Inductively Coupled Plasma Reactor Research Articles

Numerical Investigation for $\hbox{CF}_{4}$ Decomposition Using RF Low-Pressure Plasma

Among perfluorocompounds, is considered to be one of the most stable and difficult-to-decompose gases. We investigated the decomposition of CF4 using the inductively coupled plasma (ICP) reactor with radio frequency (RF) power supply, which was used for semiconductor cleaning process as well as decomposition. This technology was confirmed to achieve an extremely high efficiency and more economical system in comparison with the conventional system. The purpose of this paper is to perform numerical simulation of CF4 decomposition using RF low-pressure plasma. The experimental investigation of CF4 the decomposition was performed to validate the computed results, and the reaction products such as CO, CO2, and COF2 upon the CF4 decomposition were measured CO2 using a analyzer and a Fourier transform infrared spectrophotometer. Then, the numerical simulations of the CF4 decomposition using commercially available code were performed to obtain the gas temperature, electron temperature, electron number density, gas velocity, and chemical species number density distribution to cope with CF4 the decomposition in the ICP reactor.

On the applicability of self-consistent global model for the characterization of Cl 2/Ar inductively coupled plasma

This work reports the results on the influence of gas mixing ratio, gas pressure (0.26–3.3 Pa) and input power (400–900 W) on the Cl 2/Ar plasma parameters in the planar inductively-coupled plasma reactor. The investigation combined plasma diagnostics by Langmuir probe and quadrupole mass-spectroscopy with plasma modelling given by the self-consistent global (0-dimensional) model with Maxwellian approximation for the electron energy distribution function. It was shown that, for the given range of experimental conditions, the model showed an outstanding agreement with the experiments and provided the near-to-adequate data on kinetics of plasma active species, their densities and fluxes.

Experimental characterization of an inductively coupled acetylene/hydrogen plasma for carbon nanofiber synthesis

A plasma-enhanced chemical-vapor deposition process was employed to synthesize carbon nanofibers (CNFs) on glass substrates patterned with Ni catalytic films. At the gas pressure of 20mTorr and the substrate temperature (surface) of ∼500°C, the isolated and vertically aligned carbon nanofibers have been successfully synthesized. This paper reports experimental investigation of plasma properties characterized by the optical emission spectroscopy of the spectral line intensities of the various species such as hydrogen, C2, and CH, as well as the rf characteristics at the biased substrate stage measured by an impedance meter. The measurement results reveal that the C2 density increases with the acetylene/hydrogen flow ratio and the inductively coupled plasma (ICP) source power, as expected. The atomic hydrogen density, however, decreases with the flow ratio but increases with the ICP power. The resulting growth rate of CNFs increases with the C2 density if atomic hydrogen density also increases accordingly, e.g., as the ICP power increases. The trend is reversed if the atomic hydrogen density decreases, due to too much amorphous carbon (a-C) layer formed as a result of oversupply of carbon but not enough atomic hydrogen to remove a-C. The experimental results also show that the etch effect upon the effective removing of the a-C on the surface of catalytic nanoparticles is further enhanced by ion bombardment, e.g., when either the flux (or current) or energy of the ions incident on the substrate surface increases, to give rise to an increase in growth rate. In our ICP reactor, the ion current increases with the ICP power, but it changes little when the bias power is varied. The ion energy increases with the bias power, but it decreases as the ICP power increases while the bias power is fixed. The latter one is because the plasma density increases with the inductively coupled plasma power.

Development of a Dual Inductively Coupled Plasma Source for Direct and Remote Plasma Generation in a Reactor

A dual inductively coupled plasma (ICP) system consists of a remote ICP reactor with small volume and a main ICP reactor with a substrate. Two ICP antennas were connected in parallel and a variable capacitor (Cvar) was installed in series at the end of the main ICP antenna. By adjusting the capacitance of the variable capacitor, the plasma densities in the remote region and the main region are controlled. For the remote region, the plasma was considerably changed such that it had high density and the electron temperature was higher than that in the main region because of its small volume. As such, reactive species in the remote region appeared to be effectively generated. The dual ICP system was applied to Si etching. It was observed that Si etch rate increased by 20% as the plasma density in the remote region increased, even though the plasma density in the main region decreased. This might be understood by considering the role of the remote ICP as a radical generator.

