Effects of Ion Implantation Doping on the Formation of Titanium Silicide on the Diffusion Layers
- PDF / 282,272 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 77 Downloads / 208 Views
H.MATSUI, H.OHTSUKI,
M.INO, and S.USHIO
VLSI Research and Development Center OKI Electric Industry Co., Ltd. Hachioji, Tokyo 193, Japan ABSTRACT Si samples, with and without masking oxide films,
implanted with
various doses of As+, P+, or BF2 + have been evaluated on the formation of titanium silicides from titanium films. In all cases, silicide reaction for implantation with masking oxide films is implantation without masking oxide films.
more difficult than that for
Silicide reaction becomes more
difficult with decreasing implant energy in the range over a critical dose. In the case of implantation with masking oxide films, knocked oxygen has been found at the surface of Si substrate.
Silicide formation after removing the surface layers containing considerable amount of knocked oxygen
with argon back-sputtering is as easy as silicide formation for implantation without masking oxide. The difficulty of Ti silicidation for implantation with masking oxide films is believed to be due to the effects of interference from knocked oxygen. INTRODUCTION As device sizes are reduced,
the resistivity of the interconnections
becomes the dominant limiting factor in the speed of the device. A self-aligned TiSi 2 process of gate and source/drain layers has been studied as the most promising candidate because resistivity of TiSi2 is the lowest among various silicides. However, reports on effects of ion implantation doping on the formation of TiSi2 have been rarely presented.1),2) In this work,
the difference of Ti silicide formation between the
samples implanted with and without masking oxide films has been studied. It has been found that silicide reaction for implantation with masking oxide films is more difficult than that for implantation without masking oxide films,
and the phenomenon is believed to be due to the effect of the knocked
oxygen.
Mat. Res. Soc. Symp. Proc. Vol. 54. ý 1986 Materials Research Society
770
EXPERIMENTAL PROCEDURES Fig.1 shows the basic experimental procedures.
(100)Si wafers, with
and without 20nm thick SiO2 films, were prepared.
These samples were
implanted with As,
P, or BF2 ions under various conditions,
950*C for 45 minutes in a N2 atmosphere.
and annealed at
Subsequently, wafers were cleaned
by H2 S0 4 +H2 02 solution and etched in HF solution immediately before the These Ti films were deposited by deposition of 50nm thick Ti films. The deposited Ti films were sputtering in a Perkin Elmer 4410 system. changed to Ti silicide films by two-step annealing3) in a N2 ambient furnace tube. The first annealing was performed at 600oC for 30 minutes. After removing the unreacted Ti films in NH4 OH+H202 solution, the second annealing was carried out at 8000C for 30 minutes. Sheet resistance of these samples were measured using a four-point Auger electron Spectroscopy (AES)
was used for the qualitative
analysis of the formed Ti silicide films.
Doping impurities and oxygen
probe.
profiles were examined by Secondary Ion Mass Spectroscopy (SIMS).
RESULTS AND DISCUSSIONS Dose Depe
Data Loading...
