Processing, Structure and Properties of PA6/Carbon Composites

Access to files

•   FULL-TEXT.PDF (pdf)

Abstract

ABSTRACTThe aim of this research was to study the structure-property relationships of polyamide 6 (PA6) micro- and nanocomposites produced using two particulate carbon fillers; namely graphite, G, and graphite nanoplatelet, GNP. The GNP and G are similar in lateral dimensions, but differ greatly in their thickness (by at least an order of magnitude), the size-scale of which defines G as a micron-scale filler and GNP as a nano-scale filler, and consequently, the specific surface area available for matrix interaction and the aspect ratio of each filler also differed significantly. Size scale was considered in the choice of processes to incorporate the G and GNP into a PA6 matrix. Firstly, in situ polymerisation using anionic polymerisation of epsilon caprolactam (EC), which enabled polymer/carbon interactions on the molecular scale and, secondly, melt extrusion using commercial grade PA6 as the matrix. For the G-based studies, composites with at least five stepwise incorporations between 5-25 wt. % loading were produced; whereas GNP was incorporated at loadings an order of magnitude lower (0.5-2.5 wt. %) reflecting their difference in size-scale. For both G and GNP, composites were produced via in situ polymerisation and by melt processing using a Haake Minilab bench-scale twin screw extruder, (TSE) in two groups which compare the effects of processing conditions. In in situ polymerisation, processing conditions designed to deliver the same sonication power were used and were coded as 40/10 (a sonication amplitude of 40% was applied for 10 minutes to disperse carbon filler in the molten EC) and 20/20 (sonication amplitude of 20% for 20 minutes). Similarly, in melt extrusion processing conditions designed to deliver the same strain magnitude were used and were coded as 200/3 (screw frequency of 200 rpm was applied for 3 minutes) and 100/6 (100 rpm for 6 minutes). A ninth G-based system (GL) was produced using an industrial lab-scale TSE for occasional cross-comparison with the bench-scale processed systems. All systems were characterised using thermal analysis (DSC, TGA, DMTA), tensile testing, impedance spectroscopy and electron microscopy. Overall, the best property increases were observed for the 20/20 processing conditions for the in situ polymerised systems and the 100/6 processing conditions for the melt extruded. The 20/20 in situ processing condition produced a dispersed state where particle aspect ratio was retained and less particle fragmentation occurred, the latter also giving higher reaction rates (and hence a higher molecular weight matrix) compared to the 40/10 processing conditions. In the in situ polymerised 20/20 GNP systems the best overall mechanical properties were obtained using the lowest loading applied, 0.5 GNP wt. For example, in tensile tests although addition of 0.5 GNP wt. % did not give the best tensile modulus, it gave the best tensile strength, yield stress and elongation at break. Tensile data also indicated that the dispersed states of the carbon fillers achieved using the 100/6 processing conditions are superior to that obtained using the 200/3 processing conditions. For example, tensile strength (TS) values increased above unfilled PA6 in the NP 100/6 system, where all the average TS values of the composites are higher than that of the unfilled PA6. In addition, in the G-based systems, it was only with the 100/6 processing condition that some TS values higher than the unfilled PA6 were obtained. Overall, despite the advantages in size scale of the GNP over G, the GNP nanocomposites did not consistently exhibit better properties since the growth in property values either increased slowly or fluctuated with GNP loading. In terms of electrical conductivity, with the exception of the GL system, where electrical conductivity sufficient for electrostatic applications was attained at 15 G wt. %, all other melt processed composites of both the G and GNP remained as electrical insulators with up to 50 G wt. % loading and 25 GNP wt. % loading in both the 200/3 and 100/6 processing conditions. For the in situ polymerised composites, however, electrical conductivity percolation thresholds values of <10 wt. % were measured in both the G and GNP composites, although GNP did not show a clear advantage over G despite the difference in their order size-scale.

Bibliographic metadata

Institutional metadata

Record metadata