<tex>$hboxCF_4$</tex>Decomposition of Flue Gas From Semiconductor Process Using Inductively Coupled Plasma

Flue gases from the semiconductor process, SF/sub 6/, NF/sub 3/, perfluorocarbons (PFCs) such as CF/sub 4/ and C/sub 2/F/sub 6/, and hydrofluorocarbons such as CHF/sub 3/ are being regulated internationally because their gases have extremely large global warming potential in comparison to CO/sub 2/, and they have a long lifetime. PFCs are used for wafer etching and cleanup of chemical vapor deposition chambers. CF/sub 4/ is one of the most stable gases among PFCs and its decomposition is extremely difficult. The purpose of this paper is to develop a PFCs removal system which can be used as the same power supply for both plasma processing and a CF/sub 4/ emission cleanup system in semiconductor manufacturing processes. This achieves higher efficiency and is more economical than conventional systems. CF/sub 4/ decomposition was investigated at low pressure (29/spl sim/53 Pa) using the inductively coupled plasma (ICP) reactor. When the total flow rate was below 0.189 normal liters per minute and O/sub 2/ concentration exceeded the 0.9 stoichiometric ratio, the complete CF/sub 4/ decomposition efficiency was achieved at 1.2 kW. The amount of O/sub 2/ should be a 1/spl sim/1.45 stoichiometric ratio of CF/sub 4/ to achieve good CF/sub 4/ decomposition. The argon enhanced the decomposition of CF/sub 4/. On the other hand, the helium was insignificant. CF/sub 4/ decomposition efficiency decreased in the presence of N/sub 2/. For CF/sub 4/ decomposition, the energy efficiency for CF/sub 4/ decomposition was 21.0 g/kWh at optimal condition using the ICP reactor. From the Fourier Transform Infrared Spectrophotometer analysis, CO and carbonyl fluoride (COF/sub 2/) were produced except for CO/sub 2/. COF/sub 2/ is toxic, however, it hydrolyzes into HF and CO/sub 2/, so it is easy to eliminate COF/sub 2/ using the scrubber which was located downstream of the dry vacuum pump.

Coupling of plasma and flow in materials processing

Flow fields of radicals and gases are necessary in simulating electron-neutral species collisions whereas rates of dissociation and ionization are required in calculating the flow fields of neutral species. This is a coupling between plasma and flow. A method to treat the coupling is proposed and employed to examine the structures of plasma and flow in an inductively-coupled plasma reactor. Gas is chlorine.

Electron beam fluorescence temperature measurements of N2 in a semiconductor plasma reactor

The rotational temperature of nitrogen molecules in inductively coupled plasma (ICP) discharge has been measured using the electron beam fluorescence (EBF) technique. The neutral gas temperature is an important parameter in understanding the energy balance and neutral radical uniformity in plasma processing. The EBF technique has been used to study rarefied flows, but has not been applied previously to characterize a semiconductor plasma reactor. In this work an electron beam was integrated into an inductively coupled semiconductor plasma reactor, and was used to excite neutral nitrogen molecules from the N2X 1Σg+ state into the excited molecular ion N2+B 2Σu+ state. The electron beam excitation process maps the rotational population distribution of the original ground state into the excited molecular ion state following a dipole model with ΔK=±1. The rotational temperature, which is thought to be equivalent to the neutral gas translational temperature under plasma reactor conditions, can be determined from the rotational intensity distribution of the fluorescence spectrum from the N2+B 2Σu+ state. For the range of neutral gas temperatures found in the plasma reactor a dipole model of the excitation-emission process provides good agreement with the measured spectrum. A least-square fitting procedure comparing the model and measured emission spectra is used to determine the rotation temperature. In this article the first EBF measurements of gas heating in an ICP reactor at different gas pressures (20, 50 mTorr) and plasma source input powers (250, 500, and 1000 W) are reported.

