The substrate bow in a large area processing (LAP) was simulated using: a f
inite element analysis (FEA). The structures considered were aluminum (Al)
and glass substrates of various thicknesses, and a coating from Photosensit
ive film thickness. It was found that the deflection of a large area substr
ate, e.g., 400 mm(2), could not always be obtained from the linear, small d
eflection theory even if the curvature might be small and the stress-strain
behavior in the linear elastic regime, In this case, the nonlinear, large
deflection theory had to be adopted. Also, the gravity effect from the subs
trate weight turned out to be very significant and had to be incorporated a
s well. The simulation incorporating these two factors agreed well with the
experimental data, which was generated by spin coating and curing the BCB
formulation on Al substrates, 400 x 400 x 1.27 mm.
As a means of flattening out the curvature, subjecting a vacuum underneath
the substrate was simulated. Significant reduction of the substrate deflect
ion was observed by applying only a very small vacuum. This result suggeste
d that the use of double-stick tape on the bottom of the substrate, for exa
mple, might also be feasible to completely eliminate the bow.