I. Debusschere et al., IMPORTANCE OF DETERMINING THE POLYSILICON DOPANT PROFILE DURING PROCESS-DEVELOPMENT, Journal of vacuum science & technology. B, Microelectronics and nanometer structures processing, measurement and phenomena, 14(1), 1996, pp. 265-271
During the development of a 0.5 mu m, n-type poly, complementary metal
-oxide-semiconductor (MOS) process the influence of different process
steps on the PMOS gate doping level was investigated. The application
of spreading resistance measurements and secondary ion microscopy anal
ysis for determination of the polysilicon dopant profile allowed us to
define critical process conditions within which the process could be
optimized using short loop tests. The influence of the polysilicon thi
ckness and phosphorus doping level, of the p(+) source/drain implantat
ion dose, of the subsequent temperature steps, and of the silicidation
were all investigated. It was concluded that a combination of certain
process conditions, such as a polysilicon implant dose of 4.0 x 10(15
) combined with a p-plus dose of 3.0 x 10(15), does lead to unacceptab
ly low active carrier concentrations, which are proven to result in ga
te depletion. Furthermore, it was noted that in case of low net impuri
ty concentrations the modification of any of the processing conditions
has a large influence on the final polysilicon spreading resistance p
robe profile. A detailed discussion of the different parameters and th
e resulting doping profiles are given in the article. Based on the res
ults of the short loop tests the p-plus implant dose was fixed at 2.0
x 10(15). A full transistor lot was processed in which the remaining p
arameter, i.e., the polysilicon implantation dose, was varied. The dev
ice characteristics determined on this lot are in agreement with predi
ctions, based on the results of the experiment. (C) 1996 American Vacu
um Society.