Joshua M. Katzenstein

Polymeric Cross-Linked Surface Treatments for Controlling Block Copolymer Orientation in Thin Films

The orientation of cylinder-forming poly(styrene-block-methyl methacrylate) [P(S-b-MMA)] was investigated on two sets of polymeric surface treatments: 10 para-substituted polystyrene derivatives with <10 mol % poly(4-vinylbenzyl azide) and a series of poly(styrene-random-4-vinylbenzyl azide) [P(S-r-VBzAz)] copolymers with 5-100 mol % poly(4-vinylbenzyl azide). The copolymers were spin-coated to form thin films and then cross-linked by heating. The resulting films exhibited a range of surface tensions from 21 to 45 dyn/cm. Perpendicular orientation of P(S-b-MMA) cylinders was achieved with poly(p-bromostyrene) and all the [P(S-r-VBzAz)] copolymer surface treatments, most notably the homopolymer of poly(4-vinylbenzyl azide). Films made from these simple copolymers are as effective as random terpolymer alignment layers commonly made from both block monomers and a cross-linkable monomer.

Precision Marangoni-driven patterning

A Marangoni flow is shown to occur when a polymer film possessing a spatially-defined surface energy pattern is heated above its glass transition to the liquid state. This can be harnessed to rapidly manufacture polymer films possessing prescribed height profiles. To quantify and verify this phenomenon, a model is described here which accurately predicts the formation, growth, and eventual dissipation of topographical features. The model predictions, based on numerical solutions of equations governing thin film dynamics with a Marangoni stress, are quantitatively compared to experimental measurements of thin polystyrene films containing photochemically patterned surface energy gradients. Good agreement between the model and the data is achieved at temperatures between 120 and 140 °C for a comprehensive range of heating times using reasonable physical properties as parameter inputs. For example, thickness variations that measure 102% of the starting film thickness are achieved in only 12 minutes of heating at 140 °C, values that are predicted by the model are within 6% and 3 min, respectively. The photochemical pattern that directed this flow possessed only a 0.2 dyne cm(-1) variation in surface tension between exposed and unexposed regions. The physical insights from the validated model suggest promising strategies to maximize the aspect ratio of the topographical features and minimize the processing time necessary to develop them.