Confinement induced ordering in dewetting of ultra-thin polymer bilayers on nanopatterned substrates

We report the dewetting of a thin bilayer of polystyrene (PS) and poly(methylmethacrylate) (PMMA) on a topographically patterned nonwettable substrate comprising an array of pillars, arranged in a square lattice. With a gradual increase in the concentration of the PMMA solution (Cn-PMMA), the morphology of the bottom layer changes to: (1) an aligned array of spin dewetted droplets arranged along substrate grooves at very low Cn-PMMA; (2) an interconnected network of threads surrounding each pillar at intermediate Cn-PMMA; and (3) a continuous bottom layer at higher Cn-PMMA. On the other hand the morphology of the PS top layer depends largely on the nature of the pre-existing bottom layer, in addition to Cn-PS. An ordered array of PMMA core-PS shell droplets forms right after spin coating when both Cn-PMMA and Cn-PS are very low. Bilayers with all other initial configurations evolve during thermal annealing, resulting in a variety of ordered structures. Unique morphologies realized include laterally coexisting structures of the two polymers confined within the substrate grooves due to initial rupture of the bottom layer on the substrate followed by a squeezing flow of the top layer; an array of core-shell and single polymer droplets arranged in an alternating order etc., to highlight a few. Such structures cannot be fabricated by any stand-alone lithography technique. On the other hand, in some cases the partially dewetted bottom layer imparts stability to an intact top PS layer against dewetting. Apart from ordering, under certain specific conditions significant miniaturization and downsizing of dewetted feature periodicity and dimension as compared to dewetting of a single layer on a flat substrate is observed. With the help of a morphology phase diagram we show that ordering is achieved over a wide combination of Cn-PMMA and Cn-PS, though the morphology and dewetting pathway differs significantly with variation in the thickness of the individual layers.