Spin Coating of Complex Solutions Dunbar P. Birnie, III
The technique
described in the previous page shows how interference
data can be used to evaluate both flow and evaporation behavior simultaneously
-- at least for "well behaved" solutions. This limitation has been explored
by applying the technique to solutions that will clearly deviate from this
assumption -- and where the deviation can be understood to some degree.
In particular, I have used mixed solvent systems where the two solvents
have significantly different volatility (methanol + butanol). In addition,
I have done limited testing of the thinning behavior of sol-gel solutions
during coating.
The figure at right shows the "Meyerhofer plot" for a 50:50 mixture of
methanol+butanol. The solid line shows the early time behavior that is
indicative of a solution with nominally constant behavior -- i.e. there
is a straight line having reasonable slope and intercept based on the properties
of the two end-member solvents. However at later times (closer to the origin)
there is a deviation from this behavior and ultimately the data approach
a different limit (shown by the dotted line). This second limit corresponds
to a solution that is almost completely depleted of its methanol fraction.
The more rapidly evaporating methanol is preferentially removed from the
fluid as soon as the fluid thickness becomes small enough. Up to now, I
have made no attempt at modeling this evaporation and depletion effect
in any quantitative way. Hopefully I'll be able to tackle that at some
point in the future. This example demonstrates that great care must be
used when applying this technique to complex solutions and their behavior
when being used to make coatings.
In addition to using the technique on binary solvent mixtures, I have applied
it to the coating formation using sol-gel solutions. The figure at the
right shows the "Meyerhofer plot" for a PLZT 2/50/50 sol-gel. One linear
regime is visible at early times. At the latest stages, when the coating
forms, there is a distinct drop in thinning rate -- as one might expect
since this is when the sol-gel has probably gelled to some degree.
Future work
will be directed toward improving the sensivity of the technique (both
in speed and absolute reflectance). With improved equipment it may be possible
to reach a better understanding of the preferential evaporation effects
and the later stage of coating formation when the coating finally "sets".
A more complete
description of the above work has recently been published and can be found
in this journal:
D. P. Birnie, III, "Combined Flow and Evaporation During Spin Coating
of Complex Solutions", J. Non-Crystalline Solids, 218, 174-178 (1997).
The figure at right shows the "Meyerhofer plot" for a 50:50 mixture of
methanol+butanol. The solid line shows the early time behavior that is
indicative of a solution with nominally constant behavior -- i.e. there
is a straight line having reasonable slope and intercept based on the properties
of the two end-member solvents. However at later times (closer to the origin)
there is a deviation from this behavior and ultimately the data approach
a different limit (shown by the dotted line). This second limit corresponds
to a solution that is almost completely depleted of its methanol fraction.
The more rapidly evaporating methanol is preferentially removed from the
fluid as soon as the fluid thickness becomes small enough. Up to now, I
have made no attempt at modeling this evaporation and depletion effect
in any quantitative way. Hopefully I'll be able to tackle that at some
point in the future. This example demonstrates that great care must be
used when applying this technique to complex solutions and their behavior
when being used to make coatings.
In addition to using the technique on binary solvent mixtures, I have applied
it to the coating formation using sol-gel solutions. The figure at the
right shows the "Meyerhofer plot" for a PLZT 2/50/50 sol-gel. One linear
regime is visible at early times. At the latest stages, when the coating
forms, there is a distinct drop in thinning rate -- as one might expect
since this is when the sol-gel has probably gelled to some degree.