Electrowetting of Ionic Liquids in Solid–Liquid–Liquid Systems - Drops and Bubbles in Contact with Solid Surfaces - page 259

mating the limit of applicability of the simple thermodynamic model of a sessile
drop [2, 17]. It does not predict contact angle behaviour beyond this point and these
results may be useful in developing further the theory of electrowetting saturation.
A key difference between solid-liquid-vapour and solid-liquid-liquid systems is
the much higher density and viscosity of the ambient fluid (hexadecane in our case).
In early research on electrowetting in solid-liquid-liquid systems [6] it was already
suggested that an alkane film is trapped between the droplet and the solid surface.
Since then further evidence for that has been accumulated [6, 14, 36]. Staicu and
Mugele [36] found a silicone oil film entrapped under the aqueous NaCl droplet on
Teflon AF1600 surface. The oil film was initially thick ( 400 nm), then became

unstable and broke into small droplets (the same effect has been used to enhance
the performance of electrowetting display pixels [11]). In their experiments, voltage
was applied before the droplet had touched the insulated electrode and the viscosity
ratio was very high ( m

∼

100). We therefore suspect that the oil film
was hydrodynamically entrained, i.e., was transient in character. We estimate this
was not the case in our experiments because of the specifically designed measure-
ment protocol.
Quilliet and Berge [32] estimated theoretically that an equilibrium oil film may
be present due to van der Waals interactions. This seems unlikely as, for instance,
the refractive indices of bmim.BF 4 (1.421) and hexadecane (1.434) are very similar
and therefore the Hamaker constants for the two materials should not be signif-
icantly different [78]. Thus we expect only weak van der Waals forces to oper-
ate. Electrostatic repulsion, as understood in traditional colloidal terms, should be
screened very effectively by the large ionic strength of the ionic liquid [47]. An
oil film would be even more unstable under electrowetting conditions [32, 79].
Recently strong structural forces with oscillatory decaying character have been
reported in confined ionic liquids [80-83] and these may prove to be important.
In view of these developments, the zero-interfacial-tension hypothesis should be
reconsidered in relation to the Frumkin-Derjaguin model [84, 85] (the model in
which the liquid meniscus contacts a thin wetting film rather than the non-wetted
solid-fluid interface). Nonetheless, in the next section we provide a consistent inter-
pretation of the dynamics of electrowetting within a macroscopic framework which
does not consider thin wetting films.
Contact angle saturation in electrowetting is highly undesirable as it limits its
usefulness in actuation applications. It is therefore significant that, when using
AC voltage, saturation is delayed and very low contact angles are achievable—
Fig. 7 and Fig. 11. We have previously seen the same effect in solid-liquid-vapour
systems [2]. This superior performance of AC voltage in electrowetting has been
explained by Hong et al. [86]. They modelled numerically the electromagnetic field
inside a conductive droplet sitting on an insulator in air. When AC voltage (fre-
quency 1-16 kHz) was used to achieve a contact angle change identical to that
obtained with DC voltage, the local electric field near the contact line was weaker
and did not trigger contact angle saturation. Regardless of the physical details,

= μ oil /μ water ≈