No abstract

No abstract

In order to understand the mechanism of adhesion promotion by organo silane in joints where its chemical reaction with a polymer matrix may be negligible, we studied the role of γ‐aminopropyl triethoxy silane (γ‐APS) in α‐Al2O3/ polyethylene joints. When adsorbed or deposited on α‐A12O3/ and moderately dried, γ‐APS forms a multimolecular film that is not fully cured. Drying at elevated temperatures for an extended time leads to further curing, resulting in a glassy silane film with a Tg around 108°C, while prolonged dry may cause some degradation. Joint strength is markedly improved by the application of a γ‐APS film with a maximum peel strength of 2.3 kg/cm when the silane is applied from a 2 percent aqueous solution. Drying the silane at an elevated temperature prior to joint formation reduces joint strength and also changes the failure mode from cohesive failure through the polyethylene to a mixed mode. Evidence of interdiffusion between γ‐APS film and polyethylene at a temperature (149°C) that is above the melting point of polyethylene and the Tg of the silane film was obtained by measuring the Si concentration profile across the interface of the laminate of polyethylene/γ‐APS/polyethylene. Diffusion constants in the order of 10−12 cm2/s were obtained, with a teridency toward reduced diffusion as a consequence of extensive drying. DSC results indicate at least partial miscibility of the silane polymer in the amorphous region of polyethylene. It is proposed that the interdiffusion between γ‐APS silane film and polyethylene is an important mechanism for adhesion promotion of the joints investigated in this study.

The sorption isotherm and the polymer mass‐fixed diffusion coefficients, D, for toluene in butyl rubber have been measured by the incremental sorption method to concentrations of 130%, corresponding to a solvent volume fraction of 0.578. The increase in D with concentration is strongly exponential to a concentration of 30% and then begins to level out. Since the nature of the dimensional change occurring in vapor sorption was not known, the values of D were converted to solvent self‐diffusion coefficients, D1, assuming both swelling in the thickness direction (1D) and isotropically (3D). The free volume (FV) theory of Fujita was fitted to the resulting D1 with the zero concentration diffusion coefficient and the self‐diffusion coefficient of toluene as limiting values leaving only a single arbitrary parameter. In this form the FV theory was able to describe the trend of the experimental D1 for the 1D and 3D cases equally well. Values of D were back‐calculated from the FV relations for the 1D and 3D cases for comparison with the experimental results and with the diffusion coefficient determined by immersion in toluene. These comparisons favor the assumption that swelling is isotropic. It appears that the simple free volume relation is capable of providing a satisfactory representation of the experimental data with only a single fitting parameter, although there are moderate quantitative discrepancies. © 1994 John Wiley & Sons, Inc.

The cure reactions of epoxy‐diamine and its composites are monitored in‐situ using the intrinsic fluorescence of the aromatic diamine, diaminodiphenyl sulfone (DDS). With a fiberoptic fluorimeter, in‐situ cure monitoring was performed via a single fiber, distal‐end probe, in neat epoxy as well as in commercial grade prepregs containing graphite fibers and DDS curing agent. The prepregs were investigated during multiply lamination in an oven. The fluorescence excitation spectra were obtained by emitting at 420 nm with a scan range of 320 to 400 nm, and the DDS peak position was determined as a function of cure time and temperature. The DDS spectra show a progressive red shift up to 24 nm when the primary amine is reacted with epoxide to become the secondary and the tertiary amines. The spectral shift of the DDS is also correlated with the extent of epoxide reaction determined by the Fourier transform infrared (FTIR) spectroscopy. Both data exhibit a linear relation, consistent with the behavior of the DDS peak shift, which increases linearly with the amine reaction. The excitation spectra also show a temperature dependency such that the amount of red shift increases with the measurement temperature in a manner that can be described by an exponential function. The temperature effects also depend on the state of cure in the sample. The temperature correction can be made by the application of an empirically developed equation. Thus, a direct comparison can be made among the on‐line data obtained under varying conditions of cure, by reducing the spectral data to any reference temperature. This intrinsic fluorescence technique is much simpler than the previously reported extrinsic fluorophore technique, which requires the addition of an extrinsic fluorophore and an internal dye, and can be applied to any commercial prepregs containing DDS, thus making it a very powerful and widely applicable monitoring tool for composite processing.

