S. Thomas et al., MINIMIZED RESPONSE-TIME OF OPTICAL-EMISSION AND MASS-SPECTROMETRIC SIGNALS FOR OPTIMIZED END-POINT DETECTION, Journal of vacuum science & technology. B, Microelectronics and nanometer structures processing, measurement and phenomena, 14(4), 1996, pp. 2531-2536
The response times of optical emission spectroscopy (OES) and mass spe
ctrometry (MS) have been measured for plasma etching of III-V heterost
ructures. For the Ga optical emission signal at 417.2 nm, a response t
ime as fast as 0.2 s was obtained. The minimum response time of the (A
sCl2+)-As-145 partial pressure, measured by MS, was found to be 0.9 s.
The saturation times of the optical emission signal and the partial p
ressure signal have been shown to be related to the residence time of
etch gases by varying the total gas flow rate and the chamber pressure
. Decreasing residence time by reducing the pressure from 6 to 2 mTorr
and maintaining a constant flow rate caused the saturation time of th
e Ga emission signal at 417.2 nm to decrease from 7 to 3 min. The (AsC
l2+)-As-145 partial pressure signal saturated before the Ga emission s
ignal. Endpoint detection for etching an AlInAs emitter and stopping o
n a GaInAs base of a heterojunction bipolar transistor was studied. Al
gorithms which monitor the change in Ga emission intensity have been d
eveloped to automatically stop the emitter etch with similar to 2 nm o
f the GaInAs base layer removed. Additionally, etching of GaInAs on an
InP substrate was studied and the signal from OES detected the endpoi
nt before the MS signal did due to the faster response time. (C) 1996
American Vacuum Society.