1 - 2
Most diffusion furnaces are capable of heating at a rate
of about 10°C/min. This is an acceptable rate for sources
that are 2 or 3 inches in diameter. However, the slower
heating rates given in Table I
are recommended for the larger diameter sources to insure
uniform heating across their diameters. If temperature gradients
are too great, silicon damage and nonuniform doping can
occur along the outside edge.
If controlled heating and
cooling rates are not available, uncontrolled rates are
usually acceptable. In this case, however, the two end zones
must be slave-controlled to the center zone. If they are
not slave-controlled, non-uniform heating of the boat may
occur resulting in non-uniform doping across the boat.
Sufficient time must be
allowed at the deposition temperature for a solid source
system to reach thermal equilibrium and for the sources
to deposit enough B2O3 to uniformly
dope the surfaces of the silicon wafers. The minimum time
that is required to give good results increases with increasing
source diameter. Experience has shown that times less than
those given in Table I are
generally too short to obtain acceptable uniformities.
Some users may choose
to insert their diffusion carrier directly into the furnace
at the deposition temperature. Direct insertion is not generally
recommended for any size source. Process engineers who choose
to use direct insertion may observe premature warpage of
the sources and nonuniform doping of the silicon, especially
across the boat.
If the direct insertion
technique is selected, however, the diffusion engineer may
find it necessary to tilt the furnace temperature profile
to compensate for the longer time at temperature for sources
and silicon at the gas inlet end (source end) of the boat
(first-in last-out effect). An alternate technique to direct
insertion into the hot zone that sometimes works for 2 and
3 inch diameter systems is to quickly push the diffusion
carrier through the hot zone, allow the silicon and sources
to stabilize near the gas inlet end of the tube, and then
pull the boat slowly back into the hot zone. At the end
of the deposition time, the boat is then slowly pulled back
to the front of the tube from the hot zone. Since this technique
makes the first-in end of the boat also the first-out end,
variations across the boat are minimized.
1 - 2
