Transparent barrier coatings on polymers are receiving much attention in industry, for pharmaceutical, food and beverage packaging applications. Plasma-enhanced chemical vapor deposition (PECVD) is among several competing techniques which can produce thin layers of inorganic glassy barrier materials. In this article we describe the performance of silicon compounds (SiO2 and Si3N4) on 13 μm polyethylene terephthalate (PET) substrates, the barrier coatings being deposited in a dual-frequency (microwave/radio frequency) pilot-scale PECVD reactor for continuously moving flexible webs up to 30 cm in width. The volatile silicon compound used for SiO2 deposition is HMDSO (C6H18Si2O), while SiH4 serves to deposit Si3N4. Coating thicknesses, d, in the range 8 nm⩽d⩽200 nm, are measured using a variety of techniques, namely stylus profilometry, continuous wavelength optical interferometry, x-ray fluorescence, variable angle spectroscopic ellipsometry, and transmission electron microscopy, while film compositions are determined by x-ray photoelectron spectroscopy. Oxygen transmission (OTR) and water vapor transmission (WVTR) measurements are carried out with MOCON “Oxtran” and “Permatran-W” instruments, respectively. As also reported by other workers, we typically find OTR values of about 0.5 scc/m2 day and WVTR about 0.3 g/m2 day, for barrier thicknesses exceeding a “critical” value (dc, about 15 nm), but the minimum permeation values depend upon the concentration of defect sites in the coating (mostly related to substrate microroughness). In order to confirm this correlation, we have developed a technique combining reactive ion etching through the PET, followed by optical and transmission electron microscopies, to characterize the types and number densities of coating defects. On the basis of these, we find good agreement between measured and calculated values of OTR.

No abstract

Amorphous hydrogenated silicon carbonitride films were produced by remote plasma chemical vapor deposition (RP-CVD) from 1,1,3,3-tetramethyldisilazane (TMDSN) as the single-source compound using a H 2 -N 2 upstream-gas-mixture for plasma generation. The reactivity of particular TMDSN bonds in the RP-CVD initiation step has been examined using a hexamethyldisilazane model compound in the deposition experiments. The active species contributing to RP-CVD were identified by optical emission spectroscopic analysis of the plasma region. The films were examined using Fourier transform infrared spectroscopy, X-ray photoelectron spectroscopy, and atomic force microscopy. The effect of N 2 content in the H 2 -N 2 upstream-gasmixture on plasma generation of the active species, growth rate, chemical structure, and surface morphology of the resulting films is reported. {Electronic supplementary information (ESI) available: deconvoluted emission and IR spectra of a-Si-N-C-H films. See

Diffusion at the polymer/polymer interface was probed by small-amplitude oscillatory shear measurements carried out on polystyrene (PS)/polystyrene (PS) sandwich-like assembly as a function of the time of welding in the molten state. It was found that the dynamic complex shear modulus G*(t) at a fixed frequency increases with the time of contact in two time regimes. First G*(t) increases proportionally to t 1/2 and then a second regime takes place where G*(t) increases proportionally to t 1/4. At longer times, G*(t) tends asymptotically toward G* of pure polystyrene. The results were interpreted in terms of reptation theory and the time of transition between the two scaling law regimes was found to be in agreement with the time needed for the transition from the Rouse mode to the reptation mode. Special attention was given to the initial state of the polymer surfaces before contact by performing experiments on (i) freshly prepared surfaces, (ii) presheared samples, (iii) fractured samples, and (iv) corona-treated samples. The results showed that the diffusion mechanism is strongly dependent on the initial chain-end distribution at the surface before contact. Diffusion at surfaces with an excess of chain ends proceeds in a Rouse-like mode, whereas for surfaces without excess of chain ends, the diffusion proceeds in a reptational-like dynamics, as was predicted by de Gennes and by Prager and Tirrell.