Jérôme Dupuy

Kinetics modeling of a modified epoxy-amine formulation cured by thermal and microwave energy

The reaction kinetics of a rubber-modified epoxy formulation cured by microwave or thermal energy were investigated. Two phenomenological models were developed to predict the time and temperature dependence of the conversion for the neat and the modified systems. Good agreement was observed between the kinetic models and experimental results generated by chromatographic and calorimetric techniques. The same kinetic behavior was observed whatever the curing process (conventional or microwave heating).

Modification of Ziegler‐Natta catalysts by cyclopentiadienyl‐type ligands: Activation of titanium‐based catalysts

Conventional Ziegler-Natta catalysts, based on TiCl 4 supported on MgCl 2 , were modified by substituting a part of the chlorides by cyclopentadienyl (and derivatives) ligands. Although these catalysts are very active (activities up to 10 5 g PE/g catalyst/h) they exhibit a conventional Ziegler-Natta behavior (methylaluminoxane is not necessary, polyethylene produced with a rather broad molar weight distribution, low sensitivity to hydrogen). It was attributed to cluster effects: an increase of the number of conventional TiCl 4 active sites by adding cyclopentadienyl ligands on titanium neighbors.

FT-NIR monitoring of a scattering polyurethane manufactured by reaction injection molding (RIM): Univariate and multivariate analysis versus kinetic predictions

Near InfraRed (NIR) spectroscopy was used to monitor in situ a polyurethane synthesis during a RIM process. Univariate and multivariate analysis of transmittance spectra were used to calculate the reaction extent. A very good agreement was observed between multivariate analysis (PCA), univariate (Beer-Lambert) analysis and kinetic predictions. It was demonstrated that, in this case, PCA methods can provide a good estimation of the reaction extent versus time, without time-consuming calibration. The spectral range of PCA has to be carefully chosen.

In situ monitoring of polyurethane cure using fibre-optical FT-NIR spectroscopy

In-line monitoring of the reaction extent of polyurethane during a reactive injection moulding (RIM) process is carried out using fibre-optic near infrared (NIR) spectroscopy. Up to 250 transmission spectra are recorded during the reaction. Univariate and multivariate analysis of transmittance spectra were used to calculate the chemical conversion. A good agreement is observed between first principal component of principal component analysis (PCA), and univariate (Beer—Lambert) results. It is observed that, in this case, the PCA method can provide a good practical estimation of the time-concentration profile during the reaction, without the need of the time-consuming calibration methods. The scores of PC1 are merely linearly correlated to the level of conversion and contain enough information for the quantitative analysis. As expected interactions and hydrogen-bonding play an important role. Hence the spectral region of PCA analysis has to be carefully selected to obtain a good agreement with the Beer—Lambert law. The NIR spectroscopy and the PCA are easy-to-use techniques for on line monitoring of polyurethane reactions and these results open up a low cost effective opportunity for monitoring the fast RIM process.

Kinetics modeling of the reaction for an Epoxy/PMMA/MMT ternary system

Abstract The effect of thermoplastic PMMA and clay Cloisite 30B addition on the cure kinetics of an epoxy/amine thermosetting system was investigated using differential scanning calorimetry (DSC) and modelled using a single kinetic model. For obtaining a good agreement between the experimental results and theoretical modeling, it was necessary to separate the ternary system in two formulations: matrix/PMMA and matrix/clay systems. It appeared that the addition of PMMA delays the reaction up to the phase separation phenomenon due to a dilution effect of the reactive species. In opposition, the presence of clays accelerates the reaction (probably due to a catalytic effect of some metals ions introduced with the clay) but it has no sensitive effect on the cloud point conversion in the presence of PMMA. The modeling results are in good agreement with those experimentally obtained.

Tuning morphologies of thermoset / thermoplastic blends Part 1: Kinetic modelling of epoxy-amine reactions using amine mixtures

Abstract Different epoxy amine networks based on a mixture of diamine hardeners have been studied and their kinetic behaviour was modelled in order to understand the global and particular behaviour of each diamine in such blends. 4,4’-Diaminodiphenylsulfone (DDS) and 4,4’-methylenebis-(3-chloro-2,6- diethylaniline) (MCDEA) were used for this study by varying the DDS molar ratio from 0% to 50%. The determined kinetic model allowed us to calculate the composition for each diamine of the reactive epoxy-amine system in the whole range of epoxy conversions. Then the influence of 10% thermoplastic additive was checked and it was found to be negligible in the kinetic behaviour.

Tuning morphologies of thermoset/thermoplastic blends - Part 2: Phase separation of poly(vinyl methyl ether), PVME, using amine mixtures as thermoset hardeners

Abstract Part 2 of the study investigates morphologies of epoxy thermoset / thermoplastic blends obtained with formulations of the thermoset hardeners. The thermosetting matrices are composed of one epoxy resin crosslinked by a mixture of two aromatic diamine hardeners, namely MCDEA+DDS or MDEA+DDS (DDS is 4,4’-diaminodiphenylsulfone, MCDEA is 4,4’-methylenebis-(3-chloro-2,6- diethylaniline), MDEA is 4,4’-methylenebis-(2,6-diethylaniline). The blends are made at a fixed concentration of thermoplastic (poly(vinyl methyl ether), PVME, 10wt%) whereas three different cure temperatures are chosen and the matrix composition is varied by the ratio of DDS. DDS is a PVME-insoluble diamine whereas MDEA or MCDEA are PVME-soluble diamines. The domain size of the PVME nodules is tuned from a few micrometers down to sub micron sizes, typically between 50 and 80 nm. We attempt to explain this evolution thanks to the difference in the reactivity of the amines with the epoxy oligomer coupled to the difference in miscibility of each amine towards PVME. The kinetic study developed in Part 1 of this work is used to calculate the concentrations of each amine along the reaction time and their rate of incorporation in the polymer network. The latter parameter is shown to be the limiting factor for obtaining a nano phase separation.

Kinetic parameter estimation using an adiabatic reaction, with validation using NIR spectroscopy, 1 Application to a macrodiol / isocyanate polyurethane formulation

Abstract A kinetic model of the cure of a polyurethane formulation is determined by the evolution of the conversion calculated during adiabatic experiments. The identified kinetic parameters are validated with two different methods: i) by comparing the simulated and measured evolutions of the conversion rates during these adiabatic reactions, ii) by comparing the simulated and measured (using NIR spectroscopy) evolution of the conversion into a mould during a RIM process.