The Formation of SiO2 in an RF Generated Oxygen Plasma: II . The Pressure Range Above 10 mTorr

Abstract

This paper describes the formation of SiO~ in an rf generated oxygen plasma in the pressure range above 10 mTorr. The experimental setup used for this study is the same as that described in Part I of this study. Using the silicon-mask interface as a marker and examining the cross section of silicon wafers after oxide formation, it was concluded that the mechanism of oxide formation in the pressure range 10-100 mTorr is growth of SiO2. A unique and unexpected result of the plasma process conducted at pressures greater than 10 mTorr is that oxide growth occurs primarily on the surface of the wafer facing away from the plasma, in sharp distinction to the observations made at pressures less than 10 mTorr where the oxide forms only on the surface facing the plasma. A key experimental parameter is the confinement of the plasma by the wafer which was observed only at pressures greater than 10 mTorr and at generator frequencies from 0.5 to 3 MHz. Oxide growth rate was studied as a function of pressure, power, frequency of the a~plied field, and substrate parameters to determine the optimum growth conditions. The etch rate, refractive index, stress, fixed charge, interface states, and breakdown strength of plasma oxides grown at 500~ compare very favorably to thermal oxides grown at ll00~ * Electrochemical Society Active Member.

An In-Situ Spectroscopic Ellipsometry Study of the Electron Cyclotron Resonance Plasma Oxidation of Silicon and Interfacial

“…The system with SE characteristics and alignment procedures were previously described' 11 and applied to ion beam studies. The plasma conditions were: 300W, oxygen pressure of 5x10 4 torr which resulted in a sample temperature during oxidation of about 80"C. The results that follow were made with the sample normal to the plasma and with sample biases of OV, +30V and -30V.…”

Plasma Assisted Oxidation of Si at Temperatures below 800°C

“…2 shows the oxide thickness, d , as a function of time, t, plotted as log dx vs. lo t. The oxide thickness is seen to increase as t0 for the initial 110 nm, and then to increase as t 1 . Earlier workers [6,7] showed a change from a linear to parabolic growth kinetics with increasing oxide thickness. They speculated that the mechanism for oxidation in this type system could be due to a built-in potential on the wafer and a charged oxidizing species.…”

“…The diffusion constant in equation I increases as a function of temperature. In addition to the problem just mentioned, high temperature oxidation also causes surface depletion of impurities [2], oxidation enhanced diffusion [3) oxidation induced stacking faults [3,41 and so-called bird's beaking [5] Plasma assisted oxidation offers the potential for oxide formation at temperatures below 800C free from the problems described above, with the possible additional benefits of being orientation and resistivity, and resistivity type independent [6,7].…”

Microwave‐Discharge Plasma Oxidation of Silicon in a Cusp Magnetic Field

“…Plasma oxidation or anodization proceeds in oxygen plasma generated by dc (1), RF (2,3), and microwave discharge (4)(5)(6). Therefore, the oxidation kinetics are different from those of thermal oxidation in the furnace.…”

“…Therefore, oxidation does not proceed as fast as in plasma anodization. However, the oxidation is also enhanced below 600~ at which thermal oxidation cannot grow oxide (3)(4)(5)(6). Thus, there is no doubt that active oxidants such as oxygen atoms and ions enhance oxidation.…”