Plasma Chemistry Dependent ECR Etching of GaN
- PDF / 1,031,456 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 23 Downloads / 219 Views
device structures including lasers and heterojunction bipolar transistors (HBTs) has increased the need for anisotropic, smooth, high-rate etching. A variety of plasma etch chemistries have been reported in a reactive ion etch (RIE) system using chlorineand bromine- based plasma chemistries. Etch rates for GaN up to approximately 650 A/main have been reported at dc biases of -400 V. Significantly higher etch rates have been reported in electron cyclotron resonance (ECR) etch systems where the plasma is confined by a magnetic field to provide a high density plasma at low pressure and low ion energies. Most ECR etching of GaN has been performed using Cl2/H 2based plasmas.' 3 ' 5 Highly anisotropic GaN etching was obtained at dc-biases ranging from -150 to -250 V with etch rates exceeding 2800 A/min. In this paper, we report ECR etching of GaN as a function of plasma chemistry using CI2/H 2/Ar, CI 2/SF 6/Ar, BCI 3/H2/Ar, and BCI 3/SF6/Ar plasmas. These discharge chemistries are expected to etch GaN due to the high volatility of the Ga-chlorides and the formation of volatile NHx or NF, etch products with the addition of H 2 or SF 6 to the plasma. EXPERIMENT The GaN films were grown using Metal Organic Molecular Beam Epitaxy (MO-MBE) on GaAs substrates in an Intevac Gen II system described previously. The group-III source was triethylgallium and the atomic nitrogen was derived from an ECR Wavemat source operating at 200 W forward power. The layers were single crystal with a high density of stacking faults and microtwins and were resistive as-grown. 751 Mat. Res. Soc. Symp. Proc. Vol. 395 ©1996 Materials Research Society
The GaN samples were patterned using a photoresist mask. The ECR plasma reactor used in this study was a load-locked Plasma-Therm SLR 770 etch system with an ECR source operating at 2.45 GHz. Energetic ion bombardment was provided by superimposing an rf-bias (13.56 MHz) on the sample. Samples were mounted using vacuum grease on an anodized Al carrier that was clamped to the cathode and cooled with He gas. Etch gases were introduced through an annular ring into the chamber just below the quartz window. To minimize field divergence and to optimize plasma uniformity and ion density across the chamber, an external secondary collimating magnet was located on the same plane as the sample and a series of external permanent rare-earth magnets were located between the microwave cavity and the sample. ECR etch parameters held constant in this study were: 30'C electrode temperature, 1 mTorr total pressure, 30 sccm total gas flow, 5 sccm of Ar, 850 W of applied microwave power, and 150 W rf-power with a corresponding dcbias of -150 ± 10 V. Etch rates were calculated from the depth of etched features measured with a Dektak stylus profilometer after removing the photoresist mask. Samples etched in the ECR were approximately 1 cm and depth measurements were taken at a minimum of three positions. Error bars for the etch
rates were ±10% across the sample. Limited sample supply precluded multiple runs at each condition
Data Loading...