RFプラズマによる半導体製造装置排ガスの分解(CF_4分解の実験と数値シミュレーション)(OS.4 機能性流体と先端融合化)

We investigated the decomposition of CF_4 in an exhaust gas from semiconductor manufacturing equipment using the inductively coupled plasma (ICP) reactor with RF power supply. The purpose of this study is to develop the PFCs removal system which can be used as the same power supply for both plasma processing and CF_4 emission clean-up system in semiconductor manufacturing processes. This leads to achieve higher efficiency and more economical than conventional systems. CF_4 .decomposition was carried out as a parameter of the ratio of CF_4 to O_2, pressure and RF power. We also measured CO_2 concentration after CF_4 decomposition in order to examine whether the decomposition process is based on the reaction that we intend. And we calculated electron temperature and electron density distribution in ICP reactor.

BCN coatings at low temperature using PACVD: capacitive vs. inductive plasma coupling

In this contribution, BCN coatings were deposited on polycarbonate by means of a capacitively or inductively coupled r.f.-PACVD (capacitive coupled plasma (CCP) vs. inductively coupled plasma (ICP)) using the metal-organic compound trimethylborazine as precursor with the B:C:N ratio of 1:1:1. The influence of the plasma parameters on the properties of films will be discussed. Compared to the CCP reactor, the ICP reactor design allows, through low energy and high density ion bombardment, the deposition of transparent and hard coatings on polymers.

Theoretical investigation of the evolution of electron energy distribution functions in inductively coupled discharges

The inductively coupled plasma (ICP) device is one of the high-density plasma sources being used for the present integrated circuit etching process on an ultra-large scale. Here, we develop a Fokker–Planck code for the ICP source, and evaluate an electron energy diffusion coefficient based on the solution of the wave equations for an ICP reactor. The electron energy distribution function (EEDF) dependence on various external parameters, such as wave frequency, gas pressure and magnetic field, is investigated using the Fokker–Planck code. The effects of changing external parameters on ICP heating characteristics are also discussed. It is shown that the heating of low-energy electrons is enhanced with increasing system collisionality, which is defined as the ratio of collision frequency to effective wave frequency.

1441 Ar/HBr プラズマによるシリコンのエッチング特性

By use of an inductively-coupled plasma reactor for Ar/HBr gas, the effects of Ar/HBr mole fraction, biasing power, and gas pressure on the etch rate and the etch rate distribution are examined systematically. The etch rate strongly depends on the biasing power and the spatial uniformity of the etch rate becomes worse as the power increases. The etch rate by HBr plasma is smaller than that by Cl_2 plasma.

Effect of Etching Process Deviations and Photoresist Stripping on Contact Yield of Copper Dual Damascene Metallization

Dual damascene technology imposes strict requirements on etch and strip modules. One of them is to avoid exposure of copper on via bottoms during etching and subsequent plasma photoresist stripping (PRS) as it may lead to oxidation and contamination of copper and thus decrease contact yield. Therefore, etching should be reliably tuned for retaining some thickness of the protective cap dielectric in vias of targeted size. However, deviations of etching parameters in mass production may deteriorate selectivity and uniformity, resulting in partial or full exposure of copper in vias. In this case the effect of PRS on copper surfaces must be taken into consideration. This paper presents the study of copper surface oxidation and contamination and their effect on contact resistance and yield for three different PRS recipes: plasma at 40°C in a magnetically enhanced reactive ion etching (MERIE) reactor and and downstream plasma at 250°C in an inductively coupled plasma (ICP) reactor. It was found that PRS in the MERIE reactor and PRS in the ICP reactor oxidized the surface of copper to different depths but both processes did not reduce the yield of Kelvin vias and via chains. At the same time, plasma in an ICP reactor resulted in total failure of via chains in which copper was exposed before PRS (via size larger than 0.32 μm). Further implications of these results on dual-damascene process flow, mechanism of failure, and advantages of low-temperature PRS are discussed. © 2002 The Electrochemical Society. All rights reserved.

