Domain orientations in thin films of lamellar copolymers are evaluated as a function of copolymer architecture, film thickness, and processing conditions. Two copolymer architectures are considered: An AB diblock of poly(styrene-b-methyl methacrylate) and an ABA triblock of poly(methyl methacrylate-b-styrene-b-methyl methacrylate). All films are cast on substrates that are energetically neutral with respect to the copolymer constituents. Film structures are evaluated with optical microscopy, atomic force microscopy, and grazing-incidence small-angle X-ray scattering. For AB diblock copolymers, the domain orientations are very sensitive to film thickness, annealing temperature, and imperfections in the “neutral” substrate coating: Diblock domains are oriented perpendicular to the substrate when annealing temperature is elevated (≥220 °C) and defects in the substrate coating are minimized; otherwise, parallel or mixed parallel/perpendicular domain orientations are detected for most film thicknesses. For ABA triblock copolymers, the perpendicular domain orientation is stable for all the film thicknesses and processing conditions that were studied. The orientations of diblock and triblock copolymers are consistent with recent works that consider architectural effects when calculating the copolymer surface tension (Macromolecules2006399346 and Macromolecules2010431671). Significantly, the data demonstrate that triblocks are easier to process for applications in nanopatterningin particular, when high-aspect-ratio nanostructures are required. However, both diblock and triblock films contain a high density of “tilted” or bent domains, and these kinetically trapped defects should be minimized for most patterning applications.

The depth-dependent structure of a poly(styreneb-methylmethacrylate) (PS-PMMA) line grating (46 nm pitch) was calculated from quantitative analysis of small-angle X-ray scattering profiles. These data demonstrate that domain shapes are significantly deformed near the substrate interface, where the local PS domain shape resembles an hourglass. The bulk equilibrium dimension is recovered near the center of a 64 nm thick film. Simulations based on self-consistent field theory suggest that deformations near the substrate are caused by extensive penetration of the copolymer domains into the underlying substrate coating (a PS-brush). These findings suggest that new coatings for block copolymer directed self-assembly should consider copolymer penetration lengths in addition to tailoring surface energetics. Furthermore, given the resolution and ensemble-averaging features of synchrotron X-ray scattering, we argue that it has the potential to emerge as a "gold-standard" or "benchmark" dimensional metrology and library validation tool for high density, sub-10 nm features.

A quantitative description of kinetics in acid-catalyzed polymer deprotection reactions requires proper identification of the controlling mechanisms. We examined the acid-catalyzed deprotection of a glassy poly(4-hydroxystyrene-co-tert-butyl acrylate) resin using infrared absorbance spectroscopy and stochastic simulations. We interpret experimental data with a model that explicitly accounts for acid transport, where heterogeneities at local length scales are introduced through a nonexponential distribution of waiting times between successive hopping events. A subdiffusive behavior with long-tail kinetics predicts key attributes of the observed deprotection rates, such as a fast initial deprotection, slow conversion at long times, and a nonlinear dependence on acid loading. Most importantly, only two parameters are introduced to offer a near-quantitative description of deprotection levels at low acid loadings and short times. The model is extended to high acid loadings and long times by incorporating a simple acid depletion model based on mutual encounters. Our study suggests that macroscopic deprotection rates are controlled by acid transport in the glassy deprotected polymer, which presents with a strongly non-Fickian behavior.

Mesoporous polymer and carbon thin films are prepared by the organic-organic self-assembly of an oligomeric phenolic resin with an amphiphilic triblock copolymer template, Pluronic F127. The ratio of resin to template is selected such that a body-centered cubic (Im3m) mesostructure is formed in the bulk. However, well-ordered mesoporous films are not always obtained for thin films (<100 nm), and this behavior is found to be directly correlated with the initial phenolic resin to template ratio. Furthermore, the symmetry of ordered phases is highly dependent on the number of layers of spheres in the film: Monolayers and bilayers are characterized by hexagonal close-packed (HCP) symmetry, while films with approximately 5 layers of spheres exhibit a mixture of HCP and face-centered orthorhombic (FCO) structures. Ultrathick films having more than 30 layers of spheres are similar to the bulk body-centered cubic symmetry with a preferential orientation of the closest-packed (110) plane parallel to the substrate. Film thickness and initial composition of the carbonizable precursors in the template are critical factors in determining the morphology of mesoporous carbon films. These results provide insight into why difficulties have been reported in producing ultrathin ordered mesoporous carbon films using cooperative organic-organic self-assembly.

The authors demonstrate a simple method to identify noise sources in electron-beam systems and accurately quantify the resulting errors in feature placement. Line gratings with a 46 nm average pitch were patterned with electron-beam lithography and measured with transmission x-ray diffraction (XRD) and scanning electron microscopy (SEM). All SEM micrographs were analyzed in Fourier space to facilitate comparison with the XRD data. Diffraction profiles and Fourier transforms of SEM micrographs contained numerous “satellite” peaks, meaning weak peaks adjacent to the strong primary nodes, that are characteristic of periodic extensions and compressions in the grating pitch. The wavelength and amplitude of these pitch variations were calculated with a simple scaling law by comparing the positions and intensities of satellite peaks relative to their neighboring primary nodes. This approach is remarkably easy to implement because it does not require any modeling of electron density profiles. Data were used to calculate the frequency of each noise source and the resulting variations in the grating pitch. Two persistent noise frequencies were detected in the tool studied, (62±2) Hz and (86±3) Hz, and the tool manufacturer identified likely noise sources as electromagnetic and mechanical in nature, respectively. The 60 Hz noise produced errors in a 46 nm grating pitch of 3σ=1.5 nm, where σ is the standard deviation in the grating pitch. Errors due to the 86 Hz noise ranged from 3σ=1.5 to 2.5 nm. Variations of these magnitudes can be expected to have adverse effects on coupling efficiencies, cavity quality factors, and center wavelength values in photonic devices.