Time dependence of etch-induced damage generated by an electron cyclotron resonance source

Abstract

Cl 2 plasma passivation of etch induced damage in GaAs and InGaAs with an inductively coupled plasma sourceThe effects of etch time on the damage induced during dry etching have been studied. GaAs-and InP-based materials were etched for different times in an electron cyclotron resonance source and the electrical and optical characteristics were measured. Variations in Schottky diode ideality factor (n) and barrier height ( b ) were used to measure dry-etch-induced electrical damage at the surface of GaAs after different etch times. The contact resistance of In 0.53 Ga 0.47 As, extracted from transmission line measurements, was also investigated for different etch times and compared to the etch time dependence of GaAs. Capacitance-voltage ͑C -V͒ measurements from Schottky diodes and photoluminescence ͑PL͒ intensity from GaAs/Al 0.30 Ga 0.70 As multiple quantum well ͑MQW͒ structures were used to investigate the time dependence of the etch damage farther below the surface. The damage in GaAs as shown by n and b shows marked deterioration after 10 s of etching but improves after longer etching. The contact resistance of In 0.53 Ga 0.47 As was found to increase fastest in the first 10 s, but the electrical characteristics did not improve with etch time as the GaAs did. There was no etch time dependence in the C -V measurements from the diodes or the PL spectra from the MQWs, indicating that variations in dry-etch-induced damage with etch time are mostly confined to the surface. Using wet chemical etching, the damage profile was investigated and the most damage was found within ϳ15 nm from the surface for GaAs while In 0.53 Ga 0.47 As did not return to the original value after removing 20 nm of the surface. This indicates that the damage generation, removal, diffusion, and channeling mechanisms or rates may be different for the two materials. By changing the temperature of the stage during etching from Ϫ130 to 350°C, the etch time dependence of the n and b of GaAs was affected, indicating that diffusion of defects may be mainly responsible for the variations in electrical characteristics with etch time.

Low‐Pressure Etching of Nanostructures and Via Holes Using an Inductively Coupled Plasma System

“…In addition, the faster etch rate helps to remove the generated defects. 8 Similar improvements in n and breakdown voltage were observed as the ICP source power was increased. For the diodes etched at 25 W stage power, the etch rate increased from 45 to 261 nm/min as the source power was increased from 50 to 500 W and the |V dc | decreased from 60 to 42 V. The barrier height did not show much variation and was 0.75 eV for source powers of 50 and 500 W. This indicates that the lower |V dc | obtained with high source power and low stage power are more desirable for high-performance devices because lower damage is induced.…”

“…The introduction of the Ar into the plasma increased the physical component of the etch, which has been shown to increase damage induced in the material. 8 The control diode was wet etched in a solution of H 2 O:NH 4 OH:H 2 O 2 at 130:3:1 ratio with an etch rate of 300 nm/min. To fabricate the HDBR structures in GaAs, the samples were patterned using electron-beam lithography with 200 nm thick 950 K polymethylmethacrylate (PMMA), which consisted of 4% solid by weight cast in anisole.…”

“…Etch-induced damage studies have been done in ECR plasma systems by using quantum wires for sidewall damage studies 6 and Schottky diodes for surface damage studies. [6][7][8] Previous studies have shown that the sidewall damage depth in GaAs from dry etching can be reduced by using lower stage powers and 40 nm wide quantum wires with etch depths of 1.3 m have been demonstrated. 6 The effects of dry etching on IPG transistors showed a decrease in drain current with decreased stage power, possibly due to lower leakage along the sidewalls from etch-induced damage.…”

The influence of post-etch InGaAs fin profile on electrical performance

“…[19][20][21][22] While the dry etch approach provides advantages such as process flexibility and simplicity, 23) it may also introduce complications at the etched surface, including stoichiometric or structure modification and atomic intrusion. 24,25) As a result, the electronic properties of InGaAs post etch could be modified. Mole fraction modification of In, for example, has been shown to vary the InGaAs charge neutrality level affecting device parameters such as the subthreshold swing.…”