Average Etch Rate Research Articles

Electron cyclotron resonance plasma reactor for SiO2 etching: Process diagnostics, end-point detection, and surface characterization

An electron cyclotron resonance plasma reactor for low-pressure etching of SiO2 layers on Si is described. Under typical operating conditions of 1.1 mTorr neutral pressure, 5 sccm CF4/46 sccm He, and 400 W net microwave power, an average etch rate of 35–40 nm/min is obtained. Ion densities were measured by using a quadrupole mass analyzer (QMA) immersed in the downstream plasma; an optical multichannel analyzer was used for actinometric measurements of atomic F and O concentrations. The results are correlated with process variables and SiO2 etch rates. Real-time end-point detection was effected by actinometry and by monitoring the concentrations of etch products using a differentially pumped QMA. H2 addition inhibits SiO2 etching but improves SiO2/Si etch selectivity. For selective etching, in situ, off-line surface analysis by Auger electron spectroscopy demonstrated an increase in the surface C concentration during SiO2 etching and deposition of a fluorocarbon layer at the end point which strongly inhibits Si etching.

Low energy electron-enhanced etching of Si(100) in hydrogen/helium direct-current plasma

Low energy electron-enhanced etching of Si(100) has been achieved by placing the sample on the anode of a dc discharge in hydrogen/helium mixtures. Over a broad range of gas composition, gas pressure, and discharge current, nonpatterned samples gave etch yields of 0.01–0.02 atoms/electron, and average etch rates of 2000–3000 Å/min. Postetch examination by atomic force microscopy revealed surface roughness of 2–3 nm. These results are related to incident flux of H atoms and electrons through a simple model of the anode sheath layer above the sample.

Investigation of modulated radio frequency plasma etching of GaAs using Langmuir probes

Radio frequency ClCH3/H2 plasma etching of GaAs is examined in the 10–140 mTorr pressure range using square-wave modulation of the excitation source to control the etching. A Langmuir probe is used to measure time-resolved electron density, characteristic temperature, and floating potential during the plasma afterglow period. The ClCH3/H2 plasma electron energy is found to be 1.2±0.3 eV. The near afterglow plasma density decay has a time constant in the order of τ=30 μs at 140 mTorr for 10%–20% ClCH3 in H2 and τ=100 μs for H2. The floating potential continues to decay into the far afterglow, with a characteristic time of the order of milliseconds. The Langmuir probe measurements indicate that in ClCH3 plasmas the near afterglow is dominated by electron attachment, whereas the far afterglow is dominated by ambipolar diffusion. The GaAs etch rate experiments show that surface reactions continue into the far afterglow, dominating the behavior of the time average etch rate.

Anisotropic Chemical Pattern Etching of Copper Foil: III . Mathematical Model

A mathematical model for wet chemical etching of copper in acid chloride solutions was developed for exploring the phenomenon of anisotropic etching reported in Parts I and II of this study. The model considered a two‐dimensional rectangular cavity and accounted for laminar convection, diffusion, migration, homogeneous equilibria, and heterogeneous electrochemical reactions involving eight ionic species. Two compositions and two cavity aspect ratios were considered. Calculation of the flow field demonstrated the existence of recirculating eddies which play a significant role on local transport, particularly in the interior corners. Calculation of the concentration contours of the reaction product, , indicated that the solution was supersaturated for 1:1 cavities as well as for 5:1 cavities in the case of the solution. For 5:1 cavities in solution, supersaturation occurred only in the corner regions for . The predicted local etch rate was nonuniform owing to the presence of recirculating flow eddies. The predicted average etch rate for the two aspect ratios was compared with experimental data over the range and was found to agree to within 10 to 15% for the 0.5M solution, and to within 15 to 30% for the 3.5M solution.

