a) b - École Polytechnique de Montréal

dispersed phase-matrix, fibrillar, and co-continuous morphologies, relatively few studies have
focused on the fundamentals of three or more components with complex morphologies(Moussaif
& Jérôme, 1999; Reignier & Favis, 2000, 2003a; J. Zhang, et al., 2007).
Ternary A/B/C systems can exist in two possible states. One is known as complete wetting and
the other is partial wetting. Complete wetting can result in B and C droplets individually
dispersed in an A matrix. It can also result in the case where one B or C phase completely
engulfs the other in a matrix of A. In this latter case, the minimization of interfacial free energy
occurs when, for example, phase B is situated at the interface of, and completely wets, phases A
and C. Phase B prevents any contact between phases A and C(Figure 4-1a). In the partial wetting
state, all three phases have an interface with each other(Figure 4-1b). In such a case intact
droplets of B, for example, can be placed at an A/C interface. Recently some very novel
structures have been generated via partial wetting(Virgilio, Desjardins, et al., 2009; Virgilio,
Marc-Aurele, & Favis, 2009).
Strict thermodynamic conditions need to be met in order to have either discrete phases,
encapsulated structures or partial wetting. Torza and Mason(Torza & Mason, 1970) and then
Hobbs et al.(Hobbs, et al., 1988) employed a modified Harkins spreading theory (Equation 4-1)
to predict whether the morphology of a ternary blend is dominated by complete wetting or is
partial wetting.
λ = γ −γ
−γ
Equation 4-1. 12 23 13 12
where λ is the spreading coefficient, γ represents the interfacial tension for various polymer pairs
where the sub-indexes refer to each component. If one of the spreading coefficients such as λ12
has a positive value, a complete wetting case occurs in which phase 1 separates phases 2 and 3
(Figure 4-1a). In another case, where all spreading coefficients have negative values, the system
demonstrates a partial wetting case in which all phases have contact with each other(Figure 4-
1b). It has been shown that Harkins equation is, on the whole, a good criterion to predict the
position of phases in ternary polymer blends(Reignier & Favis, 2000, 2003b; Virgilio,
Desjardins, et al., 2009; Zhang, et al., 2007). As the number of phases increases to four however,
it is difficult to determine the morphology by such a simple model. To date, there is no equation
96