In the annealing case there is a diminishing influence of the
transient behavior recognized (see Fig. 4.3-6). Due to the
increased dopant diffusivity the profiles are determined by the dopant self
diffusion and the point defect enhancement. However, there is still
significant increase in diffusivity due to point defect enhancement.
A similar trend is exhibited at annealing temperature (see
Fig. 4.3-7). At this temperature the enhanced diffusion lasts
only a few seconds and the dopant redistributions predicts the final
profile. The shoulder of the profile is coincident with the solubility limit
and the boron precipitates are completely dissolved. The disagreement of the
profiles in the tail regions might be related to the choice of the initial
point defect profiles. The experimental profiles suggest some additional
enhancement at the slope of the distributions.
Figure 4.3-5: Simulation
results for boron transient enhanced diffusion at for 15min and
4h. Boron concentration above the solid solubility limit is immobile due
to clustering.
The given diffusion model results show good agreement with the experimental
data and are able to predict the transient enhanced diffusion effect during
furnace as well as RTA annealing. The most effective modeling parameters are
the recombination velocity of the point defect concentrations and
the interstitial mechansim enhancement factor
. Where
determines the available amount of point defects above the thermal
euqilibrium,
specifies the resulting dopant diffusion enhancement.
Figure 4.3-6: Simulation
results for boron at for 30min, 50min and 10h. The influence
of the transient enhancement is diminished due to higher dopant
diffusivity.
Figure 4.3-7: RTA simulation results
for boron transient enhanced diffusion at for 10s and 5min. The
boron aggregates are completely dissolved. Data are taken from
[Cow90a].