Assessment of Surface Damage of Gallium Arsenide due to Reactive Ion Etching
- PDF / 323,981 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 47 Downloads / 252 Views
ASSESSMENT OF SURFACE DAMAGE OF GALLIUM ARSENIDE DUE TO REACTIVE ION ETCHING M.S.Puttock , H.Thomas , D.V.Morgan*, U.Rossow', D.R.T.Zahn*, W.Richter', I.P.Hilton , J.Woodward£. University of Wales College of Cardiff, Electrical, Electronic and Systems Engineering, UýCC, Cardiff, Wales. *Institute of Solid State Physics, Berlin West Germany. Royal Signals and Radar Establishment, Malvern, Worcs, UK. ABSTRACT Crystal damage of GaAs(100) caused by Reactive Ion Etching (RIE) using a mixture of C12 and Ar gas has been assessed using Surface Roughness (Ra), Resonant Raman Spectroscopy (RRS), Schottky diodes, and Spectroscopic Ellipsometry (SE). Plasma conditions for minimum induced damage have been determined and compared to optimised RIE processes using plasma gases SiC1 4 , CH4 -H 2 , CCl 2 F2 and Ar. The SiC1 4 plasma was found to produce the least crystal damage. 1 INTRODUCTION RIE of GaAs is becoming increasingly important for device fabrication [1] because it offers advantages compared with wet etching in terms of the control of sidewall profile, etch rate and uniformity. However, RIE can induce appreciable radiation damage to the GaAs surface [2]. Previous reports have shown that the degree of induced damage is detectable using techniques such as surface roughness, Raman Spectroscopy [3], and the electrical characteristics of Schottky diodes made on the etched surface [2]. It is important to assess the extent of damage caused to GaAs so that RIE processes can be optimised. To date, no comparisons have been made of the results yielded by each technique and no detailed analysis made of how the degree of damage varies with plasma parameters such as rf power, gas composition, and gas pressure. A C1 2 -Ar gas mixture has been investigated because it is potentially of use for device fabrication [4]. This gas chemistry is unlikely to produce polymers and should be a simple system in which one can move from a physical to chemical etch process by varying the gas composition. Etching has also been investigated using CC1 2 F 2 [5], SiCl 4 , and CH4 -H2 [6] gases. These were established processes already used for device fabrication and served as a comparison with the C12 -Ar system. 2 EXPERIMENTAL 2.1 Etching
Etching was performed in a Plasma Technology RIE80 described elsewhere [5], which featured a Plass chamber and a 13.56MHz rf discharge, with electrode area -f 227cm . A turbomolecular pump was used to give a base pressure of 10 Torr. A flow of 15sccm was used for all experiments, the pressure being set by throttling the vacuum line. The process gases, Cl 2 and Ar were mixed by flow ratio. Semi-insulating GaAs(100) masked with AZ1350 photoresist, was placed on a Si cover plate and etched for three minutes, yielding etch depths ranging from 10nm to 3.5p. A selected condition gave an average etch depth of 0.93p with a standard deviation of 0.08 over five successive etches. This repeatability was maintained over two months after which time, chamber and pump cleaning was required. An H2 0 2 /NH3 wet etch was used to provide
Data Loading...
