XPS study of sub-monolayer native oxide on HF-treated Si Surfaces
- PDF / 402,465 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 54 Downloads / 172 Views
ABSTRACT Sub-monolayer native oxide on Si is quantitatively characterized by conventional XPS, using an Ols binding energy as a reference in spectrum decomposition of Si2p. This gives the average thicknesses of Si dioxide and Si suboxide in the sub-monolayer region. Using this technique, we investigate various native oxidation processes. We have found that oxygen dissolved in the HF solution influences native oxidation speed afterward. Furthermore As implantation at high concentrations (2 x 1015 cm- 2 ) in Si dramatically changes the oxidation process: the layer-bylayer feature clearly observed in undoped samples is entirely obscured in implanted samples. INTRODUCTION Suppressing the native oxidation of Si is very important in fabricating deep-submicron ULSIs. HF-treatment of the Si surface is of great interest because it removes surface oxide, makes the Si surface flat, and forms a stable surface with hydrogen atoms terminating dangling bonds 1.2. The oxidation processes for hydrogen-terminated Si surfaces have been widely investigated by X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared spectroscopy (FTIR), as well as other analytical techniques 1-7. The present work investigates the possibility of using conventional XPS to characterize Si native oxidation. Because peaks representing Si dioxide and Si suboxide are obscure, it is hard to quantitatively characterize Si oxide in the early stages of native oxidation. However, a new spectrum treatment technique is introduced in which the Si2p peak representing Si suboxide is decomposed, allowing the average thicknesses of Si dioxide and Si suboxide to be calculated.
This facilitates the quantitative analysis of sub-monolayer oxide 7 . In studying applications of this technique, we investigated the influence on Si native oxidation of the Si-crystal orientation, the dissolved-oxygen concentration in HF solution, and As implantation. EXPERIMENTAL Sample preparation Table I shows the sample preparation parameters. The wafers used in the experiment are B doped p-type CZ(100) and (111) wafers with a resistivity of 9 - 12 tcm. To see the influence of the Si-crystal orientation, Si(l 11) samples and Si(100) samples were prepared using HF solution with a dissolved-oxygen concentration of 8 ppm. To investigate the influence of the dissolvedoxygen concentration, Si( 111) samples were dipped in HF solutions with a dissolved-oxygen concentration of 8 mano 5 ppb. Finally, to see the influence of As implantation, As-implanted (1 x 10l0 ~ x 1010 cm-/, 25 kV) p-type 4 0 -off Si(100) wafers were annealed at 8500C for 30 minutes and treated in HF solution. All samples were cut into small pieces for convenience in XPS measurement, dipped in 5 mol% HF solution for 10 minutes, blown dry in N2 gas, exposed in a clean-room atmosphere, and analyzed by XPS. As reference samples, Ar-sputtered clean Si and thermally oxidized Si were used.
519 Mat. Res. Soc. Symp. Proc. Vol. 355 ©1995 Materials Research Society
TABLE 1.Sample preparationparameters wafer
As implantation
a
Data Loading...
