Polycarbonate-silsesquioxane and polycarbonate-siloxane nanocomposites: Synthesis, characterization, and application in the fabrication of porous inorganic films

Three types of poly(bicycle[2.2.1]heptane carbonate) or poly(norbornane carbonate) or PNC oligomers were synthesized and characterized via spectroscopic methods and elemental analyses to validate their chemical structures. End-group analyses were used to estimate the degree of polymerization of the oligomers via the use of proton nuclear magnetic resonance (1H NMR) results. Random-coil and rigid-rod models were used to estimate the sizes of individual PNC chains based on the degrees of polymerization calculated from NMR data. Due to the small sizes of the PNC chains, dynamic light scattering (DLS) was incapable of measuring the hydrodynamic radii, RH, of individual chains. Attempts at using gel permeation chromatography (GPC) data to estimate the hydrodynamic radii of individual chains consistently provided values that were an order of magnitude smaller than the estimated sizes of individual chains based on random-coil calculations. The thermal properties of PNCs were determined via differential scanning calorimetry (DSC) and thermogravimetric analyses (TGAs). All three types of PNC structures were both thermally-labile and acidolytically-labile, allowing them to be used as sacrificial materials in both direct-write and thermally-processed template systems. TGA data was used to determine the kinetic parameters for the thermolytic decomposition reactions and evolved-gas analysis via mass spectrometry (TGA-MS) was used to validate the mechanisms for polycarbonate thermolysis reactions that have been previously proposed in literature.;PNC oligomers were freely-mixed with hydrogen silsesquioxane (HSQ) to form solutions that were spin-coated to form templated films. Ellipsometry and dielectric measurements were used to track the changes in the optical and dielectric properties of templated films and effective medium approximations were used to estimate the level of porosity incorporated within each porous film. Transmission electron microscopy (TEM) showed that the free-mixing of PNCs with HSQ resulted in the agglomeration of the porogen molecules during the spincoating step. This phase-segregation led to the formation of domains with dimensions much larger than those of the individual chains, and during decomposition large pores were produced. To combat the phase segregation, hydrosilylation reactions were used to covalently bond vinyl end-capped PNC chains to silane-functionalized siloxane and silsesquioxane molecules. These matrix-like materials served as compatibilizers in order to improve the phase-compatibility of the sacrificial polymers in HSQ films. NMR and GPC analyses showed that the solids recovered from the hydrosilylation reactions were binary mixtures of hybrid nanocomposite molecules and residual ungrafted chains. All attempts at isolating the hybrid molecules proved to be unsuccessful and the solids were templated as blends in HSQ films.;TEM imaging showed that the domains in these nanocomposite films had bimodal size distributions due to the presence of two components in the mixtures. The hybrid molecules produced pores ranging in size from about 6-13 nm as a result of improvements in the phase-compatibility of the grafted oligomers. However, the residual ungrafted oligomers in the blends produced larger domains measuring 30-40 nm. Although the siloxane and silsesquioxane molecules were shown to fulfill the stated goal of compatibilizing the PNC chains with HSQ and the hybrid molecules produced domain sizes comparable to those of templated films reported in literature, the difficulty in isolating the hybrid molecules from the ungrafted oligomers limits the benefits of using these blends as porogen materials. It is believed that separation difficulties can be avoided if the physical and chemical conditions used in the vinyl termination reactions can be adjusted to ensure 100% conversion of all the terminal hydroxyl groups to vinyl groups. Doing so would allow all PNC chains to be grafted during hydrosilylation reaction; thus, avoiding the recovery of ungrafted oligomers. The recovery of pure hydrosilylation products would allow monodisperse domains with sizes ranging from 6-13 nm to be produced in templated films. Additionally, improvements can still be made in the morphology of hybrid films by successfully grafting polycarbonate chains directly to HSQ prior to spincoating thin films. Although all attempts at performing this reaction using PNC chains were unsuccessful, it is expected that another polycarbonate with a less sterically-hindered chemical structure may be successfully bonded to HSQ. Based on literature data, such a template system can be expected to produce pores as small as the individual polycarbonate chains, which is an ideal morphology for low-k applications.