The true surface temperature of a silicon wafer and the related etch rate in a CF4 plasma

Abstract

By means of a resistance thermometer structure diffused into the surface of a silicon wafer the temperature has been found to increase within a minute to over 100°C in a CF4 plasma. In contrast the system temperature rises only about 5 °C in the same time. This applies to a system in the tunnel configuration. In a parallel plate system the rise is far less steep and even within the usual etching time an equilibrium temperature is reached. The dependence of the etch rate on the surface temperature has been measured. In the interval from R.T. and 100 °C the etch rate stays nearly the same. Therefore temperature is no significant parameter for establishing desired etch rates

The Temperature Dependence of the Etch Rates of GaAs , AlGaAs , InP , and Masking Materials in a Boron Trichloride:Chlorine Plasma

“…The latter is due to the hostile plasma environment for which a metal thermocouple is poorly suited. However, as has been demonstrated repeatedly by several authors over many years (18)(19)(20)(21), the temperature rise of a substrate can be well over 100~ during plasma etching. For etching III-V compounds, some of which have relatively involatile etch products, temperature can play an extremely important role, for InP temperature plays the dominant role in determining how quickly this material etches and whether it will etch at all.…”

“…Manuscript submitted March 27, 1987; revised manuscript received Nov. 23, 1987. This was Paper 848 RNP presented at the San Diego, CA Meeting of the Society, Oct. [19][20][21][22][23][24] 1986.…”

The Effect of Temperature and Flow Rate on Aluminum Etch Rates in RF Plasmas

“…Several workers have developed techniques for accurately measuring sample temperature in a plasma environment, and their findings show that temperature rises of 100~ are not uncommon (2)(3)(4)(5)(6). It appears that hearings due to energetic particle bombardment and to exothermic surface reactions are both significant (2,3).…”

“…Several workers have developed techniques for accurately measuring sample temperature in a plasma environment, and their findings show that temperature rises of 100~ are not uncommon (2)(3)(4)(5)(6). It appears that hearings due to energetic particle bombardment and to exothermic surface reactions are both significant (2,3). Also, it has been proposed that the ac magnetic field present in an RF discharge induces a substantial eddy current within highly conductive materials, which in turn causes sample heating through ohmic loss (7).…”