Chemical vapor deposition of undoped and in-situ boron- and arsenic-doped epitaxial and polycrystalline silicon films grown using silane at reduced pressure
J. Pejnefors et al., Chemical vapor deposition of undoped and in-situ boron- and arsenic-doped epitaxial and polycrystalline silicon films grown using silane at reduced pressure, J APPL PHYS, 88(3), 2000, pp. 1655-1663
A nonselective epitaxial growth process for heterojunction bipolar transist
ors has been studied. The difference in growth rates for epitaxial and poly
crystalline films could be used to monitor the thickness of the intrinsic a
nd extrinsic base layers. The films were grown using chemical vapor deposit
ion on Si < 100 > (epitaxy) and on silicon dioxide (polycrystalline) at red
uced pressure (20-80 Torr) for undoped and in situ B or As doping. The depo
sitions were carried out using silane diluted in hydrogen. Diborane and ars
ine were used as the source gas for dopants. For the undoped Si films, the
deposition of polycrystalline films had a substantially higher rate than th
at of epitaxial ones. The growth rate of both epitaxial and polycrystalline
depositions decreased with increasing total pressure. It was, however, lin
early proportional to the silane partial pressure, p(SiH4). The dependence
of the growth rate on the hydrogen partial pressure was proportional to p(H
2)(-0.82) for epitaxial and to p(H2)(-0.60) for polycrystalline depositions
. The apparent activation energy was 2.1 and 1.6 eV for the epitaxial and p
olycrystalline depositions, respectively. A growth mechanism assuming the d
issociative adsorption of silane on the Si surface, in combination with fir
st-order hydrogen desorption kinetics, was employed to describe the experim
ental observations, including the differences in deposition rates, dependen
cy on the hydrogen partial pressure as well as apparent activation energy.
In situ B doping influenced neither the epitaxial nor polycrystalline depos
itions. In situ As doping, on the other hand, largely reduced the growth ra
te compared to the undoped films to such an extent that there was no apprec
iable difference in growth rate between the epitaxial and polycrystalline S
i. The doping concentration in the epitaxial B and As films were of the ord
er of 10(18) cm(-3), identical deposition conditions yielded a 5 and 20 tim
es larger dopant incorporation in the B and As doped polycrystalline films,
respectively. (C) 2000 American Institute of Physics. [S0021-8979(00)03315
-6].