Analysis of distributive mixing during polymer blending in twin screw extruders using reactive polymer tracers

Abstract

A novel quantitative method for analyzing distributive mixing during polymer blending in a co-rotating twin screw extruder was developed. This method employed a mixing limited interfacial reaction between two reactive polymer tracers to gain direct evidence of the generation of interfacial area during polymer blending. The tracers were based on a low molecular weight amorphous polyolefin wax containing a high concentration of terminal double bonds, which were targeted for functionalization with anhydride and primary amine functional groups. These functional groups were selected because their coupling reaction is extremely fast under the conditions employed during polymer processing in extruders. A melt-phase Alder Ene reaction between the terminal double bond of the polyolefin and the double bond of maleic anhydride was used to introduce a terminal succinic anhydride functional group. Hydroboration followed by amination was completed in solution with THF to introduce terminal primary amine functional groups. For the mixing experiments, the reactive polymers were blended into polypropylene (PP) resins at a concentration of 5wt%, which was adequate for quantitative analysis of the anhydride functional group conversion using a FT-IR spectrometer.

Distributive mixing, or the generation of interfacial area, was investigated during melt-melt blending of two segregated PP streams in a co-rotating twin screw extruder. Each PP stream contained one of the reactive polymers, which come into contact and react at the growing interface. A tandem single/twin screw extruder apparatus was used to perform the melt-melt blending experiments. One stream was metered to the beginning of the twin screw extruder, and the second was melt fed at a desired downstream position using the single screw extruder. As verified using model interfacial reactions, the coupling of anhydride and primary amine functional groups was mixing limited, and the anhydride conversion was linearly related to the interfacial area available for the reaction. A slit die at the end of the twin screw extruder was used to prepare a film for FT-IR analysis of the anhydride conversion, which was a direct measurement of the overall distributive mixing performance of the melt-melt blending section of the twin screw extruder. Experiments were completed to investigate the effects of operating conditions, screw configuration, and polymer viscosity on the overall distributive mixing. In addition, a washout of the anhydride polymer tracer was used to measure the cumulative RTD in the melt-melt blending section. Distributive mixing increased with the average residence time, but it was not related to the macromixing. With respect to the kneading block geometry, the overall distributive mixing performance followed the trend of: forward > reverse > neutral at 50 g/min and reverse > neutral > forward at 100 g/min. Distributive mixing with neutral and reverse kneading block was controlled by the average residence time, the shear rate, and the fully filled volume. Conversely, the screw configuration containing the forward kneading block exhibited different trends with respect to the operating conditions, and its superior mixing at the low flow rate was attributed to possible regions of flow stagnation. In addition to the effects of operating conditions and screw design, the melt-melt blending of lower viscosity PP resins resulted in significantly greater distributive mixing.

Specially designed sampling devices were used to investigate the distributive mixing profile along the length of the twin screw extruder during melt-melt blending. The mixing profiles for screw configurations containing neutral and reverse kneading block were similar. In particular, 60 to 85% of the overall distributive mixing was completed in the conveying section prior to each kneading block. In contrast, the best local distributive mixing was completed in the forward kneading block, especially at lower flow rates. Mixing in conveying sections predominately occurred in the fully filled region due to the higher shear rates applied to the polymer melt. The fully filled fraction was the controlling variable for distributive mixing in the conveying section because it incorporated the effects of both the operating conditions and the pressure at the end of the conveying section. Local residence time measurements were completed in the conveying section and kneading block using a carbon black tracer and an IR temperature probe. This method was valid for the determination of the local average residence time in the extruder, but the distribution of residence times was affected by the local flow field at the probe position. The combination of a longer local average residence time and a polymer backup in the conveying elements prior to the partially filled forward kneading block suggested the existence of stagnant flow regions. These stagnant flow regions were also manifested by a significantly larger melt temperature rise across the forward kneading block. Flow through the high shear rate gaps in the forward kneading block caused its superior local distributive mixing as well as the higher melt temperature rise.