RF-excited Plasma Research Articles

Fiber optic sensor for substrate temperature monitoring

The use of an optical fiber Fabry–Perot interferometric temperature sensor for monitoring substrate temperature in a rf sputtering system is demonstrated. The sensor head consists of a continuous length of single mode silica fiber which contains two internal mirrors a distance of 1.08 cm apart to form the interferometer cavity. The laser light source for the sensor is located outside the vacuum system and connected to the sensor head via a fiber optic feedthrough. The accuracy of the sensor was 0.05 °C in this experiment, but considerable improvement is possible with better signal averaging techniques. The fiber optic sensor provided substrate temperature information during rf plasma excitation, when a nearby thermocouple ceased to function due to electromagnetic interference.

Modification of Y2Ba4Cu8O16 thin-film surfaces by interaction with a radio frequency excited oxygen plasma

The use of rf excited oxygen plasma for oxidizing and possible etching of SrTiO3 (100) supported Y2Ba4Cu8O16 thin films is studied by photoemission. Oxygen plasma modifies the surface and eliminates photoemission features identified with surface impurities like BaCO3. The resulting surface is barium deficient but shows a Fermi level suggesting it is metallic. Application of a −200 V dc bias to the sample to increase the kinetic energy of the incident oxygen ions enhances either surface sputtering or atomic mixing and produces a surface which resembles YBa2Cu3O7, because the photoemission spectra are similar to those published for cleaved crystal surfaces or films.

Influence of an Insulating Film on Plasma Silicon Dioxide Deposition Rates

The dependence of plasma silicon dioxide deposition rates on the thickness of underlying electrically insulating thermal oxide grown on sample wafers prior to plasma deposition has been studied. Insulating films can reduce the deposition rates by approximately 25% if both wafers facing each other in the susceptor assembly are insulated from it, and by approximately 15% if only one is insulated. A semiquantitative description of the effect is given, based on variations of RF plasma excitation power due to insulating target wafers from the susceptor.

Direct formation of ferrite films in wet process

We have prepared a polycrystalline Co-ferrite film by the electroless ferrite-plating technique of our invention [J. Appl. Phys. 55, 2614 (1984)] in an aqueous solution at 70 °C on substrates of organic compounds (polyethylene terephthalate, polymethyl methacrylate, polycarbonate, Teflon), a stainless steel, and a polyester fiber. We used no intermediate layer, which we needed previously to enhance the adhesion of the ferrite film to the surface. By exposing the organic substrates to an rf-excited air plasma, the wettability of the surface improved, which enhanced the adhesion of the film to the substrate. Here we have measured the adhesion by a cross-cut test. The plasma forms hydrophilic groups such as -OH and -COOH, which improve the wettability. The OH group adsorbs the metal ions in the reaction solution, initiating the ferrite film formation. The glass shows a strong adhesion to the film even when it is not exposed to the plasma, because the glass has the OH group on the surface inherently. The stainless steel does not enhance the adhesion by the plasma treatment because the plasma forms no OH group on the metal surface. The adhesion power higher than ∼10 kg/cm2 has been obtained between the Co-ferrite film and the glass (not plasma treated), the polyethylene terephthalate (plasma treated) and the Teflon (plasma treated).

Nature and distribution of radicals in rf and dc silane discharges; effects on deposition rate and physical properties of a-Si:H

Anodic and cathodic processes for rf and dc excited plasmas of silane gas differ in the relative role of charged and neutral radical species. By studying the thickness distribution of films deposited on substrates immersed in the plasma, the relative importance of these species as contributors to the film thickness is assessed. In addition, the impact of the different deposition processes on the spin density in the deposited hydrogenated amorphous silicon films and a key photoelectronic property, the hole range, is discussed.