Feature‐Size Dependence of Etch Rate in Reactive Ion Etching

Reactive ion etching (RIE) of silicon trenches has been studied as a function of trench width, ranging from 0.4 to 2.5 μm, in a plasma. The average etch rate inside a Si trench becomes slower as the trench width decreases. This etching phenomenon is called RIE lag. Also, we have measured the ion saturation current under a negative dc bias voltage with a miniature Faraday cup inside a cylindrical hole. The loss of the ion current occurs mostly within a depth equal to the hole diameter, independent of the hole size, and the current loss is directly proportional to both pressure and RF self‐bias voltage at the substrate electrode. The lighter the mass of the ions, the more ions are deflected to the sidewall. Theoretical simulation shows that the electric field is perturbed primarily by the shape of the trench, and that the space charge near the trench plays an important role in ion deflection. Since the field perturbation is greater at the edge of the trench, the ion loss becomes more severe in a narrower trench, leading to the RIE lag. The sidewall profile observed in Si trenches shows a consistent picture of the field effect, particularly at low pressure where ion‐neutral collisions become insignificant.

Laser‐Assisted Etching of Manganese‐Zinc‐Ferrite

A focused argon ion laser beam (514.5 nm) has been used to induce etching of manganese‐zinc‐ferrite ceramics in aqueous solutions. In the absence of the laser beam the etching process did not take place even at the boiling point of the selected etchant. For concentration above 2N, bubble‐free etching took place in the presence of the laser beam. Within the range of 2–8N for concentration, the depth of the etched grooves increased with increasing laser power, decreasing scan speed, and increasing concentration. Evidence of decreasing average etch rate due to mass transport limitation was observed under low concentration, high laser power, and low scan speed conditions. The width of the etched grooves increased with increasing laser power, but was relatively insensitive to the scan speed and the concentration. The experimental findings of the laser‐assisted etching of Mn‐Zn‐ferrite can be qualitatively interpreted by a thermally activated reaction mechanism: formation at and above a threshold temperature of an active chemical environment on the surface for the initiation of the etching process; more concentrated etchant allows more abundant supply of the reactive species and requires less heat input to reach the active state. Etching of ferrite occurs as a result of the reaction between the active etchant (possibly fused ) and either the solid or the molten ferrite, depending on the laser power. The etching process generates soluble product species and the profiles of the etched patterns are strongly influenced by the mass transport characteristics of the etching system. A mathematical model for the surface temperature distribution, which takes into account heat loss into the liquid, is also presented.

The micro-environment of dislocations in undoped, semi-insulating GaAs

The environment of dislocations in undoped, semi-insulating GaAs has been studied by modified A B etch. The residual impurity distribution is clearly delineated by this etchant, resulting in regions of lower than average etch rate around dislocations.

Laser-Assisted Liquid-Phase Etching of Copper Conductors and its Application to Advanced Integrated Circuit Interconnect

ABSTRACTWe have demonstrated a method f or the selective-area, laser-assisted liquid-phase etching of 6-μm thick copper conductors, by focusing the 488-μm line of an argon ion laser onto the surface of samples immersed in a dilute solution containing sulfuric acid and hydrogen peroxide. Average etch rates of up to 5.0 μm/s have been achieved, in the process of completely severing 14-μm wide copper lines, with little attendant damage to the underlying layer of polyimide. Two etchant formulations were identified which exhibited large etch rates at elevated temperatures and low enough background etch rates (approximately 1 μm/hr at 0 °C) to be useful in practical applications.

Laser-induced chemical etching of ceramic PbTi1?xZrxO3

The laser-induced etching of ceramic PbTi1−xZrxO3 in a hydrogen atmosphere and in air has been investigated. Visible Ar+ and Kr+ laser radiation was employed in most of the experiments. In H2 atmosphere, regular patterning of the ceramic is possible. Average etch rates reach up to about 250 μm/s.

The effect of gamma dose on the alpha response of CR-39

Abstract The effect of gamma irradiation on the alpha-particle track response of CR-39 is investigated. The effects are dramatic above doses of 500Krad. The alpha-particle tracks in highly dosed plastic exhibit a ‘threshold’ residual range below which the track etching appears to saturate. The effects can be simulated by the use of a suitable VT(R) vs range curve and is therefore not a ‘skin’ effect. The effects appear to be due to the addition of the two doses, the effects being present for alpha dose + gamma dose and gamma dose + alpha dose. The average track etch rates VT/VB, remain constant up to 500Krad but then exhibit a fall-off at higher doses. The observations have relevance for neutron spectrometry and dosimetry measurements where large gamma fields may be present. The bulk etching behaviour is also investigated. The change in bulk etching rate is too small to be of any value in a gamma dosemeter for doses below 100Krad.