnetic field that provides for ECR heating or without a static magnetic field via ohmic, resonance and other non-ohmic and stochastic heating mechanisms. Even when ECR strength magnetic fields are present these other heating mechanisms that occur in unmagnetized plasmas can be important and may even dominate. This paper examines via a two-dimensional, self-consistent microwave field and plasma model the heating of low pressure, overdense, magnetized and unmagnetized plasma discharges. Low-pressure (collision frequency /spl Lt/ excitation frequency) microwave plasma simulations that model the spatial variation of the microwave heating fields and plasma discharge are difficult to use in the local regions where the plasma frequency is near the excitation frequency. In these regions resonance effects occur and the microwave electric field can become large. Because of the localized nature of the resonance (high microwave field strength) region, stochastic (non-collisional) heating can occur as the electrons are accelerated/heated in this region and/or transverse through this resonance region via their initial momentum. This paper explores the self-consistent modeling of microwave discharges including resonance effects and stochastic heating effects. Further the models are constructed to closely match an experimental system that has been extensively characterized. This experimental system is a 2.45 GHz resonant cavity plasma source that has been studied while running argon discharges at pressures of 4-60 mTorr using Langmuir probes to determine the plasma density and electron temperature and using microwave field probes and optical diagnostic techniques to measure the microwave field strength.

Chlorine-based high density plasmas, commonly used in the etching of elemental and compound semiconductors, are characterized using mass spectrometry, optical emission spectroscopy, and electrostatic probes. Plasma fluxes are characterized by three-dimensional Langmuir probe measurements and optical emission spectroscopy. The flux is further characterized at the substrate platen by mass spectrometry to determine its makeup in terms of charged or neutral species and atomic or molecular species. Langmuir probe investigations show variations in electron temperature (2–6 eV), plasma density (1×1010 to 1×1012 cm−3), and plasma potential (5–25 V) as process conditions (microwave power, total pressure, and fraction of Cl2 in Ar) and measurement location are varied. Concurrent optical emission spectroscopy measurements of ionized species are in general agreement with Langmuir probe results. Further, optical emission spectroscopy of neutral and ionized species provides global insight into the variation of atomic/molecular fractions in the plasma as it is transported to the substrate processing region. At the substrate, mass spectrometric characterizations show Cl+ and Ar+ dominating the flux for low pressure and high powers, while Cl2 and Ar dominate at high pressure and low power. For Cl2 fractions greater than 25% molecular chlorine begins to dominate the flux to the substrate. These observations of processing space are discussed with respect to implications on semiconductor etching and regions most suitable to high rate, anisotropic processing conditions are identified.

Summary form only given, as follows. The electron temperature in a pulsed inductively coupled plasma (ICP) can be significantly lowered when it is operated in a pulsed mode. Lower electron temperature, together with the increase in negative ions, was found to be able to reduce notching effects in etching processes. The local plasma properties in a pulsed ICP can be measured by Langmuir probes. The probe current can be measured as a function of delay time for each pulse cycle at different but fixed bias voltages. The current versus delay time data for different bias voltages can then be used to reconstruct the typical I-V curves for each delay time. The plasma density and electron temperature as a function of delay time can then be determined from these I-V curves. For an Ar plasma with a chamber height of 40 cm and diameter of 40 cm, operated at 30 mtorr gas pressure, 13.56 MHz RF, and 10 kHz pulse rate at 50% duty cycle (50 /spl mu/s on and 50 /spl mu/s off), both the temperature and the plasma density were found to decay initially very fast and then much slower in the power off period. The decay time constants are expected to depend on the gas pressure which affects the ionization rate and diffusion coefficient, and also on the chamber size which affects the loss rate of the plasma. Experiments will be carried out for various gas pressures and chamber heights to see how the decay time constants vary with these two parameters.

