Processing, Structure and Properties of Poly (ethyleneterephthalate)/Carbon Micro- and Nano-composites

  • Basheer Alshammari

Student thesis: Phd

Abstract

Incorporation of conductive carbon fillers into polymer matrices can improve electrical,thermal and mechanical properties of the resulting composites. In this work, three differentconductive carbon fillers were used; i.e. graphite, graphite nanoplatelets (GNP) and asreceivedmultiwall carbon nanotubes (A-MWCNT). In addition, A-MWCNT were modifiedusing mixed acids and named as T-MWCNT. These four fillers were incorporated into poly(ethylene terephthalate) (PET) to prepare four types of PET/carbon micro- and nanocomposites.These composites were prepared by melt compounding using a Haake Minilabextruder equipped with a co-rotating twin screws. The extruded samples were compressionmoulded to films of 1 mm thickness and were subsequently quenched to obtain lowcrystallinity samples. The extruded samples were also injection moulded to obtain dumbbellshaped specimens. The electrical, morphological, thermal and mechanical properties of thesecomposites were studied and characterized as a function of carbon filler types and contentsusing a wide range of analytical and testing techniques: namely; impedance spectroscopy,DSC, TGA, SEM, TEM, FTIR, DMTA and tensile testing. The results demonstrated that theaddition of graphite, GNP and A-MWCNT produced electrically conductive composites andthat the conductivities were found to be dependent on several factors; including filler type,filler content and processing conditions. The PET/A-MWCNT nanocomposites showed anexcellent electrical conductivity (~ 0.2 S/m at 2 wt. % A-MWCNT) with a low percolationthreshold (Fc ~ 0.33 wt. %). In contrast, PET/T-MWCNT nanocomposites displayed similarelectrical conductivity to that of pure PET and no percolation threshold was observed in thiscase (until 2 wt. % of CNT), this was attributed to the acid treatment which disrupted theinherent electrical conductivity of the CNT and also reduced their aspect ratio. However, TMWCNTshowed better dispersion and distribution into the PET matrix as well as reducedCNT-CNT interactions and therefore do not as readily form network structures. This resultedin better mechanical properties in comparison to the PET/A-MWCNT nanocomposites. Interms of processing, increasing screw speed during mixing was found to enhance theelectrical conductivities of PET/carbon nanocomposites (GNP and A-MWCNT), but onlyabove the percolation thresholds values, by ~ 2 - 3 orders of magnitude. However, nosignificant change was observed in the electrical conductivities of PET/graphitemicrocomposites. All the carbon fillers, with different dimensions, were found to act asnucleating agents for the PET matrix and hence accelerated crystallization and increased thedegree of crystallinity. CNT were found to accelerate the crystallization at lower loadingscompared to GNP and graphite. In addition, it was found that quenched PET and compositesamples were not fully crystallized after processing and therefore (cold) crystallized duringthe first heating cycle in both DSC and DMTA, as indicated by crystallisation peaks duringthe DSC first-heat and a rise in storage moduli above Tg during the DMTA first heat. Ingeneral, TGA showed that carbon fillers improved the resistance to thermal and thermooxidativedegradation under both air and nitrogen atmospheres. However, a reduction inthermal stability was observed for the composites containing T-MWCNT in air. The carbonfillers increased the storage and tensile moduli of the composites compared to pure PET.However; tensile strength and elongation at break were reduced except for compositecontaining T-MWCNT which showed no significant change at lower loadings. The tensile23moduli of nanocomposites were predicted using Halpin-Tsai models, which showed goodagreement at low loadings of A-MWCNT (smaller or equal to 0.2 wt. %) and GNP (smaller or equal to 2 wt. %). However,poor agreement was observed at higher loadings of fillers where the composites displayedreduced reinforcement efficiency. This correlates with results from
Date of Award1 Aug 2015
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorArthur Wilkinson (Supervisor) & Ian Kinloch (Supervisor)

Cite this

'