Material Processing with Broad-Beam Ion Sources

During the past two decades, ionized gases have been increasingly applied to etching, modification, and deposition of materials in thin film form (1), often using radio frequency (rf) excited plasmas. In parallel, the technology of broad-beam, multi aperture ion sources has been extensively developed for space propulsion (2). Only in the past few years have these broad-beam ion sources been applied in a systematic fashion to materials problems. The purpose of this review is to describe the growth ofthis new area of ion beam processing of materials. To define the field more precisely, the ion beams referred to here are typically several cm to tens of cm in diameter, not mass­ separated, with ion energies of tens of eV to a few keY. Not included are finely focused ion beams for surface analytical studies; ion implantation and ion-mixing beams; ion beams for fusion injection; and single-aperture beams. The primary advantage of ion beam processing of materials lies in the control available with an ion beam as compared to other plasma processes. The ion flux and ion energy are easily measured and independently controlled, and the direction of ion impact on surfaces is controlled, since the beams operate in a low background gas pressure of typically 0.01 Pa. This review begins with a description of broad-beam multiaperture ion sources, emphasizing the design advances that have stimulated recent

X-ray photoelectron spectroscopy study of hydrogen plasma interactions with a tin oxide surface

Tin oxide thin films were exposed to an rf excited hydrogen plasma (100 mTorr) for various times. The plasma system was operated in a self-biased mode. X-ray photoelectron spectroscopy analysis of vacuum transferred samples shows that SnO2 is reduced. A thin self-limiting layer of elemental Sn forms on the reduced film to a thickness of ∼10 Å.

Oxide growth in an rf plasma

A mechanism is considered for oxide growth on an electrically isolated metal sample in the presence of an rf-excited oxygen plasma. The assumption is made that the rate of rf oxidation is limited by transport of ionic species through the already-formed oxide layer. Thermally activated hopping of ionic defects in the presence of electric fields due to the surface potential established by the rf discharge and modified by the space charge of the mobile ionic defects is hypothesized. The origin of the voltage across the oxide is discussed in terms of a balance between the transport of negatively charged O ions and transport of electron holes created by ion neutralization of positive ions from the plasma. This model is developed analytically and evaluated numerically by employing the continuum limit of hopping transport equations valid for the very-high-field limit. A three parameter fit gives excellent agreement of the theory with the published data of Greiner for the rf oxidation of lead. The fitting parameters are the voltage across the oxide film, the ionic diffusivity, and the rate at which the oxide is removed by sputtering. The resulting parameter values are shown to be in reasonable accord with available experimental data. Parametric curves are presented to illustrate the dependence of the kinetics on the various microscopic parameters. The dependence of the limiting thickness on the primary parameters is likewise presented. It is shown that the growth rate may be significantly retarded by high levels of space charge in the oxide.

The electrical characteristics of rf magnetron and non-magnetron planar systems

The electrical equivalent circuit for an rf excited plasma is presented. The plasma parameters, viz. conductance and capacitance are determined for an rf planar magnetron system. The measurements of the plasma parameters for non-magnetron planar systems are repeated. The results confirm the calculation presented in an earlier paper 1 though the geometry of the target assembly is different in this case.

Experimental and design information for calculating impedance matching networks for use in rf sputtering and plasma chemistry

To calculate impedance matching networks for capacitively coupled rf excited plasmas (as used for sputtering and some type of chemical reactors) it is necessary to know the values of the plasma parameters, i.e. the capacitance and the conductance of the discharge and the variation of these with the operating conditions of the system, i.e. the pressure, the frequency and the input power. The paper presented here gives information on the foregoing and shows how the results can be used to design suitable matching networks. To provide basic information on the glow discharge conditions experiments have been carried out using an rf power supply with one side of the matching network connected via a blocking capacitor to an electrode and the remaining side grounded. This system has been operated under various conditions, i.e. the frequency has been varied between 2 and 14 MHz, the air pressure from 1 × 10 −1 down to 5 × 10 −3 torr and the input power density from 0.5 to 2 W cm −2. The results obtained for 14 MHz are in close agreement with those reported for 13.56 MHz by Logan et al 1. From the values obtained for the plasma parameters different types of matching network (L, π, T> and tuned transformer) have been calculated. The design equations for these networks are presented with calculated and tested examples to demonstrate their application. The procedure used, although common electrical engineering practice, should with the measured data be of value to those undertaking rf system design without experience of matching conditions.

Plasma deposition of GaP and GaN

We report the preparation of thin films of gallium phosphide and gallium nitride by deposition from low-pressure rf-excited plasmas in mixtures of trimethylgallium with ammonia and phosphine, respectively. With the deposition conditions used, the gallium phosphide is found to be amorphous, while the gallium nitride is polycrystalline. We present preliminary measurements of optical and electrical properties.