This paper describes the application of a molecular sensor for in‐situ monitoring of epoxy‐diamine cure via remote sensing fiberoptic probes. A custom‐built, fiberoptic fluorimeter allows on‐line recording of fluorescence spectra directly from the cure environment. Cure reactions in epoxy‐diamine network, such as diglycidyl ether of bisphenol A‐diaminodiphenyl sulfone (DGEBA‐DDS) or diglycidyl ether of butanediol‐diaminodiphenyl sulfone (DGEB‐DDS), have been monitored by a reactive molecular sensor, diamino azobenzene (DAA). DAA exhibits sensitive changes in UV‐visible and fluorescence spectra due to the conversion of its primary amine groups to secondary and tertiary amine groups. Fluorescence intensities are correlated with extent of reaction in epoxy network and processing parameters, such as cure temperatures and time. The use of an internal reference dye for normalization of fluorescence intensities is necessary for the quantitative correlation of spectral signals with the network structure. Variables affecting the fluorescence intensity such as excitation volume, probe location, excitation intensity fluctuation, temperature, and background intensities from optical fiber can be calibrated by normalizing the signal intensities against the internal reference. Sulforhodamine 101 was found to be a satisfactory reference dye which provides stable, readable signals over temperatures up to 200°C.

Carbon fibers and pyrolytic graphite blocks were treated with plasma of acrylonitrile (AN) and styrene (ST) monomers, using an induction‐coupled, RF‐plasma reactor. Both substrates were stable towards plasma, leading to a deposition of thin, coherent coatings of 400Å∼1000Å thickness. Both monomers produced surfaces which are substantially more polar (γc=54 dynes/cm for AN and 40 dynes/cm for ST) than the untreated surfaces (γc=32 dynes/cm). ESCA and IR studies indicate that the plasma polymers contain a high concentration of oxygen (12 percent in PPAN and 17.8 percent in PPST), in the form of CO, COOH, COC, and OH groups. Also, treated fibers exhibited slightly higher tensile strengths than the untreated counterparts, suggesting that the plasma coatings effectively heal some of the surface flaws of the fiber. The abundant surface polar groups combined with the improved tensile properties of the plasma treated fibers make them attractive reinforcements for advanced composite materials.

SYNOPSISDuring supposedly isothermal sorption/desorption of gases or vapors by solid polymers, latent heat effects alter local temperatures with the result that diffusion behavior may appear to be non-Fickian. Even when sorption curves are seemingly Fickian, spurious values of the diffusion coefficient, D, may still be inferred. These phenomena are examined in an experimental and theoretical study of incremental sorption/desorption of acetone vapor in the fluoroelastomer, vinylidene fluoride/hexafluoropropylene copolymer. A theoretical analysis developed earlier to model water vapor sorption in wool and cellulose is shown to successfully reconcile measured mass transfer rates and temperature changes. Because diffusion time scales typical of organic vapors dissolved in rubbery polymers are often comparable to or smaller than those of heat transfer between the solid and vapor phases, heat effects accompanying sorption are slowly dissipated, and neglecting them may produce large errors in D.The primary effect of a perturbation in solid temperature is on vapor solubility. In sorption, elevated temperatures reduce solubility, which slows mass uptake. In desorption, reduced temperatures increase solubility, slowing devolatilization. In both cases, the result is artifactual lowering of the apparent D.A secondary effect is associated with the dependence of D on temperature. The increase in apparent D manifests itself primarily at the start of sorption/ desorption. Its significance decreases as the increment in vapor activity decreases.' Because of the small activity increments in the experimental study, D was treated as a constant in the theoretical analysis outlined below. The phenomena investigated here were first analyzed rigorously in a study of water vapor sorption in wool and polymer^.^^^ Because of water's large heat of vaporization, its heat of sorption, A H S , is much greater than those characteristic of organic vapors in polymers, which is probably why heat effects in the latter systems have generally been neglected. However, our recent studies of organic vapor sorption in rubbery polymers495 have demonstrated significant heat effects and correspondingly abnormal sorption behavior, especially at high vapor activities.Investigations of heat effect^^^^^^^ have heretofore focussed on the determination of correct D values from sorption/desorption data, and calculation of time courses of solid-phase temperature. Large temperature excursions were measured during recently reported integral desorption measurement^.^ Apparently no previous investigation has reconciled model predictions with both experimentally measured mass transfer rates and temperature data, as we do here. In addition, we explore the utility of a single dimensionless parameter that emerges from 937