AVS 47th International Symposium
    Manufacturing Science and Technology Friday Sessions
       Session MS-FrM

Paper MS-FrM4
Standard Practice for Temperature Calibration of the Silicon Substrate in Temperature Programmed Desorption Analysis

Friday, October 6, 2000, 9:20 am, Room 304

Session: Langmuir Award/Ultra Clean Society and Contamination Free Manufacturing
Presenter: T. Matsunaga, Matsushita Inc.
Authors: T. Matsunaga, Matsushita Inc.
N. Yabumoto, NTT Adv. Tech. Co.
N. Hirashita, Oki Electric Ind. Co., Ltd.
H. Okumura, Toray Res. Center, Inc.
M. Nishiduka, Toshiba Corp.
I. Nishiyama, NEC Corp.
M. Matsuura, Mitsubishi Corp.
M. Morita, Osaka Univ.
A. Shimazaki, Toshiba Corp.
T. Jimbo, Hitachi, Ltd.
T. Ajioka, NTT Electronics Corp.
Correspondent: Click to Email
Ultra Clean Society (UCS)’s Equipment Standardization Working Group proposed a standard practice covering temperature calibration of the silicon substrate, ranging from 400 to 1000°C, for temperature programmed desorption (TPD) analysis. Although TPD has been widely used to characterize materials and fabrication processes in ULSI devices, the temperature is not accurate enough to develop reliable fabrication processes. The desorption temperature was found to differ over 100°C between interlaboratory TPD measurements. In order to solve this problem, the ramping temperature of TPD instrument was calibrated to silicon substrate temperature by this proposed standard practice, which consists of heating silicon calibration materials at a controlled rate in TPD instrument, measuring characteristic desorption peak temperatures, and quadratic calibration fitting to the standard temperatures. The calibration materials are (1) CaC@sub 2@O@sub 4@.H@sub 2@O dropped on a Si wafer and dried, (2) Ar and (3) H ion implanted into Si wafers. The standard temperatures of the characteristic desorption, associated with decomposition, structural transformation and lamination of silicon, were determined by special TPD instrument with the highest isothermal space around the specimen among the interlaboratory. The precision of this practice was determined in an interlaboratory test in which 4-5 laboratories participated using two instrumental models. This test using a few common specimens proved that the average standard deviation, measured in different laboratories for all the measurements with ramping rates of 10, 30 and 60°C /min, was estimated to be 6.0°C between 400 and 1000°C.