Plasma processing of AISI 304 stainless steel using radio frequency hollow cathode discharge

In this paper, a 105GHz millimeter wave interferometer system is used to measure the electron density and temperature of an atmospheric pressure helium plasma driven by submicrosecond pulses. The peak electron density and electron-neutral collision frequency reach 8×1012cm−3 and 2.1×1012s−1, respectively. According to the electron-helium collision cross section and the measured electron-neutral collision frequency, the electron temperature of the plasma is estimated to reach a peak value of about 8.7eV.

Electron beam-generated plasmas (EBPs) have been used to modify the surface properties. In certain applications, EBPs are transversely confined and their properties are of value to the treatment. In this paper, the characteristics of an electron beam-generated argon plasma, confined within a narrow gap, are investigated using a two-dimensional particle-in-cell simulation. The employed particle-in-cell/Monte Carlo collision model accounts for the electron and ion kinetics, as well as collisions between electrons and the background gas, including the elastic scattering, excitation, and impact ionization. Our simulations reveal a strong correlation between the plasma density and the beam density within the plasma bulk. The excitation of obliquely growing waves is observed, which is found to have a significant impact on the transport of beam electrons, thereby leading to the non-uniformities of plasma density and electron temperature. Specifically, the obliquely growing waves increase the local plasma density while reducing the electron temperature. These contrasting effects compensate for each other, and therefore, to some extent, smooth out the distributions of ion flux and energy flux. We further examine the variations of plasma parameters with respect to the beam current density, beam energy, and gas pressure. Increasing the beam current density or decreasing the beam energy results in higher plasma density and electron temperature, while increasing pressure leads to a higher plasma density but electron temperature scarcely changes. Based on the simulation results, we propose an approach to achieve independent control of the ion flux and energy flux by adjusting beam current density, beam energy, and pressure.

The Langmuir probe (LP) is generally accepted as an effective and relatively simple <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in situ</i> space plasma instrument (plasma density, electron temperature, and spacecraft potential). As LPs transition to small spacecraft, their implementation encounters new technical challenges. For example, a negative charge is induced on the spacecraft while positively biasing an LP due to the small surface area ratios (spacecraft surface area to probe surface area ≪ 1000). This results in a varying spacecraft potential that degrades the accuracy of electron temperature and electron density measurements, reducing the LP’s effectiveness as a diagnostic tool. To mitigate the effects of this spacecraft charging, the twin-probe method (TPM) was developed. The TPM corrects LP measurements with tracked spacecraft potentials, measured by a separate high-impedance probe. By accounting for the changes in the spacecraft potential, the LP sweeps can be reconstructed to provide more accurate measurements of the ambient plasma’s properties. Here, we detail the TPM and present laboratory experiments that study its effectiveness and summarize constraints. Through these experiments, we found that it is possible to correct for spacecraft charging effects that can cause deviations in temperature and density as large as 20% and 136%, respectively. Furthermore, we will demonstrate a clear correlation between area ratio, spacecraft charging, and the negative impact on temperature and density measurement accuracy.

Surface wave-sustained plasma (SWP) is one of the low-pressure, high- density plasma. Applying this technique, diamond-like carbon (DLC) films with excellent characteristics can be prepared by physical vapor deposition (PVD) method. However, the films' application is restricted in some degree, because it is difficult to control the film properties. In this paper, SWP was excited along a conductive rod at a frequency of 2.45 GHz without magnetic fields around the chamber wall. The fundamental theories of plasma diagnostic were presented and plasma properties were studied with a Langmuir probe under the conditions of depositing DLC films by PVD method with a graphite target. Plasma density, electron temperature, plasma potential and target current were measured at difference technique parameters such as gas pressure, microwave power, and so on. As a result, it was proved that plasma properties are greatly affected by microwave power, target voltage and argon gas pressure in chamber. The gas mass flow rate had almost no effect on plasma characters. At the same time, the results indicated that electron density is up to 10¹¹-10¹²cm^-3 even at the low pressure of 1 Pa.