Three-dimensional simulation of an inductively coupled plasma reactor

A three-dimensional finite element fluid model and a corresponding simulation tool have been developed for studying azimuthal asymmetries and their effect on etch uniformity in inductively coupled plasma (ICP) reactors. For silicon etching with chlorine in an ICP reactor with a planar coil, four different cases were examined: (a) uniform power deposition without a focus ring, (b) uniform power deposition with a focus ring, (c) nonuniform power deposition without a focus ring, and (d) nonuniform power deposition with a focus ring. When etching is ion driven, the power deposition profile is most important for etch uniformity, because azimuthal nonuniformities in the ion production rate can persist even down to the wafer level. For uniform power deposition, the effect of asymmetric pumping becomes more important. A focus ring can play an important role in alleviating azimuthal nonuniformities, especially in the nonuniform power deposition cases. Gas inlets pointing parallel to the wafer plane introduce only local disturbances in the species profiles.

Two-dimensional pulsed-plasma simulation of a chlorine discharge

A two-dimensional (r,z) continuum model was developed to study the spatiotemporal dynamics of a pulsed power (square-wave modulated) chlorine discharge sustained in an inductively coupled plasma (ICP) reactor with a planar coil. The self-consistent model included Maxwell’s equations for the power deposition profiles coupled to the electron energy equation and the species mass balances. Simulation results showed separation of the plasma into an electronegative core and an electropositive edge during the active glow (power on) and the formation of an ion–ion plasma ∼15 μs into the afterglow (power off). During the early active glow, the negative ion flux was convection dominated near the quartz window of the ICP reactor due to the formation of large electrostatic fields, leading to a self-sharpening front propagating into the plasma. The negative ion density profiles were found to have a strong spatial dependence underlying the importance of spatial resolution in negative ion density measurements. The time dependent ion and radical flux uniformity was also studied. Simulation results were compared with experimental data and reasonable agreement was observed.

Modeling and experimental comparison of pulsed plasma deposition of aniline

A model for predicting deposition rates of plasma polymerized aniline films in a pulsed inductively-coupled plasma reactor is presented. The model proposes a two-step deposition mechanism: dissociation of monomer by electron collisions (fast) and subsequent (slow) diffusion of radicals toward the substrate. A free radical mechanism is proposed and the dissociation rate constant is calculated based on the best available cross section information. Results of the model are compared with axial and radial, polyaniline deposition rate profiles measured in the reactor. Plasma parameters such as initial electron temperature, electron density and dissociation cross-section were varied in order to arrive at a good agreement between the model and measured deposition rate data. The effect of these parameters on the model output is described.

Detection of C2 radicals in low-pressure inductively coupled plasma source for diamond chemical vapor deposition

Abstract The C 2 radical density was measured in a low-pressure, radio frequency (rf), inductively coupled plasma (ICP) employing a CH 3 OH/H 2 /H 2 O mixture. Measurement was carried out under the conditions where predominantly diamond can be formed. At the typical growth conditions for the low-pressure, rf ICP reactor used for diamond formation, the C 2 radical density in the ICP region was of the order of 10 13 cm −3 . The correlation between the C 2 radical density and the quality of diamond films was investigated.

K-1709 誘導結合CF_4プラズマの粒子モデル解析(J08-3 熱流体現象の原子・分子論的アプローチ(3))(J08 熱流体現象の原子・分子論的アプローチ)

For our future analysis of radicals production in inductively-coupled CF_4 plasmas, we analyzed the electromagnetic field in an inductively-coupled plasma reactor. We solved Maxwell's equations by use of ADI method considering the rf antenna current and plasma current. The computational conditions were chosen to be equal to the experimental ones at the industry.

