Key points are not available for this paper at this time.
(i) A film of a nonwetting liquid is not stable. If it is thick (thickness e ≥ 1000 Å) it is metastable and evolves via nucleation and growth of a dry spot. If it is thin, it is unstable against spinodal decomposition (amplification of thermal undulations) and breaks into microscopic droplets of size ~e 2 /a (a = a molecular size), (ii) A film made with a wetting liquid is metastable and tends to shrink whenever e < e S (the equilibrium thickness of the wetting layer). Contrary to case (i) where the growth of a dry patch is controlled by capillary forces, the central role for drying in case (ii) is played by long-range forces, and the only process is nucleation and growth.
Wyart et al. (Sat,) studied this question.