Synthesis and characterizations of novel solid polymer electrolytes with carbon nanotube as a filler / Suriani Ibrahim

• Main Author: Suriani , Ibrahim
• Format: Thesis
• Published: 2011
• Subjects: T Technology (General); TJ Mechanical engineering and machinery

Nanocomposite solid polymer electrolyte based on poly (ethylene oxide) (PEO) as a host matrix, doped with lithium hexaflurophosphate (LiPF6) salt, plasticized with ethylene carbonate (EC) and dispersed with amorphous carbon nanotubes (αCNTs) as nanocomposite filler were prepared by solution-casting technique. The polymer electrolytes were characterized by Impedance Spectroscopy (IS) to obtain the composition of additives which gives the highest conductivity value for each system. At room temperature, the highest conductivity was found to be 1.30 × 10-3 Scm-1 with a composition of PEO - 20wt% LiPF6 - 15wt%EC and αCNTs - 5wt%. The ionic conductivities of the nanocomposite polymer electrolytes increased with temperature and obeyed the Vogel-Tamman-Fulcher (VTF) law. The crystallinity, chemical reaction, thermal behaviour, morphologies and optical properties of the nanocomposite polymer electrolytes were examined by using FTIR, XRD, DSC, TGA, UV Vis and PL Spectrophotometer. FTIR indicates the existence of interactions among PEO, LiPF6, EC and αCNTs. XRD and DSC studies indicate that the increased conductivity was due to the increase in amorphous content. This enhanced the segmental flexibility of polymeric chains and the disordered structure of the electrolytes. TGA studies indicate that the stability of nanocomposite polymer electrolytes was decreased when 5wt% of αCNTs filler was added into plasticized polymer electrolytes. The SEM micrographs showed surface changes when LiPF6, EC and αCNTs were added into the polymer systems. The optical band gaps for PEO - 20wt% LiPF6 - 15wt% EC - 5wt% αCNTs was 4.60 eV. A neural network model was developed which could predict the Nyquist plot of nanocomposite polymer electrolyte system (PEO - LiPF6 - EC - CNT). The Bayesian neural network was found to be successful in predicting the experimental results, which climates the time-consuming studies.