AVS 45th International Symposium
    Manufacturing Science and Technology Group Wednesday Sessions
       Session MS-WeM

Paper MS-WeM6
Evolution Effects of Reactor Inner Wall Surface on Fluorocarbon Plasma Parameters

Wednesday, November 4, 1998, 10:00 am, Room 317

Session: Advanced Process Equipment and ES&H
Presenter: T. Ichiki, Toyo University, Japan
Authors: H. Oshio, Toyo University, Japan
M. Ogata, Toyo University, Japan
T. Ichiki, Toyo University, Japan
Y. Horiike, Toyo University, Japan
Correspondent: Click to Email
SiO@sub 2@ contact hole etching using fluorocarbon plasmas cause various issues such as etch stop, relating closely to the change of plasma state due to the deposition on the reactor wall. To clarify wall effects, variations of CFx(x=1-3) radical densities, pressure and wall temperatures with the C@sub 4@F@sub 8@ discharge time were investigated. Inductively coupled plasma was generated by 13.56 MHz power supplied to an antenna wound around a 130 mm @phi@ quartz bell jar connected to a 150 mm @phi@ SUS reactor. A 130 mm @phi@ Cu barrel, whose temperature was controlled by water was inserted into the reactor. An orifice head of the appearance mass spectroscope(AMS) was set at 15 cm apart from the antenna. Etching experiments were carried out on a stage equipped with a load-locked system. At first the reactor was cleaned by ashing, then pressure evolution with discharge time was measured for different residence times,@tau@. At @tau@=100 msec the pressure once dropped, and then gradually increased after 3 min. AMS revealed the pressure increase was mainly attributed to increase in CF@sub 3@ radical. The initial decrease resulted from singnificant adsorption of radicals on the cleaned wall. The run-to-run variations were measured by repeating a cycle of 3 min discharge on and 3 min off for 20 times. The wall temperature rise reached its steady state at 50 - 80 ° after the 5th run, while the pressure kept on increasing even over the 20th run. SiO@sub 2@ and Si etch rates as well as plasma potentials decreased corresponding to these increases. In contrast, at @tau@=30 msec run-to-run variation slightly occured after the 3rd run and detailed measurements of pressure evolution with time revealed no initial drop but the rapid increase after plasma ignition, caused by the high deposition rate due to the high radical density. Furthermore, water cooling of the bell jar suppressed the increase in pressures. Consequently to adopt short residence time and control the wall temperature is important for keeping CF@sub 3@ radical density constant to achieve good reproducibility.