THE SCHOOL OF MATERIALS SCIENCE AND ENGINEERING

GEORGIA INSTITUTE OF TECHNOLOGY

Under the provisions of the regulations for the degree

DOCTOR OF PHILOSOPHY

on Monday, November 30, 2020

10:00 AM

via

Bluejeans Video Conferencing

https://bluejeans.com/472186811

will be held the

DISSERTATION DEFENSE

for

Morgan Watt

"Effect of Network Formation and Particle Morphology on the Electrical Properties of Composites"

Committee Members:

Prof. Rosario A. Gerhardt, Advisor, MSE

Prof. Kyriaki Kalaitzidou, ME

Prof. Jonathan Colton, ME

Prof. Satish Kumar, MSE

Prof. Jud Ready, MSE

Abstract:

Composites are multi-phase systems that have a filler and a matrix material. Their properties are a summation of the properties of the individual materials plus they sometimes have extra properties not from the original materials. The overall properties of composites vary based on multiple conditions including processing and particle morphologies which will be explored in this study. Commonly composites are made and studied for structural applications. In this study, MWCNT/PMMA, SiC/PMMA, and SiC/Glass composites were fabricated to investigate the effect of particle morphology and processing steps on the network formation and resultant microstructural and electrical properties. Understanding of these effects would allow for tailoring composites for specific applications and properties and avoid unnecessary trial and error.

The MWCNT/PMMA composites were made using three different mixing methods (mechanical, melt, and solution) to create three different networks of conductive fillers within an insulating matrix (segregated network, distributed network, and agglomerated microstructures) and study their effect on the electrical properties. Of the three mixing methods, mechanical mixing resulted in composites with the lowest percolation threshold (0.087 phr). When the impedance data was fit to equivalent circuits, before percolation, the circuits were the same for all composites. After percolation, the circuits varied but were very similar since the same two precursors were used. The melt mixed samples were the most different due to the even dispersal of the MWCNT which allowed PMMA to have a stronger effect.

To compare the effect of the filler particle morphologies, three types of SiC (micron-sized, nano-sized, and whisker) were used to make SiC/ PMMA composites by mechanical mixing. The nano- SiC/PMMA composite formed the most defined segregated network composite, provided stable results, the highest conductivity, and a low percolation threshold of 2.35 phr. Conversely, the SiCw /PMMA composite had the high percolation threshold of 10phr, while the micron-SiC/PMMA composites had unusual results. Despite these differences, the SiC/PMMA composites all had the same circuits representing them since they were made with the same type of material and processing. The differences in values of the circuit elements present directly correlates to the interaction between the matrix and filler.

Two sizes of anisotropic SiC (whisker and nanowire) were incorporated into glass matrix composites to compare the effect of anisotropic filler size. It was theorized that the smaller size of the nanowire would allow it to segregate into the grain boundaries easier and create a lower percolation threshold due to  a higher aspect ratio aiding the chance of making network connections. This case proved true in that the SiCnw/glass composites did have a lower percolation threshold than SiCw/glass composites. However, the conductivity value was lower, and samples became too fragile to continue increasing the concentration. Circuit fitting showed that SiCw / glass composites had a variety of circuits as concentration increased. Conversely the SiCnw/glass composites only needed two circuit models to represent their electrical properties as concentration was changed.

During the course of this study, it was discovered that the SiC/glass samples had a strong sensitivity to the humidity at the time of testing. PMMA composites were found to be unaffected by the humidity. From 0-2.5 phr, the conductivity increased with increasing humidity. For 7.5 and 12.5 phr, there was an initial sharp increase then a gradual decrease in conductivity with increasing humidity. For better understanding of the mechanism of conduction with both water and SiC contributing, normalized M” and Z” were plotted together. The separate M” and Z” peaks indicated both localized and long-range conductivity.

This study has demonstrated that very small differences in material characteristics and processing method used to fabricate the composites may result in very different properties. This study provided direct comparisons to verify the changes observed along with equivalent circuit fitting to correlate the contributions of the filler and matrix materials.