The thickness and refractive index of poly(dimethylsiloxane) (PDMS) thin films on silicon, exposed to high-pressure carbon dioxide, have been measured in-situ by spectroscopic ellipsometry. The swelling of the film is determined from the thickness and the sorption of CO2 from the refractive index. The sorption and swelling values for the thin films exceed those of the bulk films. These increases may be attributed to excess CO2 at the polymer thin film−CO2 and polymer−silica interfaces and the influence of the silica surface and CO2 on the orientation of the polymer. Pressure-induced birefringence of the fused-silica windows was addressed in the model calculations and corrections were found to be successful for a thermal oxide reference wafer and the PDMS films. Large surface excesses of CO2 on the reference wafer were observed in regions where CO2 is highly compressible due to critical adsorption.

Carbon dioxide produces an anomalous maximum in the swelling of poly(methyl methacrylate) (PMMA) thin films that is not present in bulk films, as shown with in-situ spectroscopic ellipsometry. This maximum and a corresponding minimum in refractive index are observed in regions of pressure where CO 2 is highly compressible near the critical point. An effective excess thickness, determined from the height of the anomalous swelling maxima, increases proportionally with increasing initial film thickness, h0, in the range studied of 85 nm < h0 < 325 nm. Therefore, the anomalous swelling maxima suggest concentration inhomogeneities in the thin film, i.e., the onset of polymer/CO2 phase separation, that are influenced by the compressibility of the system and the confinement of the film. The PMMA swelling isotherms are insensitive to changing the substrate from silicon to gallium arsenide.

We examined the dynamics of dewetting of a thin symmetric diblock copolymer film on a substrate above the bulk order−disorder transition temperature, T ODT, of the copolymer using atomic force microscopy. The dewetting mechanism proceeded with the formation of discrete holes without their characteristic peripheral rims. During this early stage, the hole radii R increased exponentially with time. This stage was followed by a narrow intermediate regime where the rim develops and R ∼ t. When the rim was fully developed, R increased as t 2/3. The shape of the rim was highly asymmetrical and its width L increased as t 1/2. At the final stage of the process, droplets of the copolymer, a few microns in diameter and with heights on the order of tens of nanometers, existed on a dense copolymer “brush” of uniform thickness 7 nm anchored to the substrate. This clearly indicates that the process is autophobic, a phenomenon first documented in small molecule liquids.

Thin symmetric poly(styrene-b-methyl methacrylate) (PS-b-PMMA) diblock copolymer films at temperatures higher than the bulk order−disorder transition temperature T ODT are shown to dewet silicon substrates, forming topographical features that depend on the initial film thickness h. Films of thickness h < 3.5 nm dewet the substrate, forming bicontinuous spinodal-like patterns. When 3.5 nm < h < h L = 7 nm, discrete holes are observed randomly throughout the film. For films of thickness in the range h L < h < h H = 35 nm, the copolymer exhibited autophobic behavior, whereby the top layer of thickness (h − h L) dewets a dense “brush” of ordered copolymer of height h L anchored to the silicon substrate. The morphologies, which include a bicontinuous spinodal pattern for films of thickness in the range h L < h < 19 nm, and discrete holes, for films of thickness in the range 19 nm < h < h H, eventually evolve into droplets. Films of h > 35 nm remained stable, with smooth surfaces. The time-dependent evolution of the “spinodal” structures that evolve in the autophobic regime is discussed. In addition, the existence of surface-induced ordering of the copolymer at temperatures above the bulk T ODT is also discussed.

Since the early seminal theoretical work by Brochard and coworkers, and experiments by Reiter over a decade ago, considerable progress has been made toward the development of a comprehensive picture of the "equilibrium" and dynamic behavior of unstable thin polymeric films. Generally, theoretical work has carefully guided the design of many experiments conducted in this field. Recent experimental findings, however, raise new questions that could probably not have been foreseen by theory and now need to be revisited. In this paper we highlight three problems in two general areas, (1) the use of the effective interfacial potential describing film substrate interactions and (2) the dynamics of dewetting and the associated connection to "slip" phenomena and fingering instabilities. We suggest that in addition to experiments, analytical theory and simulations will play a critical role toward elucidating the ultimate goal of a universal picture of "equilibrium" and dynamic behavior of instabilities in thin films.