Fluorocarbon polymer deposition kinetics in a low-pressure, high-density, inductively coupled plasma reactor

Maintaining dimensional control and adequate throughput during the etching of submicron features requires plasma etch tools that operate at low pressures and high densities, such as inductively coupled plasmas (ICPs). Unfortunately, in this regime, it has proven difficult to achieve a stable, reproducible chemistry for selective oxide etching of contacts and vias. In particular, it is difficult to control the passivating polymer film which provides etching selectivity to silicon, nitride, and photoresist. As a first step toward sorting out the complicated oxide etching chemistry, we have measured and modeled the kinetics of the polymer film deposition in an ICP reactor for C2F6/H2 and CHF3 chemistries. Using a unique application of statistical design of experiments, we have explored the pressure range of 3–15 mTorr, power range of 300–2000 W, residence times from 0.5 to 1.0 s, and magnetic field from 0 to 24 G. Polymer deposition rates on a bare Si wafer are measured using a laser interferometer. The concentration of fluorocarbon radicals, CF, CF2, and CF3, are measured in the plasma using wavelength modulated infrared diode laser absorption spectroscopy. Additional measurements include actinometric F atom density and ion saturation current. These measurements are analyzed in terms of a polymer deposition model and the important physical phenomena are inferred. Significantly, we find a unique polymer deposition mechanism over the entire range of tool parameters including direct deposition of CF and ion-assisted deposition of CF2.

A computational investigation of the effects of varying discharge geometry for an inductively coupled plasma

In this paper, a numerical investigation of ionization and density profiles in a low pressure inductively coupled plasma (ICP) discharge is presented. ICP reactors have in recent years become a popular choice for semiconductor processing. To ensure high yields, it is important to optimize process uniformity. Off-axis ionization can result in the hollowing of the density profile of the plasma. Previous numerical studies [Keiter (1996), Beale (1995), Mumken (1999), Kortshagen (1999)] have focused on electron kinetics and the resulting ionization rates as the main factor in producing nonuniform ion density profiles. Here we extend some of the previous work to also examine the effects of the ion transport in detail. Two models are used for the study. The first model is a hybrid model that approximates electron transport by solving the Boltzmann transport equation (BTE) using a spatially dependent two-term spherical harmonic expansion (SHE), and neutral and ion transport with a simple fluid model. The second model is a new kinetic model for ion transport in complex geometry based on the convected scheme (CS). This model uses an "unstructured mesh" and "Long Lived Moving Cells". By using the CS model for ion transport, it is possible to investigate the limitations of the fluid model and issues pertinent to process uniformity which are not accessible to the fluid approximation. As the ratio between ion mean-free path and system size is often looked to as the key parameter for determining the applicability of a fluid model, the two parameters we have varied in this study are the reactor chamber height and the neutral temperature. Ion density profiles produced by the fluid model are qualitatively similar to those of the kinetic ion model, although the kinetic model tends to produce more uniform ion density profiles.

Abatement of perfluorocarbons with an inductively coupled plasma reactor

Preliminary tests using an inductively coupled plasma (ICP) reactor to abate C2F6, CF4, and CHF3 show promising results. In conjunction with this, some of the operating parameters, which need to be considered when trying to obtain maximum abatement efficiency, are explored. Near 100% destruction efficiency (DRE) for the persistently fluorinated compounds (PFCs) mentioned earlier has been achieved with O2 as the additive gas. Similarly, the abatement of CF4 with CH4 and/or CH4/O2 as the additive gas, has shown comparable results. For the non-CF4 PFCs, like C2F6 and CHF3, the formation of CF4 as an abatement by-product was monitored and shown that it can be controlled by varying the operating parameters of the ICP reactor. The trends in the DRE observed while manipulating the ICP operating parameters such as ICP power, residence time, amount/type of additive gas, and Ar dilution, show that there are trade offs which must be taken into account. A quadrupole mass spectrometer was used for analyzing the abatement effluents.